LLM serving is increasingly dominated by decode attention, which is a memory-bound operation due to massive KV cache loading from global memory. Meanwhile, real-world workloads exhibit substantial, hierarchical shared prefixes across requests (e.g., system prompts, tools/templates, RAG). Existing attention implementations fail to fully exploit prefix sharing: one-query-per-CTA execution repeatedly loads shared prefix KV cache, while one-size-fits-all tiling leaves on-chip resources idle and exacerbates bubbles for uneven KV lengths. These choices amplify memory bandwidth pressure and stall memory-bound decode attention.

The hardware scheduler on NVIDIA GPUs is highly inefficient in utilizing micro-architectural hardware resources. It places blocks from the same kernel within the same GPU Streaming Multiprocessor (SM) core, resulting in a stacking colocating problem, where identical blocks are placed within the same SM core, saturating only a subset of intra-SM hardware resources while leaving others underutilized. The primary challenge in addressing this issue is that the NVIDIA hardware is closed-source, preventing us from directly modifying the hardware scheduler. To bridge the semantic gap between the resource demands of kernels and the scheduler, we introduce µ Share, which enables intra-SM scattered colocating of kernels through a non-intrusive half-plus blocksize shaping method. It shapes the blocksize of kernels to a halfplus blocksize (i.e., slightly more than half of the SM's thread capacity), scattering identical blocks of the same kernel across different SMs. It further adopts a time-shifted launching method to reduce intra-SM resource contention. Compared to state-of-the-art systems, µ Share does not require intrusive modifications to hardware or kernel code, yet it can still improve inference throughput by 26.90%-54.09% and increases low-level hardware utilization by 38.53%-61.15%.

Device-free Wi-Fi localization is a promising technology to localize users who do not carry smart devices. The basic idea is to separate and analyze the signals reflected off human body from the multi-path signals. However, previous works could only localize moving users, because they can not distinguish the signals reflected off static users or objects like walls and furniture. This paper presents the Respiration Localization system,RespLoc, which for the first time enables device-free Wi-Fi localization system for static users. To recognize static users, the key insight is that people can breathe but objects cannot. However, it is non-trivial to extract user’s location from the respiration signal, because the respiration signal is significantly weaker than the regular motion signal. To this end, we propose the equivalent analysis method. Instead of using traditional signal separation, which suffers from severe noise due to residual signal components, we propose to construct an equivalent signal with the following properties: First, the equivalent signal follows the same variation law as the respiration signal; Second, the equivalent signal is not affected by irrelevant static signals. Based on this equivalent signal, we are able to resolve location features from the multi-path signals directly without separating them. We implementRespLocon commodity Wi-Fi devices, and extensive experimental results demonstrate thatRespLoccan localize static users with a median error of 0.89 meters.

Massive MIMO has emerged as a cornerstone technology for 5G-Advanced and future 6G networks, yet its practical deployment remains limited by hardware cost and power consumption. Switch-based architectures, which share a small number of RF chains among many antenna elements, provide a scalable alternative, but create a new bottleneck: only a subset of antennas is observable at any given moment, leaving the channel state of the remaining elements unknown. Lacking this information prevents the system from exploiting advanced physical-layer functions such as digital beamforming or multi-stream MIMO. In this paper, we present ARGUS, a generative channel reconstruction framework that infers the CSI of unobserved antennas from partial observations. The key idea is that all antenna responses are governed by the same underlying wireless propagation environment, enabling the task to be formulated as a generative inference problem. We employ a variational autoencoder to capture the latent spatial structure and reconstruct unobserved channels through sampling. Extensive experiments show that our reconstructed CSI incurs less than 2.5% achievable rate loss, and real-world measurements demonstrate a more than 90% antenna-selection match rate, confirming the practicality of the proposed approach.

Accurate CSI feedback is crucial for Massive MIMO systems, yet it remains challenging in resource-constrained IoT scenarios due to strict pilot overhead constraints. Under such extreme data sparsity, conventional data-driven methods often fail to generalize. To address this, this paper proposes a Physics-Aware Multi-Feature fusion approach (PAMF), a deep learning framework that systematically integrates data-driven learning with wireless propagation physics. The framework includes dedicated feature extractors based on spatial, frequency-domain, statistical, and physics-based methods, along with a deep residual reconstruction network. A key innovation lies in its dual-level physical constraint mechanism, which incorporates domain knowledge at both the feature and loss levels to ensure physically plausible channel estimates. By leveraging multi-modal feature representations and physics-aware optimization, PAMF effectively recovers the complete channel matrix from sparse pilot signals, which not only improves feature discrimination, but also leads to greater robustness particularly under the dynamic conditions typical of urban mobile networks. Experimental results demonstrate that the proposed method consistently outperforms existing approaches across diverse datasets including MIMO configurations of various scales, different modulation schemes, and real-world Wi-Fi CSI Specifically, on real-world Wi-Fi data, PAMF achieves an NMSE of 0.1002, approximately 62% lower than the ChannelNet baseline (0.2669). Overall, this study contributes a practical and physical-aware framework for channel estimation, paving the way for more reliability and efficiency next-generation wireless systems, with direct implications for large-scale IoT deployments.

The blockchain-based federated learning framework has garnered widespread attention to ensure data privacy and trustworthy computing in IoT devices. Improving the accuracy and scalability are paramount for the increasing demands of IoT tasks. However, existing solutions, such as SIFL and ChainFL, utilize traditional single-chain architecture and fixed models, which exhibit limitations in terms of model generalizability and scalability. To overcome the above problems, this paper proposesFedShard, an FL framework that integrates a layered dual-track incentive mechanism and a privacy knowledge distillation module.When implementingFedShard, we address two technical challenges: (1) to ensure efficient training among the numerous clients in the sharding architecture, we propose a layered dual-track incentive mechanism that provides both long- and short-term rewards; and (2) to enhance the framework’s convergence and privacy in sharding-based FL, we design a knowledge distillation module that incorporates local differential privacy. Furthermore, we propose a bucket-based hash ring to manage clients, enabling the framework to adapt to dynamic network environments. To validate the framework’s generalizability and scalability, we implementFedShardon a 48-core high-performance server using Fabric to conduct both on-chain and off-chain experiments. Our comprehensive experiments, comparingFedShardwith SIFL, PEFL, and ChainFL, reveal that our solution outperforms the state-of-the-art methods by achieving a notable 29% increase in accuracy and a 22% improvement in throughput.

This paper presents ArmNet, a mobile sensing system for capturing the trajectory of the wrist using measurements from wrist-worn IMU devices, specifically targeting robust interaction within Internet of Things (IoT) ecosystems. Unlike existing solutions that directly map IMU data to joint positions, ArmNet integrates physical kinematic constraints with neural network modeling. This hybrid approach is crucial for resource-constrained IoT devices where computational overhead and sensor limitations are primary concerns. Specifically, kinematic priors capture spatial dependencies between the elbow and wrist, generating physics-guided intermediate features that reduce the solution space. These features, together with raw IMU data, are fed into a recurrent neural network to learn joint displacement vectors, which are then integrated into continuous trajectories. Extensive experiments show that ArmNet achieves robust and accurate arm tracking, generalizing well across users and motion patterns, thereby enabling a new modality for seamless human-computer interaction in smart environments.

Facial Expression Recognition (FER) serves as a foundational sensory interface for Human-Centric IoT, supporting applications such as smart healthcare monitoring and affective intelligent environments. However, real-world performance is often hindered by theSemantic Gap, where models confuse visually similar expressions that arise from fundamentally different physiological muscle movements. To bridge this gap, we propose the Physical-Semantic-Aware Multimodal Framework (PSM-FER), which introduces 3D Blendshape (BS) coefficients as explicit physical priors to encode high-level muscle motion semantics. Our framework utilizes two synergistic pathways:Direct Physical Gating(DPG) for robust feature modulation andSemantic-Guided Spatial Attention(SGSA) for anatomical spatial recalibration. Additionally, an auxiliary physical regression task enforces anatomical consistency by regularizing the latent features to follow underlying biomechanical laws. Extensive experiments on the RAF-DB dataset demonstrate that PSM-FER achieves an accuracy of 92.37%, establishing a robust and interpretable foundation for affective sensing in complex IoT ecosystems.

Memory disaggregation architecture presents unique challenges in ensuring data consistency due to the limited computational power available at memory servers. One line of pessimistic solutions utilizes lock tables to deal with this challenge. Since the lock release signal cannot be immediately synchronized to the compute server, such solutions can only achieve suboptimal memory utilization. To address the aforementioned limitation, another line of solutions operates optimistically, polling the memory server until the operation completes successfully. However, these solutions trigger a massive number of unnecessary retries, resulting in performance collapse in high contention scenarios.high memory utilization while avoiding the need for retries. Our key idea is that compute servers proactively predict the appropriate retry interval and schedule accordingly. By analyzing the properties of requests on memory disaggregation, we find that the retry interval mainly consists of network fluctuations and host congestion delay. ECC design builds upon insights gained from this analysis. We have integrated ECC into Clover—a state-of-the-art memory disaggregation system and evaluated it through simulations and testbed experiments. Our testbed results show that ECC improves throughput by 50.8% over Clover.

Function-as-a-Service (FaaS) is becoming increasingly prevalent, and the number of daily service requests in our system ByteDance has even reached 49.6 billion. These requests exhibit “3H” characteristics:high concurrency,high burstiness, andhigh fluctuation, causing the instance scheduling system to face severe bottlenecks even earlier than the data plane.In this paper, inspired by the idea of on-demand scaling of FaaS applications, we proposeGalaxy, a highly scalable instance scheduling system for FaaS platforms.Galaxydecouples the instance scheduling system into functions and introduces a top controller to dynamically adjust the number of instances for each function based on the request load, effectively eliminating bottlenecks under high concurrent requests. We evaluateGalaxywith real-world workloads and the experimental results show thatGalaxycan complete the scaling of over 89,935 instances within 1 second, excluding cold start overhead.Galaxyincreases the scheduling throughput by 8.6 ~ 12.7× compared to the existing control plane.

The rapidly expanding blockchain ecosystem faces a critical interoperability crisis—evidenced by siloed total value locked and bridge losses. Existing interoperability solutions are trapped in a paradigm dilemma: centralized relays offer performance but introduce single points of failure, while decentralized relay-chains provide security but suffer from poor scalability and rigidity. To overcome these limitations, we introduce theHybrid Relay Architecture, a new decentralized interoperability paradigm that reconciles the efficiency of centralized relays with the security guarantees of decentralized systems via dynamic role stratification and probabilistic trust coordination. We instantiate this paradigm asCelestial, which realizes hybrid relaying through (1) a unified interchain data unit for coordination, (2) a geo-aware topology that enables adaptive role assignment, and (3) a proof of cross-chain transmission protocol that incentivizes honest node participation. Implemented in 12K+ LoC and evaluated across 90+ nodes spanning three continents,Celestialachieves 3,340 TPS with sub-600ms P99 latency, recovers from faults 6.4× faster than state-of-the-art systems, and reduces verification cost by 63% under adaptive attacks.

Function caching is one of the fundamental techniques in FaaS platforms to alleviate coldstart overhead. However, as cache instances consume significant cloud resources (e.g., memory), it is challenging to balance function performance and cache cost. Current systems use simple and rudderless cache polices with a “local cache control” design, which ignores function characteristics such as workload skewness from hot functions and results in either cache contentions or cache resource waste.In this paper, inspired by software-defined networks, we proposeFlame, an efficient cache system to manage cached functions with hotspot-aware instance scheduling and cache allocation. It consists of a two-layer design. Firstly, by decoupling the cache control plane from worker nodes and introducing a centralized cache controller,Flamecan schedule functions from a global view of the cluster’s status, thereby reducing inter-node workload skew. Second,Flamedivides the prior monolithic cache pool within each node into multiple partitions and dynamically assigns them to different hot functions, thereby further mitigating intra-node cache contention. Experimental results from realworld workloads show thatFlamecan reduce cache resource usage by 36% on average while improving function performance by nearly 7× compared to the state-of-the-art method.

The dynamic nature of tasks and environments in Internet of Things (IoT) systems require deep learning models to continuously retrain on evolving data to ensure their effectiveness. Existing continual learning (CL) methods aim to mitigate catastrophic forgetting, where the model loses knowledge of previous tasks when learning new ones. However, these methods often ignore the memory resource competition caused by the parallel execution of multiple applications, which limits the realworld IoT application of CL in resource-constrained edge devices. In this article, we propose EDCL, a novel approach that enhances the training efficiency and model accuracy of CL methods while ensuring the uninterrupted operation of high-priority inference programs. Specifically, we first implement a custom batch sampler that can dynamically load batches and measure the memory usage and training time recorded via offline profiling. In the online stage, by monitoring the resource consumption of high-priority programs, EDCL can dynamically select batch policies that meet resource constraints and facilitate efficient training. Additionally, we propose an adaptive hierarchical buffer swap method to enhance the model’s ability to retain previously learned knowledge and mitigate forgetting. Extensive experiments show that EDCL effectively balances training efficiency and model accuracy while preventing high-priority inference programs from failing due to memory contention, demonstrating promising performance compared to baselines.

Millimeter-wave (mmWave) 5G-V2X communications face significant challenges in real-time beam alignment within high-mobility vehicular networks. While environmentaware beam prediction methods mitigate channel estimation overhead, their efficacy is severely compromised by modality data loss stemming from lighting variations, adverse weather, or sensor failures. To address this issue, we propose a Cross-modal Losstolerant Beam Prediction model (CLBP). CLBP robustly fuses RGB camera and LiDAR data, employing a novel cross-modal attention mechanism to achieve resilient feature alignment across these heterogeneous modalities. Furthermore, a Branch Features Dynamic Fusion (BFDF) module adaptively reweights modality features, suppressing noise from degraded inputs and promoting effective information propagation to enhance resilience. To facilitate realistic evaluation, we introduce a Data-Conditioned Missingness Mechanism (DCMM), which augments the DeepSense 6G V2V dataset with meticulously simulated sensor failure scenarios. Experimental results demonstrate CLBP's superior performance, achieving 94.48% Top-5 beam prediction accuracy even under 10% modality loss, and a 29% reduction in average power loss compared to baseline methods. These findings demonstrate CLBP's significant robustness in dynamic vehicular environments and its capacity to maintain consistent, high-performance beam prediction despite challenging data imperfections.

360$^\circ$video streaming emerges as an innovative video presentation form that offers users an immersive and interactive experience, where the quality of experience (QoE) is a vital indicator to measure user viewing perception. In multicategory 360$^\circ$video streaming, existing QoE-driven approaches typically assume a fixed request distribution to enhance users' average QoE, prioritizing the optimization of edge caching and bitrate selection decisions for the video category with a larger request number. Inevitably, these unfair approaches would lead to average QoE reduction in real-world scenarios, in which the request distribution exhibits significant variations and is challenging to predict accurately. To this end, we propose a fairness-aware 360$^\circ$video streaming strategy in cloud-edge collaboration networks for improving users' average QoE. Specifically, we first formulate the joint edge caching and bitrate selection problem as a multi-agent cooperative input-driven Markov decision process to maximize users' average QoE and guarantee QoE fairness for users. Subsequently, we devise an adaptive learning-based multi-agent deep reinforcement learning (MADRL) approach, which can adaptively adjust the learning rate of each agent according to the dynamic user request distribution, thus helping agents make optimal decisions. Finally, experimental results on real-world datasets show that the proposed algorithm significantly improves users' average QoE while ensuring QoE fairness for users.

Distributed stream processing has recently gained much interest due to the need of extracting meaningful results from continuous data stream. To keep the extracted results fresh, the underlying network flows are often required to transmit packets continuously. Otherwise, these results will become stale, and their staleness is determined by the slowest flow. At this point,coflowscan be semantically comprised. Hence, efficient coflow transmission is critical for streaming applications. However, prior coflow-based solutions have significant limitations. They use a one-shot performance metric—CCT (coflow completion time), which cannot continuously reflect the staleness of the output results for a streaming application. To this end, we propose a new performance metric—coflow age(CA), for coflows generated by distributed streaming applications. The CA tracks thelongest time-since-last-serviceamong all flows in a coflow. In such a context, we consider a data center network with multiple coflows that continuously transmit packets between their source-destination pairs and address the problem of minimizing the average long-term CA while simultaneously satisfying the throughput constraints from the coflows. To solve this problem efficiently, we design a randomized algorithm and a drift-plus-age algorithm, and show that they can make the average long-term CA to achieve nearly two times and arbitrarily close to the optimal value, respectively. Through extensive simulations, we further demonstrate that both of the proposed algorithms can significantly reduce the CA of coflows, without violating the throughput requirement of any coflow, when compared to the state-of-the-art solution in both scenario with the packet arrival probability being known and unknown a prior.

In-network aggregation accelerates distributed training by offloading gradient aggregation to the programmable switches. However, in multi-tenant learning environments, contention for limited switch memory can cause memory overflows that significantly degrade aggregation throughput. To mitigate memory overflow, prior work has proposed selective in-network aggregation, which allocates switch memory to a subset of jobs based on memory availability. This approach classifies jobs into INA jobs (which perform in-network aggregation) and PS jobs (which perform in-server aggregation). Despite these efforts, network congestion occurs and slows INA jobs, resulting in inefficient memory utilization and reduced aggregation throughput. In this paper, we present FlexINA, which rethinks selective in-network aggregation to deliver high aggregation throughput. The first challenge is how to avoid excessive memory under-utilization by INA jobs during network congestion. FlexINA introduces INA-aware congestion control, prioritizing reducing the sending window of PS jobs during network congestion. The second challenge is how to allow PS jobs to utilize under-utilized aggregators without affecting the aggregation of INA jobs. FlexINA implements an adaptive head-tail aggregation, optimizing memory usage by combining static head mapping for INA jobs (to use allocated head memory) and dynamic tail mapping for PS jobs (to use under-utilized tail memory). We implement a FlexINA prototype and evaluate it on both a small-scale testbed and in large-scale simulation experiments. Our evaluation shows that FlexINA improves aggregation throughput by up to$1.9\times $and$1.4\times $compared to selective in-network aggregation NetPack (ATP) and NetPack (A2TP), respectively.

Blockchain-based Federated Learning (BCFL) has attracted considerable attention in the intelligent IoT domain for its privacy-preserving and decentralized characteristics. Depending on their applicable scenarios, BCFL frameworks are categorized into two types: synchronous and asynchronous. However, synchronous BCFL struggles with low efficiency in heterogeneous IoT environments, while asynchronous BCFL suffers from slow convergence speed. In additional, Both BCFL incur significant resource consumption from blockchain consensus mechanisms which is unrelated to federated learning tasks, leading to resource wastage and poor scalability, making them unsuitable for large-scale IoT networks. To address these challenges, we propose CoCFL, a novel BCFL framework utilizing multi-chain collaboration. CoCFL introduces two lightweight sub-chains: PoCFL-CChain and PC-CChain, based on different FL strategy. PoCFL-CChain uses a synchronous FL strategy for learning devices with similar performance to generate high-accuracy models, while PC-CChain adopts an asynchronous strategy for heterogeneous devices, which can improving training efficiency. CoCFL assigns devices to suitable sub-chains based on their performance to carry out FL tasks and aggregates the sub-chain models into a global model. This multi-chain collaboration strategy enhances model accuracy and convergence speed and significantly improves the scalability of BCFL. In additional, the consensus mechanisms in CoCFL sub-chains not only maintain the blockchain ledger but also handle FL-related tasks such as detecting poisoning attacks, assigning roles, and distributing incentives. This design not only improving the efficiency of BCFL, but also enhances learning security and ensuring fair incentives. Experiments show that CoCFL improves learning accuracy by 6% and efficiency by 18% over existing BCFL frameworks. It also demonstrates excellent scalability, with time consumption liner decreasing as sub-chains increase, and can withstand up to 40% of poisoning attacks while ensuring fair incentives.

Byzantine fault-tolerant (BFT) consensus protocols are the core components of blockchain. In the process of improving the performance of BFT protocols, existing work faces the following three problems: 1) the binary dilemma between the leader’s performance bottleneck in star-based linear communication and compromised resilience in tree-based sublinear communication; 2) two- or three-round protocols restrict the phase number of one proposal, thereby limiting the number of concurrent proposals and causing high latency. 3) The fixed timeout makes the protocol sensitive to varying network delays. Therefore, this paper proposesCrackle, the first sector-based pipelined BFT protocol with a sublinear communication complexity, for a throughput improvement of consensus protocol with max resilience of$(\mathcal {N}\textrm {-} 1)/3$. We propose a sector-based communication mode to disseminate messages from the leader to a subset of replicas in each phase to accelerate consensus and split the traditional two-round protocol into$2\mathsf {\kappa }$phases to increase the basic pipeline scale. We refine the timer strategy so that the timeout$\Delta $is adjusted with the proposal submission to cope with the changing network environment. We then address two technical challenges: 1) to ensure Quorum Certificate (QC) validation, we design a$\mathit {voteMap} s$field within each block, and verify QC by the signature aggregation of$\mathit {voteMap} s$in continuous$\mathsf {\kappa }$phases; and 2) to achieve pipeline decoupling among shorter phases, we propose a vote-appending mechanism that relaxes the conditions for the leader to send new proposals. We provide comprehensive theoretical proof of the correctness ofCrackle, including safety and$\mathit{liveness}$. Moreover, we implementCracklebased on a public BFT framework and deploy it on 64 cloud servers. Real experimental results reveal that ourCrackleprotocol achieves up to 10.36x higher throughput and can dynamically adapt to network delay compared with state-of-the-art BFT protocols such as Kauri and Hotstuff.

Video large language models (VideoLLMs) have significantly advanced video question answering (VideoQA) applications, which demand both low latency and high accuracy. To meet the requirements, VideoLLMs are typically deployed on GPUs for parallel acceleration. However, the massive computational load from long video contexts often makes such acceleration insufficient. Existing techniques like token pruning and speculative decoding attempt to address this challenge by altering the computational load, but often fail to balance both speed and accuracy. Sparse self-speculation mitigates these limitations via selecting a subset of tokens on a specific token budget to draft the output and then verify it using all tokens. However, existing sparse self-speculation is designed for text-based scenarios and cannot be directly applied to VideoQA tasks, as it fails to account for discrepancies in critical multimodal tokens and the dynamic nature of optimal token budget in VideoQA, leading to suboptimal scale and inappropriate composition of parallel computational load. We argue that achieving both high accuracy and low latency in VideoQA tasks requires managing the computational load with awareness of these discrepancies and dynamics. To achieve this, we introduce ViDA, a selective sparse self-speculation inference system. It progressively searches token budgets by iteratively refining lower and upper bounds of search space derived from long contexts, aiming to find and allocate varying optimal budgets in real-time adaptively. Additionally, it leverages insights from discrepancies in critical multimodal tokens to perform a discrepancy-aware token selection approach for identifying critical tokens. Evaluations across various VideoQA workloads show that compared to state-of-the-art methods, ViDA preserves exact model outputs while reducing average time-per-outputtoken (TPOT) by 15% to 46% and average end-to-end latency by up to 28%, while decreasing the divergence from optimal token budget distribution by up to 93

To enable parallel transaction processing in blockchain systems, recent consensus protocols have adopted directed acyclic graph (DAG) structures where DAG is used to organize and parallelize the blocks. Unfortunately, these protocols suffer from high communication overhead. Our experiment on the state-of-the-art Graded DAG[12] reveals that dissemination of transaction and consensus vote messages account for the majority of network traffic. We analyze that the overall overhead is O (N2) per replica and O (N3) for the entire system, where N is the number of replicas, and note that existing approaches have not succeeded in reducing this overhead.

Serverless workflow applications, composed of multiple serverless functions, are increasingly popular in production. However, inter-function communication and cold start latency remain key performance bottlenecks. This paper introduces AlloyStack, a library operating system (LibOS) tailored for serverless workflows. AlloyStack addresses two major challenges: (1) reducing cold start latency through on-demand OS component loading and (2) minimizing data transfer overhead by enabling functions within the same workflow to share a single address space, eliminating unnecessary data copying. To ensure secure isolation, AlloyStack uses Memory Protection Keys (MPK) to separate user functions from the LibOS while maintaining efficient data sharing. Our evaluation shows that AlloyStack reduces cold start times by 98.5% to just 1.3ms. Compared to SOTA systems, AlloyStack achieves a 7.3× to 38.7× speedup in Rust end-to-end latency and a 4.8× to 78.3× speedup in other languages for intermediate data-intensive workflows.

Large language models (LLMs) achieve state-of-the-art performance in many NLP tasks but incur prohibitive memory-access and compute costs when processing very long contexts due to linearly growing KV Cache. Existing static sparsification methods rely on fixed heuristics, while dynamic schemes incur substantial runtime overhead. To address this trade-off, we propose SmartCache, a sparse inference system that exploits two-dimensional KV Cache similarity across adjacent decoding iterations and neighboring layers. SmartCache combines a similarity-driven dual-path selection algorithm, which adaptively reuses TopK KV entries from both the previous iteration and the preceding layer with a rolling-array cache index manager that reduces index storage complexity from$O(L \cdot k)$to$O(k)$. We analyze the layer and iterative sparse patterns of KV Cache in long context LLM decoding and show that SmartCache maintains semantic consistency while drastically reducing redundant computation and memory traffic. Extensive experiments on Llama-3-8B-Instruct-Gradient-1048k, Qwen2.5-7B-Instruct-1M, and glm-4-9b-chat-1m across four long-context benchmarks report up to$\mathbf{3 0. 5 \%}$end-to-end latency reduction and$\mathbf{1 5} \boldsymbol{\%} \mathbf{- 2 3 \%}$average latency reduction, with inference accuracy degradation constrained within 2% and occasional slight improvements. These results indicate that SmartCache offers a practical, high-accuracy solution for scalable long-sequence LLM inference.

Modern LLM decoding has the drawbacks of high cost and slow speed, and speculative decoding has been shown to be an effective solution to this problem. However, the inference latency still poses a significant challenge to maintaining service level objectives (SLOs) in systems that employ multiple draft models for speculative decoding. The verification phase in such systems if reliant on tree attention often constitutes a bottleneck especially when draft sequences lack common prefixes and substantially underutilizes GPU parallelism while increasing end-to-end latency. We introduce SuperSpec, an end-to-end speculative decoding system designed to co-optimize verification, sampling and draft generation. SuperSpec integrates three pivotal innovations: an Efficient Batch Verifier, which substitutes treebased flattening with batch parallel validation and layer-wise KV Cache replication; a Global Optimal Sampler, which assesses all candidate sequences within a batch to ascertain the longest valid path, thereby circumventing the local optima frequently encountered in tree-based rejection sampling; and a Dynamic Adaptive Multi-Drafter, which dynamically modulates the speculative length (K) for each drafter predicated on real-time idleness metrics and acceptance rates. Empirical evaluations of Qwen2.5-72B and the OPT-66B on various datasets show that SuperSpec improves average acceptance rate by 6.4% to 30.2%, and the end-to-end inference acceleration ratio by 7.12% to 62.06%, when compared to the state-of-the-art tree-based speculative decoding system SpecInfer. These improvements were achieved without compromising the quality of text generation, making SuperSpec an effective solution for accelerating LLM inference.

Mixture-of-Experts (MoE) architectures have emerged as a scalable and efficient alternative to dense Transformer models by activating only a subset of experts per layer. However, deploying MoE models in multi-GPU environments faces severe challenges due to expert load imbalance and the resulting inefficient GPU utilization. Existing static replication strategies fail to adapt to dynamic token distributions, while dynamic rebalancing methods incur excessive communication and memory-access overheads, often outweighing the benefits of load balancing. This paper presents MoEoM, an inference system that enhances the efficiency of MoE models by innovatively taking memory-access costs into consideration. MoEoM integrates two complementary techniques: (i) a load-aware offline expert deployer, which symmetrically groups experts across GPUs and selectively replicates high-load experts, and (ii) an I/O-aware online token reallocator, which dynamically redistributes tokens among original and backup experts to minimize the maximum latency across GPUs. Experimental evaluations on state-of-theart MoE models demonstrate that MoEoM reduces end-toend inference latency by up to 16.6 %, improves throughput by 11–20 % in the prefill stage and 11–23 % in the decode stage, and decreases cross-GPU imbalance by 65–80 % (about 75 % on average), compared to prior MoE inference baselines. These results highlight the importance of incorporating both computation and memory-access overhead into expert placement and token scheduling for efficient large-scale MoE deployment.

The increasing parameter size of Deep Neural Networks (DNNs) has significantly enhanced model performance. As large-scale DNN models typically require partitioning into multiple blocks for distributed training, existing research has predominantly focused on offline scheduling for individual or batched training tasks. However, the stochastic arrival of such tasks in edge networks poses a critical challenge for efficiently scheduling them in resource-constrained edge clusters. In this paper, aiming to minimize the average training completion time across all tasks, we first extract DNN operator graphs and partition them into coarse-grained subgraphs using a max-flow mincut algorithm. Then, we formulate the online scheduling problem for multiple distributed DNN training tasks as a Markov Decision Process (MDP) and propose a reinforcement learning-based (RL-based) solution. Extensive experiments comparing our method with three conventional baselines (FIFO, SJF, and Greedy) under diverse configurations show that our approach reduces the average training completion time by 12.95%, demonstrating its effectiveness in resource-constrained edge environments with dynamic workloads.

End-to-end semantic communication paradigms demonstrate substantial potential in reducing network load and compressing data redundancy. However, their inherent openness introduces significant security risks, such as unauthorized access that enables attackers to camouflage themselves among legitimate users. Moreover, legitimate users may exploit input-output data pairs to conduct model stealing attacks. Existing defense strategies generally lack user access control mechanisms and fail to provide targeted countermeasures against model inversion attacks from internal users. To address this gap, we propose KGSC-SAT, a Key-Gated Semantic Communication framework enhanced by Steganography Adversarial Training for Secure Transmission. The framework employs a key-based feature modulation method to identify authorized users, while adversarial steganography training facilitates deep feature-level masking. Experimental results demonstrate that KGSC-SAT effectively mitigates both unauthorized access and insider model inversion threats, while delivering reliable communication performance.

Programmable packet scheduling enables users to customize scheduling algorithms flexibly without designing new ASICs. Existing schemes prefre to approximate optimal Push-In First-Out (PIFO) using First-In First-Out (FIFO) queues in commodity programmable switches. Despite its availability, these schemes suffer performance degradation due to the unawareness of available switch buffer. To be specific, when the port buffer is drained, existing schemes discard all incoming packets, even though these packets have higher priorities than the enqueued packets. In this paper, we reveal that the problem's key culprit is the lack of coordination between buffer management and packet scheduling in the switch. To fill this gap, we present Lark, a buffer-aware building block for programmable scheduling schemes designed to solve the above problem. Its key idea is to proactively drop the low-priority packets when the allocated buffer is to be drained, thereby admitting the later-arriving high-priority packets. Lark contains two modules, a lightweight gradient-based online prediction module and a simple priority-based decision module. Lark relies the former module to identify whether the allocated buffer is to be drained and uses the later to determine whether to drop the incoming packet. We have integrated Lark into two representative schemes, SP-PIFO and AIFO. Our large-scale evaluations over three realistic workloads show that Lark can significantly optimize their key metrics without sacrificing throughnut.

Advances in deep neural networks (DNNs) have significantly contributed to the development of real-time video processing applications. Efficient scheduling of DNN workloads in cloud-hosted inference systems is crucial to minimizing serving costs while meeting application latency constraints. However, existing systems suffer from excessive module latency during request dispatching, low execution throughput during module scheduling, and wasted latency budget during latency splitting for multi-DNN applications, which undermines their capability to minimize the serving cost. In this paper, we design a DNN inference system called Harpagon, which minimizes the serving cost under latency constraints with a three-level design. It first maximizes the batch collection rate with a batch-aware request dispatch policy to minimize the module latency. It then maximizes the module throughput with multi-tuple configurations and proper amount of dummy requests. It also carefully splits the end-to-end latency into per-module latency budget to minimize the total serving cost for multi-DNN applications. Evaluation shows that Harpagon outperforms the state of the art by 1.49 to 2.37 times in serving cost while satisfying the latency objectives. Additionally, compared to the optimal solution using brute force search, Harpagon derives the lower bound of serving cost for 91.5% workloads with millisecond level runtime.

Due to the exponential growth of blockchain ledger sizes, federated-storage which enables multiple devices to jointly store data, has emerged as a promising solution for secure data storage in edge computing. However, how to achieve verifiable queries in such decentralized storage remains underexplored. Existing broadcast-based query methods lack a verification mechanism for query results, making it impossible to ensure their correctness and completeness. Meanwhile, authenticated data structure based (ADS-based) query strategies are constrained by the full ledger data and cannot provide verifiable query services for users in a federated-storage environment. To this end, this paper takes the lead to propose EVQ, a verifiable blockchain keyword query scheme tailored for federated-storage edge computing. We propose a split keyword-based ADS as the core structure of our framework which ensures that users can verify the correctness and completeness of query results while alleviating storage pressure of edge devices. Specifically, the proposed ADS is constructed through a two-phase process: top-bottom keyword index tree construction and bottomtop RSA accumulator integration. Splitting the ADS based on keywords enables distributed data storage and the generation of corresponding ADS for the stored data. To reduce the query costs incurred by edge devices during query processing, we formulate the Keyword Allocation Optimization (KAO) problem and propose a gain-ratio-based keyword allocation mechanism to determine the splitting scheme of the ADS. The experiments are conducted based on the Foursquare dataset, which contains approximately 18 months of global check-in data collected from Foursquare. The experimental results show that, compared to the merkle tree strategy that also integrates the RSA accumulator, our EVQ improves query performance by 24.77 x.

Blockchain, with its immutability and decentralization, drives innovation in finance and supply chain, but the growing data volume makes storing complete ledger replicas impractical for users, especially in the resource-constrained Internet of Thing (IoT) scenarios. Existing solutions focus on nodes storing only a partial ledger to alleviate storage burdens. Nonetheless, these approaches prioritize storage optimization by minimizing the query cost and lack control over storage cost. Furthermore, these approaches overlook the relationships between network users, thus failing to fully measure the future query cost. Thus, this article proposes BSSN, a blockchain storage technology based on social networks. The combined use of storage cost and query cost is introduced for the first time to formulate the node allocation optimization (NAO) problem, and the multipopulation genetic ant colony (MGAC) algorithm will be employed to derive node allocation strategies. Specifically, we address three technical challenges: 1) to predict the transactions that nodes will participate in the future, we employ the social ties to obtain the access frequencies among users; 2) to strike a balance between the storage cost and query cost, we jointly model the two costs as a multiobjective optimization problem to formulate the NAO problem; and 3) to solve the NP-hard NAO problem, we use the MGAC algorithm, where the storage and query populations collaboratively search for solutions based on four operations. Extensive experiments indicate that compared with existing work, BSSN can reduce the average query cost to 67% with its adjustable storage cost, ensuring a balanced data storage among users.

Sign Language recognition is practically important in various scenarios such as smart home, medical rehabilitation, and intelligent industry. Compared with wireless sensing and computer vision methods, data glove-based methods have gained a plenty of attention, because they can perform well even in the environments with multi-path noise or visual occlusion. However, existing data glove-based methods usually require complex calibration and laborious dataset collection, and suffer from accumulated error. To address these challenges, we introduce a robust sign language recognition system with cross-domain generation, called SmartGlove, the first approach to achieve robust sign language recognition. To avoid complex calibration process, we propose a customized feature set that can enable user-insensitive and unintentional system calibration. To avoid the labor cost in training data collection, we propose a cross-domain data transformation technique to generate training data in target domain. To eliminate the accumulated error of sentence recognition, we utilize a context-based calibration method considering correlation among adjacent words. We implement SmartGlove with COTS devices, and extensive experiments reveal that SmartGlove achieves accuracy exceeding 97.11% for 30 sign language words, with an average recognition time of 47 milliseconds per word. Furthermore, the system recognizes 30 common sign language sentences with accuracy of 97.17%.

Automated manufacturing is the cornerstone of the Industrial Internet of Things (IIoT) ecosystem, where vibration monitoring technology is a critical tool for maintaining industrial machinery. The prevailing approach mostly employs inertial measurement units (IMUs), lasers, and cameras, each demonstrating deployment constraints. In recent years, millimeter-wave (mmWave) radar has shown high vibration measurement performance, but it faces challenges in accurately localizing vibrating objects and determining observation points. This study introduces a new system called VibCamera, which leverages the mmWave vibration measurement technology with computer vision (CV) algorithms for vibration monitoring. With the positional assistant of CV semantic segmentation, the radar can accurately determine sufficient observation points, thereby achieving precise measurement with high directionality. VibCamera includes two camera modes, RGB-only and RGB+depth, and solves two technical challenges: 1) integrating multimodal information for vibration target localization and 2) extracting high-quality vibration signals in interference environments. VibCamera provides more consistent and precise outcomes without the need for physical contact. The experimental results indicate that the RGB-only mode has amplitude and frequency errors below$27.04 \; \mu \rm m$and 0.22 Hz, respectively, with a 90% probability, and the RGB+depth mode has errors below$23.72 \; \mu \rm m$and 0.21 Hz.

Sharding has become a significant approach to enhance blockchain scalability. However, existing sharding techniques applied in IoT scenarios suffer from transaction congestion due to imbalanced distribution of transactions across shards, which hinders intra-shard transaction processing capacity. To overcome the above problems, this paper proposes HydraChain for IoT scenarios, the first multi-agent reinforcement learning based sharding blockchain system with account graph relationships, for a throughput improvement of shards under realtime load balancing. Agents collaborate by sharing information and jointly optimizing decisions, enhancing the accuracy and efficiency of the decision-making process. We first construct a sharding blockchain environment integrated with an embedded graph encoder. Concurrently, we propose a SG-MAPPO multiagent model with decoder, which enables agents to cooperatively learn to optimize account allocation strategies based on real-time shard load and global system information. When implementing HydraChain, we address two technical challenges: (i) to extract granular behavioral features from accounts with diverse and time-varying patterns, we design a graph data encoder, which constructs a graph network based on transactional relationship; and (ii) to ensure real-time load balancing under the constraints of dynamic transaction patterns, we propose a multi-agent model (SG-MAPPO), which matches graph encoding features within the environment. Our approach leverages the ability of multi-agent model to collaborate and adapt to the changing environment, enabling efficient resource allocation and improved system performance. Moreover, we implement HydraChain and conduct experiments on a high-performance server equipped with 48 cores and 125GB of memory. Our comprehensive experiments, comparing HydraChain with DQN-Based, SAC-Based and SPRING, reveal that our solution outperforms state-of-theart solutions by achieving a notable 22% increase in transaction throughput and a 5.2% reduction in workload imbalance across shards.

Mobile crowd sensing (MCS), as an innovative data acquisition model in the Internet of Things (IoT), employs an incentive mechanism based on users’ reputation evaluation, which is a mainstream reward allocation method. However, in the existing incentive mechanisms based on reputation evaluation, unidirectional incentive strategies and nonadaptive reputation models result in unequal reward allocation. To tackle this issue, we propose an effective interactive incentive mechanism based on adaptive reputation evaluation. Specifically, we generate user status thresholds to classify, rate, and weight user behaviors, based on the average quality thresholds of tasks released or data submitted by different users in each interaction round. Meanwhile, we achieve multiparty consensus by incorporating the obtained user reputation values and combining them with the cumulative reputation values from multiple rounds to obtain adaptive reputation evaluation results. Moreover, we design an interactive incentive strategy that measures users’ incentive values based on their reputation evaluation results in each round, mutually punishing malicious behaviors from both the publisher’s and the worker’s perspectives. Extensive experiments have demonstrated that our method consistently outperforms existing advanced incentive mechanisms.

The application scenarios of drone collaboration are rapidly expanding, such as the low-altitude economy and wildfire protection. Blockchain-based drone collaboration requires a consensus mechanism to ensure efficient and secure consistency among large-scale distributed nodes. However, the existing consensus mechanism has problems with poor fault tolerance of topology and rigid proposal concurrency. To this end, this paper proposes AirBFT, an efficient and robust consensus mechanism for large-scale drone collaboration. First, this paper designs a new four-layer network topology, using upper-member and lower-member communication, while ensuring the maximum 1/3 resilience and fanout of √N. Secondly, this paper proposes a dynamic pipelining algorithm to adjust the parallelism of proposals according to the real-time network status. Finally, this paper proposes a committee sampling technology based on the EigenTrust algorithm to reduce the impact of the malicious behavior of Byzantine nodes. Experiments based on the public consensus framework show that compared with Kauri and HotStuff, the proposed AirBFT reduces transaction confirmation delay by 58%, the throughput is increased by 1.9 times, and it can ensure efficient operation with a 1/3 Byzantine node ratio.

The Wi-Fi sensing technique holds great promise for future smart homes, thanks to the widespread use of Wi-Fi devices. With this technique, we can deduce the behavior of the target based on the channel state information (CSI), which is obtained during Wi-Fi communication. However, existing Wi-Fi sensing technologies are not compatible with standard communication technologies. This is because Wi-Fi sensing usually relies on capturing CSI from high-frequency communication packets, whereas regular IoT communication does not consistently maintain such high communication rates. To achieve precise sensing even with a low packet rate, we introduce LowDetrack, an indoor human detection and tracking system at ultra-low packet rates with Wi-Fi. In particular, we utilize compressed sensing to supplement missing data compared to existing systems that rely on linear interpolation or neural networks. To detect and track the target, our insights are twofold: 1) We combine compressed sensing and Fresnel zone to a theoretical model for accurately obtaining the reflection path change rate, which can be converted into the actual velocity of the target; 2) We investigate the mapping relationship between the dynamic frequency composition ratios in different links, which can provide navigation for velocity direction and correct direction recognition errors. We implement LowDetrack on commercial off-the-shelf Wi-Fi and realize human detection and tracking, where the median tracking error is 0.76m at the packet rate of 25 Hz.

A federated learning framework enables multiple clients to jointly train models locally without uploading their private data, effectively protecting the clients’ data privacy. However, existing federated learning auction mechanisms have not considered heterogeneity in client training time, making it difficult for the server to aggregate client models effectively within a constrained time. Moreover, continuously selecting specific clients in each round can lead to overfitting. This article proposes an Uncertainty-aware Auction Mechanism (UAMARD) based on Age of Update (AoU), Reputation, and Data Quantity, which considers training time and provides guidance on the number of data points to participate in training for selected clients. Firstly, we model a reverse auction system that considers the uncertainty of training time to promote client participation. We introduce AoU to quantify the time interval required for the server to receive the latest updates from the client to avoid overfitting. Then, we prove that solving the problem of maximizing social welfare is NP-hard. Subsequently, we introduce a dynamic programming algorithm (VCG RA) to solve the problem of maximizing social welfare. To further reduce time complexity, we propose our UAMARD method, which achieves a near-optimal level of social welfare while ensuring minimal time complexity. Ultimately, simulation experiments confirmed the efficacy of UAMARD and VCG RA. When benchmarked against other mechanisms, UAMARD and VCG RA demonstrated superior performance with quicker convergence and higher accuracy in testing the MNIST and CIFAR-10 datasets.

Abstract Federated learning represents a decentralized approach to machine learning, enabling numerous devices to collaboratively contribute to model training while ensuring the privacy of individual data. However, the existing incentive mechanism of hierarchical federated learning (HFL) only considers the data contribution of a single round, which needs to be revised. For non‐IID data sets, the continuous selection of any end devices will cause the weights to diverge in a specific direction. Therefore, a new metric is needed to avoid continuously selecting a certain end device to ensure the overall effectiveness. We introduce a metric to describe the importance of updates: age of update (AoU), which can help select end devices not selected in the previous round to promote a faster model convergence. We put forward an incentive mechanism based on AoU, reputation, and data quantity in HFL (ARDHFL). We have derived the optimal equilibrium solution for the three‐stage Stackelberg game. Based on this solution, we can ensure maximum edge‐cloud utility while incentivizing end devices to engage actively in HFL tasks and providing superior data to train the HFL model. Finally, we conducted extensive experiments to prove that ARDHFL can effectively improve the performance. Compared with the fixed scheme, random scheme, FMore and InFEDge, the testing accuracy of ARDHFL in the MNIST dataset has been improved by 29.7%, 9.3%, 6.8% and 6.1%, respectively. In the CIFAR‐10 dataset, it has been improved by 40.2%, 33.1%, 16.4% and 14.2%, respectively, and demands fewer communication iterations to achieve the same testing accuracy.

Large language models (LLMs) have demonstrated remarkable performance across a wide range of tasks, largely due to their substantial model size. However, this also results in significant GPU memory demands during inference. To address these challenges on hardware with limited GPU memory, existing approaches employ offloading techniques that offload unused tensors to CPU memory, thereby reducing GPU memory usage. Since offloading involves data transfer between GPU and CPU, it introduces transfer overhead. To mitigate this, prior works typically overlap data transfer with GPU computation using a fixed pipelining strategy applied uniformly across all inference iterations, referred to asstaticoffloading. However, static offloading policies fail to maximize inference throughput because they cannot adapt to the dynamically changing transfer overhead during the inference process, leading to increasing GPU idleness and reduced inference throughput.We propose that offloading policies should beadaptiveto the varying transfer overhead across inference iterations to maximize inference throughput. To this end, we design and implement an adaptive offloading-based inference system called TightLLM with two key innovations. First, its key-value (KV) distributor employs atrade-compute-for-transferstrategy to address growing transfer overhead by dynamically recomputing portions of the KV cache, effectively overlapping data transfer with computation and minimizing GPU idleness. Second, TightLLM’s weight loader slices model weights and distributes the loading processacross multiple batches, amortizing the excessive weight loading overhead and significantly improving throughput. Evaluation across various combinations of GPU hardware and LLM models shows that TightLLM achieves 1.3 to 23 times higher throughput during the decoding phase and 1.2 to 22 times higher throughput in the prefill phase compared to state-of-the-art offloading systems. Due to the higher throughput in prefill and decoding phases, TightLLM can reduce the completion time for large-scale tasks, which involve processing and generating a substantial number of tokens, by 59.6% to 94.9%.

In-Network Aggregation (INA) solutions represent the forefront in advancing All-Reduce, utilizing limited switch memory for efficient gradient aggregation. However, existing INA solutions primarily focus on enhancing aggregation efficiency, often overlooking the efficient utilization of memory. Isolation solutions typically pre-allocate resources for each job, leading to memory wastage due to the uncontrolled use of resources. In contrast, the sharing solutions encounter significant memory contention, resulting in performance degradation within a multi-tenant environment. In this paper, we propose DynaINA, a flexible job scheduler to support multi-tenant training. The core idea of DynaINA is to provide spatial and temporal compatibility between jobs. For spatial compatibility, DynaINA utilizes multiple dynamic memory pools to provide job isolation. For temporal compatibility, DynaINA employs contention-aware job scheduling to facilitate memory sharing. Furthermore, DynaINA prioritizes communication-intensive jobs, leveraging the benefits of INA to enhance overall performance in training clusters. Extensive experiments with popular vision and language models demonstrate that DynaINA reduces training time by up to 65.16% and improves switch memory utilization by up to 85.02% compared to state-of-the-art solutions in a 100Gbps network.

Blockchain offers distinct advantages in terms of data credibility and provenance certification, and its fusion with Internet of Things (IoT) technology holds great promise. Nevertheless, IoT environments are marked by extensive node networks and intricate communication patterns, especially the sensing environment. The conventional blockchain consensus mechanism, hampered by its heavy reliance on computing resources and communication bandwidth, faces difficulties in ensuring seamless data exchange among IoT edge servers. The issues encountered by state-of-the-art Byzantine Fault Tolerance (BFT) consensus include: (i) high communication complexity between nodes; and (ii) the detrimental impact of Byzantine behavior on system performance. To overcome the above problems, we propose the lightweight blockchain consensus called AntB, firstly introducing the concept of sampling into the consensus and significantly reducing the number of participating consensus nodes from$N$to$n$, which lowers the consensus complexity to$\mathbf{2\cdot O(n)+O(N)}$. We design a dynamic reputation mechanism so that Byzantine nodes cannot control the sampling set to affect the activity of the consensus in the long term. When implementing AntB, we address three significant technical challenges: (i) to determine the optimal sample size, we propose a sampling calculation method based on statistical confidence intervals, where the sample size is primarily determined by the chosen confidence level and margin of error; (ii) to prevent Byzantine behavior, we devise a weighted random sampling mechanism utilizing reputation coefficients based on edge servers’ behaviors; and (iii) to maintain consensus activity and consistency after sampling, we propose the consensus mechanism for partial sampling and global verification to avert potential issues. We implement AntB and conduct performance evaluations in a server with 32 cores and 64GB of memory. The evaluation results indicate that, the more nodes participating in the process of consensus, the better the performance of AntB will be. Especially, compared to HotStuff, AntB has a 24.94% higher success rate and Transactions Per Second (TPS) can improve by 102.10% when the number of nodes is 300.

In recent years, sharding technology has been frequently applied in blockchain systems to increase scalability. However, when new shards are added, the system may result in significant overhead in terms of computing and networking since the data allocation approach is incompatible with dynamic changes in shards. Currently, S-Store, the state-of-the-art sharding solution built on the account model, has a high re-computing latency when growing shard numbers and an unbalanced sharded data distribution after growth. To address these issues, this paper presents Tangram, an efficient and balanced dynamic storage extension approach for sharding blockchain systems. Tangram reduces system extension overhead and latency while ensuring a balanced shard distribution. In implementing Tangram, we tackle three main technical challenges as follows. (1) Designing a novel state tree structure for the storage and maintenance of sharding state data. We introduce the Jump Merkle Tree (JMT) based on the Merkle Tree, which integrates node migration and orderliness. (2) Presenting a protocol to be compatible with dynamic shard scenarios. We devise a shard addition protocol to improve system extension availability and decrease shard extension delay. (3) Proposing an approach to guarantee system longevity after extension. We first devise algorithms for the state tree to eradicate invalid states after system expansion. Furthermore, we introduce a shard reduction protocol to enhance system storage extension support in complex scenarios, such as cleaning up inactive states to avoid bloating the state tree. We conduct extensive experiments to evaluate the performance of Tangram. Experiment results demonstrate that Tangram outperforms existing solutions, showing reduced latency and superior data balance. When compared to the state-of-the-art sharding storage solution, Tangram decreases the transaction execute time by up to 87.84%, the state data migration by more than approximately 74%, and achieves up to 7.63x improvement in the standard deviation of sharding data balance.

Given privacy and network load concerns, employing on-device multimodal neural networks (MNNs) for IoT data is a growing trend. However, the high computational demands of MNNs clash with limited on-device resources. MNNs involve input and model redundancies during inference, wasting resources to process redundant input components and run excess model parameters. Model Redundancy Elimination (MRE) reduces redundant parameters but cannot bypass inference for unnecessary input components. Input Redundancy Elimination (IRE) skips inference for redundant input components but cannot reduce computation for the remaining parts. MRE and IRE independently fail to meet the diverse computational needs of multimodal inference. To address these issues, we aim to combine the advantages of MRE and IRE to achieve a more efficient inference. We propose anadaptivemultilevelredundancyelimination framework (AMRE), which supports both IRE and MRE.AMREfirst establishes a collaborative inference mechanism for IRE and MRE. We then propose a multifunctional, lightweight policy model that adaptively controls the inference logic for each instance. Moreover, a three-stage training method is proposed to ensure the performance of collaborative inference inAMRE. We validateAMREin three scenarios, achieving up to 52.91% lower latency, 56.79% lower energy cost, and a slight accuracy gain compared to state-of-the-art baselines.

Non-intrusive hand tracking with mmWave radar technology is important in various Human-Computer Interaction (HCI) scenarios. However, existing mmWave-based solutions require users to be stationary and restrict a fixed hand motion area, which limits application flexibility and user experience. This paper proposes a novel mmWave-basedMobileHandTracking (MHTrack) system, which tracks user's hand gestures during walking. MHTrack focuses on tracking bothabsolutehand trajectory in the global coordinate system andrelativehand trajectory to the body. Specifically, we propose a wake-up mechanism for hand motion capture, in which hand point cloud can be recognized even under body interference and noise. We propose a hand tracking strategy named local spatial update, which overcomes the sparsity and instability of point clouds, to obtain absolute hand trajectory. Subsequently, we propose a hand anchor correction method to suppress anchor offset and remove the impact of body movement from absolute hand trajectory, thereby obtaining relative hand trajectory. As a case study, we project the relative hand trajectory onto a 2D image and feed it into a gesture recognition model to recognize the gestures. We conduct extensive experiments to evaluate the performance of MHTrack. Results demonstrate a 3D hand trajectory tracking error of$3.6cm$in an area of$3.2m\times 4.8m$and a gesture recognition accuracy of$99\%$with 30 gesture classes.

Liquid sensing in ubiquitous contexts plays an essential role in various scenarios. Recently, some wireless sensing systems have been proposed for liquid identification. However, existing works usually require specific equipment or capture the signals penetrating a target, limiting the deployability of liquid sensing. In large-scale scenarios, multiple devices are usually required to expand the coverage area due to the RFID reader antenna's reading range limitation. To enlarge the sensing range and make the liquid sensing method can be adopted in real moving scenarios, in this paper, we presentMobileLiquidIDentification (MLiquID), a liquid sensing system that can recognize the type of liquid in a mobile manner with commercial off-the-shelf (COTS) RFID devices. This mobile process leads to continuous variation in location, so the major challenge in this paper is how to extract signal features from the superimposed information of movement and material. The key insight is to regard movement as an opportunity to acquire data from different perspectives instead of a challenge to hinder feature extraction. We construct a Phase-RSS model by analyzing the influence of moving and liquid on the phase and RSS signals. First, we propose a method to calculate the distance from the tag to the reader antenna. Second, we explore an identification method to identify liquid type by extracting signal features Phase-RSS coefficient$C_{P-R}$and Maximum Response Distance (MRD). Experimental results demonstrate an average accuracy of 96.80% in identifying 10 common liquids, which shows the great potential of MLiquID for mobile liquid sensing.

As an advanced technology for non-contact sensing, mmWave radar enables fine-grained measurement of a wide variety of user behaviors. While creating intelligence and convenience, it also concerns behavioral privacy and security, as radar signals contain a wealth of behavioral information. Existing solutions are either incapable of customizable privacy protections or cannot cope with multi-person scenarios. This paper presents aMulti-user behavioral privacyFilter, MuFilter, a data masking system centered on the idea of dimensional signal interference. It determines the sensing signatures that need to be preserved or interfered with based on the sensing services that users want to enable and disable, thereby making targeted tampering on the radar signal. On this basis, we introduce the multi-person tracking technology to allow MuFilter to determine the number of users in unknown scenarios. Moreover, a subspace tampering technique is proposed to ensure that each tampering only affects the target user and not other users, thus supporting personalized privacy protection for multiple users. Experiments show that MuFilter can interfere with targeted behavioral signatures with a 100% success rate, while the degree of impact on other users’ signatures ranges from 0% to 3.85%.

Joint service caching and workload scheduling plays an important role in ultra-dense mobile edge computing (MEC) networks to satisfy the stringent requirements of latency-critical services by leveraging the aggregated edge resources (e.g., storage and computing resources) located near the users. However, most of the existing methods incorporating popularity-based and/or size-aware caching strategies fail to match the resource demands of user requests with the heterogeneous resources of edge nodes, leading to heavier cloud load and higher latency. It becomes even worse when user requests exhibit dynamic variations over time. To address these issues, we propose an online approach for resource-efficient joint service caching and workload scheduling in ultra-dense MEC networks, called CoShare. The core idea is to fully utilize the heterogeneous resources of the edge layer to further reduce the cloud load and thus the service latency. First, we formulate the joint service caching and workload scheduling problem as a mixed integer nonlinear programming problem with the goal of minimizing the cloud load. Then, an online algorithm is developed to transform this optimization problem into a series of per-slot sub-problems by leveraging Lyapunov optimization. Next, to solve these sub-problems, we design a cacheability-based alternating iterative algorithm utilizing Gibbs sampling, in which the cacheability indicator considers both service resource demands and service popularity. Finally, simulation results show that CoShare can effectively exploit available edge resources to achieve lower cloud loads compared to other strategies.

Federated Learning (FL) is a new distributed machine learning that allows end devices to collaboratively train a global model using local data to protect data privacy. However, the existing FL incentive mechanisms only consider a single round’s data contribution, which needs improvement. For non-IID datasets, continuous selection of any end device will cause weights to diverge in a specific direction. Therefore, a new metric is needed to avoid continuously selecting a particular end device to ensure the overall performance of the incentive mechanism. We have introduced a metric to describe the importance of updates: Age of Update (AoU), which can help select end devices not selected in the previous round to promote faster model convergence. We proposed an incentive mechanism for a Two-stage Stackelberg Game with Multidimensional Individual Attributes (TSGMIA). We have derived the optimal equilibrium solution for the game in two stages. Based on this solution, we can ensure maximum server utility while incentivizing clients to participate actively in FL tasks and providing high-quality data to train the FL model. Finally, we conducted extensive simulation experiments to demonstrate that our proposed mechanism can effectively improve the performance of the FL model. Specifically, when the model accuracy reaches 0.9, compared to InFEDge, the communication round of TSGMIA will be reduced by 9%. In addition, compared with the fixed selection scheme, random selection scheme, and InFEDge, after 100 rounds of training, TSGMIA improved the accuracy of the global model by 30%, 10%, and 1.7%, respectively.

Federated learning (FL) is a distributed learning paradigm that enables multiple client devices to collaboratively train a global model based on their local datasets while protecting data privacy. However, due to its distributed nature, FL is susceptible to the resources of heterogeneous client devices with different data quantities, communication resources, and computing capabilities. Heterogeneity leads to uncertain global model training time and hinders the convergence of the global model. Therefore, selecting suitable clients to participate in the FL training process is necessary to improve the efficiency of FL. This paper proposes an FL client selection method (FedAHP) based on the Analytic Hierarchy Process (AHP) to optimally balance the trade-off between model accuracy and training time during client selection. Experiments show that FedAHP outperforms the greedy method regarding training time consumption and model accuracy. Specifically, FedAHP achieves a 67% reduction in communication rounds compared to the greedy method when the model accuracy reaches 0.90. Furthermore, when taking the same 20 hours, FedAHP improves the accuracy of the global model by 6% in comparison to the greedy method.

RDMA has been widely deployed to improve the performance of applications with frequently fine-grained remote access. However, restricted on-chip resources result in cache misses under high concurrency that significantly degrade network performance. QPC-aware solutions only focus on the number of concurrent QPs, ignoring the impact of WQE within QPs. SMART limits the number of WQEs in each QP with a credit-based scheme. Nevertheless, we find that equal treatment increases the tail latency of messages ranging from 32 bytes to 1024 bytes by 2× when mixing messages of different sizes. In this paper, we introduce WQEFC, a scalable RDMA message scheduler that provides lower latency and higher throughput for applications with heavily concurrent messages. Our key insight is that there is a significant difference in the sensitivity to cache miss between messages of different sizes. For messages smaller than 32 bytes, which are sensitive to cache misses, we combine the credit limiter and sub-message poller, limiting the number of concurrent wqes to avoid cache miss while ensuring optimal message completion latency. For other messages, which are insensitive to cache miss, we assign them a higher priority and use the sub-message poller to ensure message concurrency while reducing cache miss. We implement WQEFC as a middleware between the driver layer and the application layer for flexible deployment. WQEFC outperforms the state-of-the-art solution Smart by increasing system throughput by 61.3%, and reducing the tail latency of messages smaller than 32 bytes and larger than 32 bytes by 41.6% and 72.6%, respectively.

Asynchronous Complete Secret Sharing (ACSS) in Byzantine fault-tolerant systems has become one of the essential building blocks in multiple threshold cryptosystems. However, current ACSS schemes scale poorly due to high communication costs, which are quadratic in the number of participants n. In this paper, we propose a new scheme ALCES to reduce such communication costs from O(n2) to O(cn) with a negligible probability of failure ${e^{ - \frac{c}{{18}}}}$, while guaranteeing completeness and agreement properties. The key point of ALCES is to sample c parties to construct a committee, which then verifies and distributes the encrypted shares to other parties. Additionally, we introduce a new mechanism, referred to as secret labels in ALCES, by encoding the information of labels in polynomial coefficients. This mechanism allows an arbitrary string to act as the label, binding it to a specific secret while efficiently ensuring security and privacy with minimal communication cost. Experimental results show that our technique reduces the overall communication cost in a single sharing process by 66% and 83% for very large quantities, such as 4096 and 8192 parties, respectively, when compared with prior work.

Phishing scams represent a significant criminal activity on Ethereum, driving the need for effective detection methods. The methods based on graph neural networks(GNNs) make significant breakthroughs due to their ability to model complex transaction networks. However, existing approaches often overlook the heterogeneity of Ethereum’s transaction graph during neighbor nodes aggregation. These methods typically focus on low-order neighbors, disregarding high-order ones, which limits their overall performance. To this end, we propose the High- and Low-order Transaction Aggregation Graph Network(HLTAG), which separately aggregates high- and low-order features for more effective feature representation. Specifically, we utilize biased random walk to aggregate low-order neighbors. We employ path aggregation to handle high-order neighbors. To mitigate the influence of noise and redundant information from high-order neighbors, we introduce a combination of attention decay, node similarity, and path attention mechanism, which dynamically adjust the aggregation weights. Extensive experiments demonstrate that HLTAG (94.4% Recall and 89.3% AUC) outperforms the state-of-the-art approaches in detecting Ethereum phishing scams, and exhibits significant advantages in large-scale scenarios.

With the rapid expansion of blockchain data, the demand for data exchange between chains has grown significantly. Authenticated queries have become one of the crucial methods for retrieving on-chain data due to their efficient performance and ability to ensure data security. However, existing intra-chain query approaches either face substantial maintenance overhead or exhibit low query efficiency, when dealing with the explosive growth of data in cross-chain scenarios; while current cross-chain query approaches suffer from issues including limited query types and poor scalability. To this end, this paper takes the lead to propose a novel generalized framework named CubeChain which provides various query types for intra- and cross-chain authenticated queries. We propose a highly scalable authenticated data structure (ADS) named Cube as the core structure of our framework which excels in achieving high performance while minimizing maintenance overhead by establishing data bridges between vertexes. When implementing CubeChain, we address two challenges: (i) implementing lightweight verification while supporting various query types by using a two-layer hashing structure, and (ii) further improving the query efficiency by suppressing vertexes. We substantiate the superior performance of Cube in terms of query efficiency, update overhead, and scalability through theoretical analysis. Finally, we implement the CubeChain framework based on the open-source Fabric v2.2. Real experiments with YCSB benchmark demonstrate that, compared with the state-of-the-art Bs+tree-based ADSs in MSTDB and SEBDB, our query performance improved by 23.75x in intra-chain scenarios and 10.72x in cross-chain scenarios, while maintaining a 30% reduction of cross-chain query load.

One notable drawback of traditional Federated Learning (FL) is its susceptibility to single point of failures. In recent years, Blockchain-based Federated Learning (BCFL) has been proposed as an effective solution to address this issue. However, existing BCFL frameworks face challenges in heterogeneous IoT scenarios. The heterogeneity of IoT devices poses challenges to the adaptation of blockchain consensus. The integration of blockchain imposes constraints on the learning scalability of systems, making it challenging to accommodate a large number of heterogeneous IoT devices. On the other hand, current blockchain consensus fail to sufficiently measure the contributions and destructions among heterogeneous devices in terms of learning quality, leading to low learning security and insufficient incentive fairness. To overcome the limitations of prior art, this paper introduces CoCFL, a novel blockchain-based federated learning framework based on multi-chain collaborative model. CoCFL enhances learning scalability by adopting a multi-chain asynchronous collaboration approach that partitions both learning and communication granularity of the system. Within each sub chain, CoCFL introduces a lightweight, secure and incentive-fair blockchain-based federated learning consensus, called Proof of Contribution to FL (PoCFL). In PoCFL, partic-ipants' contributions to the learning and the consensus process form the basis for delegating consensus responsibility and dis-tributing rewards. Furthermore, we introduce a novel malicious model detection algorithm into PoCFL, called the Trustee Nearest Algorithm. Through Trustee Nearest, PoCFL effectively mitigates poisoning attacks. Experimental results demonstrate that CoCFL exhibits better learning scalability compared to traditional FL and and avdanced BCFL frameworks in the same scenarios and can effectively withstand poisoning attacks initiated by at least 40% of malicious participants. Moreover, CoCFL demonstrated good incentive fairness during the learning process.

An innovative architecture called execute-order-validate (EOV) has been proposed by Hyperledger Fabric that enables concurrent processing of transactions. However, the architecture suffers from issues such as excessive invalid transactions and serialization limitations in scenarios with high transaction conflicts, which restrict its applicability in real-time and high-performance settings. To address the aforementioned limitations, we propose ParFabric to enhance the EOV architecture. Firstly, we analyze four essential characteristics required for the transaction reordering algorithm within this architecture. We propose a heuristic dynamic reordering algorithm to reduce the number of invalid transactions. This is achieved through real-time identification and early abortion of transactions based on weighted pre-ordering and the construction of a transaction conflict graph. Secondly, leveraging the transaction conflict graph, we introduce a novel optimal block packing strategy based on transaction dependencies. This strategy replaces the total transaction order with partial order, enabling parallel validation and commit at the block level, thereby leading to increased system throughput while reducing transaction latency. Experimental results indicate that, ParFabric demonstrates excellent performance in terms of vertical scaling of peers. Additionally, at the same infrastructure cost, ParFabric provides 2.2x and 1.6x higher throughput than FabricPlusPlus and FabricSharp in high-conflict scenarios.

Serverless computing enables a new way of building and scaling cloud applications by allowing developers to write fine-grained functions. The execution duration of a cloud function is typically short, usually ranging from a few milliseconds to a few seconds. FaaS providers charge users based on the execution duration of cloud functions with a granularity of 1 millisecond. Existing mixed scheduling methods collocate functions with varying execution times, which may prolong their execution duration and lead to unfair charges for FaaS users. To address this problem, we propose a partition scheduling approach, placing functions with varying execution times on different partitions. We introduce Synergy, a solution in serverless computing that leverages a collaboration of central and local scheduling to partition functions and employs suitable scheduling algorithms for these partitions respectively. Synergy also supports the dynamic switching of scheduling algorithms for partitioned nodes to adapt to highly fluctuating loads. We evaluate Synergy using real-world, representative benchmarks. Experimental results demonstrate that, compared to state-of-the-art and conventional approaches, Synergy can reduce the average function execution duration by $63 \%$.

Large-scale datacenter networks are increasingly using in-network aggregation (INA) and remote direct memory access (RDMA) techniques to accelerate deep neural network (DNN) training. However, existing research trends suggest that these two techniques are on an inevitable collision course. To fill this gap, we present FreeINA, a host-driven in-network aggregation aimed at providing RDMA reliable connection (RC) for multi-tenant learning settings. FreeINA relies on dual transmission paths to support RC compatibility, with one path for INA and another one for aggregation on end-host parameter server. With dynamic control of these two paths, FreeINA can leave the traditional in-server aggregation unaffected while ensuring INA’s reliability without modifying RDMA network interfaces (RNICs). To support multi-tenant learning, FreeINA employs all-reduce-level memory allocation, which can capture the well-known "on and off" DNN training pattern and thus improve switch memory efficiency. We have implemented a FreeINA prototype using P4-programmable switch and commercial RNICs, and evaluated it extensively using 100Gbps testbed. The results show that compared to the state-of-the-art solution—ATP, FreeINA improves single-job training speedup ratio by 1.20×, while improving the aggregation throughput by 2.65× in multi-job scenario.

Blockchain systems widely employ Byzantine fault-tolerant (BFT) protocols to ensure consistency. Improving BFT protocols’ throughput is crucial for large-scale blockchain systems. Frontier protocols face crucial problems: (i) the binary dilemma between leader bottleneck in star-based linear communication and compromised resilience in tree-based sublinear communication; and (ii) 2- or 3-round protocols restrict the phase number of one proposal, thereby limiting the scalability and parallelism of the pipeline. To overcome the above problems, this paper proposes Crackle, the first sector-based pipelined BFT protocol with a sublinear communication complexity, for a throughput improvement of consensus protocol with max resilience of (N-1)/3. We propose a sector-based communication mode to disseminate messages from the leader to a subset of replicas in each phase to accelerate consensus and split the traditional two-round protocol into 2κ phases to increase the basic pipeline scale. When implementing Crackle, we address two technical challenges: (i) to ensure Quorum Certificate (QC) validation during continuous κ phases, we design a voteMap field within each block, and verify QC by the aggregation of continuous κ voteMaps; and (ii) to achieve pipeline decoupling among shorter phases, we propose a vote-appending mechanism that accelerates the leader’s transition to the next phase. We provide comprehensive theoretical proof of the correctness of Crackle, including safety and liveness. Moreover, we implement Crackle based on a public BFT framework and deploy it on 64 cloud servers. Real experimental results reveal that Crackle achieves up to 10.36x higher throughput compared with state-of-the-art BFT protocols such as Kauri and Hotstuff.

In the realm of industrial Internet of Things(IIoT), the data pattern is subject to change over time, necessitating the implementation of online anomaly detection to adapt to the change of data pattern. Given the multiple stages in the production process in IIoT, data at different times exhibit varying periodic characteristics. Existing training methods primarily use fixed time windows, which struggle to adapt to complex time patterns, leading to decreased accuracy in anomaly detection. Furthermore, the incremental update method which utilizes a single model cannot effectively capture changing data characteristics. This paper introduces an online anomaly detection architecture named Adaptive Time Window enabled Model Pool (ATWMP). The framework utilizes a reinforcement learning model to dynamically select the optimal time window for model update and anomaly detection. Within the model pool framework, anomaly detection is conducted based on model reliability, and model updates are performed according to concept drift in order to ensure accurate adaptation to changing data features. Comprehensive experiments conducted on eight concept-drifted datasets and IIoT datasets demonstrate the superiority of this proposed method compared with other advanced methods.

The latest research on wildfire forecast and backtracking has adopted AI models, which require a large amount of data from wildfire scenarios to capture fire spread patterns. This paper explores using cost-effective simulated wildfire scenarios to train AI models and apply them to the analysis of real-world wildfire. This solution requires AI models to minimize the Sim2Real gap, a brand-new topic in the fire spread analysis research community. To investigate the possibility of minimizing the Sim2Real gap, we collect the Sim2Real-Fire dataset that contains 1M simulated scenarios with multi-modal environmental information for training AI models. We prepare 1K real-world wildfire scenarios for testing the AI models. We also propose a deep transformer, S2R-FireTr, which excels in considering the multi-modal environmental information for forecasting and backtracking the wildfire. S2R-FireTr surpasses state-of-the-art methods in real-world wildfire scenarios.

This paper introduces PPT, a pragmatic transport that achieves comparable performance to proactive transports while maintaining good deployability as reactive transports. Our key idea is to run a low-priority control loop to leverage the available bandwidth left by the reactive transports. The main challenge is to send just enough packets to improve performance without harming the primary control loop. We combine two unconventional techniques: an intermittent loop initialization and an exponential window decrease, enabling us to dynamically identify and fill the spare bandwidth. We further complement PPT's design with a buffer-aware flow scheduling scheme to optimize the average FCT of small flows without prior knowledge of flow size information. We have implemented a PPT prototype in the Linux kernel with ~400 lines of code and demonstrated that compared to Homa, it delivers up to 46.3% lower overall average FCT and even 25%/55.5% lower average/tail FCT of small flows in an Memcached workload.

Voice command systems have been widely deployed on many smart devices for remote control. To further enrich the intelligence of these smart devices, the location of sound plays an important role in context-aware acoustic services. Despite initial steps made toward reliable voice localization, the state of the arts rely on prior knowledge of device location, device orientation and an indoor electronic map. To mitigate this additional cost, this paper presents VoiceMap, an autonomous mapping system of acoustic devices for voice localization. The insight behind VoiceMap is to explore the cooperation of sweeping robots and voice devices. Specifically, the sweeping robot is responsible for exploring the electronic map of the environment, while the microphone array is responsible for localizing the sweeping robot, so that we can establish the positional relationship between them. The core challenges are how to accurately locate the continuously moving robot, and how to synchronize the coordinate systems of the sweeping robot and the voice devices. To this end, we first design an inertial-based super-resolution method to estimate the angle of arrival (AoA) with respect to the robot. Then, we develop an effective coordinate synchronization mechanism, so that VoiceMap can automatically locate the voice devices on the electronic map generated by the robot. Finally, we implement a prototype system using commercial devices, and conduct comprehensive experiments to verify the proposed system. The experimental results show that we can realize a median error of 0.12m in terms of device localization.

Wi-Fi-based noncontact respiratory monitoring technology plays a significant role in smart elderly care by eliminating the need for users to wear additional hardware devices. Although respiration sensing has shown impressive accuracy in ideal scenarios, achieving precise respiration monitoring in interfering scenarios, such as when other users, are engaged in activities, remains challenging. This difficulty arises primarily from the weak nature of the respiratory signal, which is susceptible to interference from activities. To address this issue, this article introduces the RespEnh system, a respiratory monitoring system that utilizes human location information to mitigate interference. First, We enhance the quality of respiratory signal by leveraging the frequency diversity of Wi-Fi signal, effectively visualizing respiratory patterns even amidst interference. Second, we apply short time window shift to remove walking noise in the time domain while preserving respiratory signal characteristics. Finally, we integrate multiantenna signals to enhance spatial domain performance. Experimental results demonstrate that our system achieves a relative anti-interference performance of 2.4 m, even when breathing as far away as 6.4 m. It effectively monitors person’s respiration, with a detection rate of 80% within an error range of 1 bpm.

As the demand for the Internet of Things (IoT) grows, it becomes crucial to possess systems capable of detecting any data leakage used for authentication. Within IoT camera systems based on facial bio-metric recognition, there is a risk of Deepfake Bypassed Facial Feature Authentication due to the widespread use of deepfake video technologies, such as DeepFaceLive and expression manipulation. Traditional Face Anti-Spoofing Detection techniques may struggle to detect real-time deepfake videos within IoT contexts. Moreover, constrained by the scale of Face Anti-Spoofing Detection datasets, current detection models primarily focus on recognizing the entire face in videos, neglecting the inter-component correlations of facial features. However, our investigation indicates that different parts of the face have varying impacts on deepfake detection. To address this issue, we segment the face into several regions within video frames and explore the relationships between these regions. Our approach involves constructing feature graphs that represent such correlations, aiming to leverage the relationships between facial regions and the temporal characteristics of real-time facial manipulation videos for use in live facial detection cameras. Initially, features for each facial region are extracted via Convolutional Neural Networks (CNNs). Subsequently, with these features as vertices and their correlations as edges, a feature graph of the entire video is constructed. Ultimately, a Graph Neural Network (GNN) is employed to determine whether the video has been tampered with. Experiments conducted on several publicly accessible datasets demonstrate that our proposed method outperforms other state-of-the-art Face Anti-Spoofing Detection techniques in most scenarios. Thus, the aforementioned advanced Graph Neural Network model exhibits exceptional performance in real-time deepfake detection tailored for live facial detection cameras.

360° video streaming has gained increasing attraction in the current popular virtual reality, AR, and MR applications, which can provide users with an immersive experience. In tile-based 360° video streaming, edge caching and bitrate selection strategies are jointly designed to improve users’ Quality of Experience (QoE), which incorporates video quality and rebuffer. However, the existing QoE-driven approaches use a unified QoE function to guide the decisions of edge caching and bitrate selection, which neglect the impact of video quality and rebuffer on different categories of 360° videos, thus failing to provide high-average QoE for users. In this article, we propose a multiagent deep-reinforcement-learning-based joint edge caching and bitrate selection strategy for multicategory 360° video streaming to improve users’ average QoE. The key idea is to employ different edge caching and bitrate selection strategies for different video categories to enable fine-grained performance optimization. Based on multicategory 360° video streaming, we first model a joint edge caching and bitrate selection problem as a multiagent cooperative Markov decision process with the goal of maximizing users’ average QoE. Next, an Field-of-View-aware multiagent soft actor–critic (FA-MASAC) algorithm is designed to help agents collaboratively learn optimal edge caching and bitrate selection decisions in a distributed way, in which each video category is treated as an agent. Finally, experimental results on real-world data sets show that our proposed strategy can greatly benefit users’ average QoE compared to existing strategies.

360° video streaming, as an immersive and interactive form of video presentation, has consistently strived for higher video quality and lower latency. A lot of effort has been devoted to improving video quality and reducing latency by utilizing the storage or computing resources of the edge/end layer with edge caching and super-resolution (SR) techniques, respectively. However, it has been conspicuously ignored by existing work that fully leveraging the aggregated caching and computing resources of both edge and end layers can further improve video quality and reduce latency. To this end, in this paper, we propose a MADRL-based Edge-Client collaborative Caching and SR (ECCSR) strategy for high-quality low-latency 360° video streaming. First, we construct a Quality of Experience (QoE) function that involves not only video quality, temporal smoothness, and rebuffering time, but also device energy consumption. Subsequently, we formulate the problem of edge-client collaborative caching and SR as a multi-agent cooperative Markov decision process with the goal of maximizing users’ average QoE. Furthermore, to cope with the decision coupling between agents, an adaptive learning-based multi-agent double actors regularized critics (AL-MADARC) algorithm is developed to help agents make optimal collaborative caching and SR decisions. Through extensive experiments using real-world datasets, we show that ECCSR makes a great improvement in users’ average QoE compared to existing strategies.

Wireless signal analytics in IoT systems can enable various promising wireless sensing applications such as localization, anomaly detection, and human activity recognition. As a matter of fact, there are significant correlations in terms of dimension, spatial and temporal aspects among wireless signals from multiple sensors. However, none of the wireless sensing research currently in use directly incorporates or exploits the signal correlations. Therefore, there is still substantial scope for improvement in regards to accuracy and robustness. We are introducing a novel framework called Signal Correlation Learning (SCL). This framework utilizes a directed graph to explicitly represent the signal correlation across various wireless sensors. We use signal embedding to depict the correlation features of a multi-dimensional sensor that arise from a multi-sensor system. Then, we perform Kullback-Leibler (KL) divergence on embedding vectors of any pair of sensors in the system to construct a subgraph at a given time point, which can measure the spatial signal correlation of sensors. Subsequently, several subgraphs spanning a specific time frame are fused into a coherent universal graph based on the small-world theory. This universal graph represents the three types of signal correlation simultaneously. A signal correlation aggregation structure is utilized to extract the features from the universal graph. These features can be used to address target sensing problems. We implement SCL in real RFID, Bluetooth, WIFI, and Zigbee systems, and evaluate its performance in three common wireless sensing problems including localization, anomaly detection, and human activity recognition. Extensive experiments demonstrate that our SCL framework significantly outperforms state-of-the-art wireless sensing algorithms by increasing$80\%\sim 190\%$in terms of accuracy, and by increasing$160\%\sim 220\%$in terms of robustness.

Sharding is an important technology that utilizes group parallelism to enhance the scalability and performance of blockchain. However, the existing solutions use a historical transaction-based approach to reallocate shards, which cannot handle temporary overload and incurs additional overhead during the reallocation process. To this end, this paper proposes LMChain, an efficient load-migratable beacon-based sharding blockchain system. The primary goal of LMChain is to eliminate reliance on historical transactions and achieve the high performance. Specifically, we redesign the state maintenance data structure in Beacon Shard to effectively manage all account states at the shard level. Then, we innovatively propose a load-migratable transaction processing protocol built upon the new data structure. To mitigate read-write conflicts during the selection of migration transactions, we adopt a novel graph partitioning scheme. We also adopt a relay-based method to handle cross-shard transactions and resolve inter-shard state read-write conflicts. We implement the LMChain prototype and conducted experiments in a real network environment comprising 17 cloud servers. Experimental results show that, compared with state-of-the-art solutions, LMChain effectively reduces the average transaction wait latency of overloaded transactions by 30% to 48% in different cases within 16 transaction shards, while improving throughput by 3% to 10%.

Multi-dimensional performance evaluation is crucial for blockchain systems as it enables appropriate blockchain choosing for a given scenario and helps to pinpoint the bottleneck module of a blockchain system to optimize its performance. However, the existing evaluation frameworks for blockchain suffer from low system generality, inefficient workload execution, and incomprehensible evaluation metrics. In order to overcome their limitations, we design and implement the Generalized and Fine-grained Blockchain Evaluation (GFBE) framework. Specifically, we abstract 3 types of Universal Evaluation Interface (UEI) via the dynamic proxying approach to enable generalized evaluation of heterogeneous blockchain systems. Through the design of Lua-based workloads plugin with high flexibility and reusability, GFBE improves the efficiency of workload execution. To achieve comprehensive measurement, we define 15 key performance metrics across hierarchical layers of blockchain architecture. We also implement and deploy GFBE on 16 machines each with 8 CPUs and 16GB RAM, and evaluate three open-source blockchain systems namely Ethereum, ChainMaker, and Haihe smart chain. The experimental results demonstrate that GFBE efficiently and accurately measure 15 key performance metrics such as Contract Execution Efficiency at the contract layer, Consensus Agreement Time Ratio at the consensus layer, and State Query Time at the data layer. Compared with state-of-the-art frameworks such as BLOCKBENCH, Log-based, and Caliper, GFBE distinguishes itself as the only framework that encompasses the appealing features of universal interface, reusable workload, and all-layer metrics.

Many existing datacenter transports perform one-shot packet priority tagging at end-hosts and leave them fixed during the packet's transmission. In this paper, we experimentally show that: (1) such fixed packet priority is not sufficient for FCT (flow completion time) minimization, and (2) adjusting packet transmission priority in the network requires effective coordination among switches. Building on these insights, we present Anole, a new datacenter transport that advocates packet re-prioritization in near-bottleneck switches to minimize FCT. To this end, Anole integrates three simple-yet-effective techniques. First, it employs an in-network telemetry (INT) based approach to dynamically detect the bottleneck for each flow. Second, it adopts an on-off rate control mechanism for each sender to pause heavily congested flows but send lightly- and non-congested ones. Last, it leverages an altruistic scheduling policy at each switch to let the flows whose next hops are bottleneck switches give way to others. We implement an Anole prototype based on DPDK and show, through both testbed experiments and simulations, that Anole delivers significant performance advantages. For example, compared to EPN, Homa, and Aeolus, it shortens the average FCT of all (small) flows by up to 61.6% (89.1%).

The ubiquitous Internet-of-Things (IoT) devices generate vast amounts of multimodal data, and the deep multimodal fusion network (DMFN) is a promising technology for processing multimodal data. Deploying DMFNs locally on embedded IoT devices is a profitable way to provide privacy-preserving and robust sensing services. However, the current compression methods suffer from the following limitations: First, they are designed based on unimodal networks or specific model structures. Hence, it is hard to extend these methods to diverse DMFNs; Second, existing works never relate their efforts to disparate computational demands of multimodal data and modalities. Easy samples and redundant modalities consume the same computational resources as powerful modalities and complex samples. We propose anAdaptiveCompressionFramework (ACF) for DMFNs to address those challenges. It enables input-dependent runtime compression locally on resource-constrained embedded devices. Specifically, we propose an offline model transformation module to upgrade the static network with two kinds of dynamic components to support online structural adjustment. Then we design a lightweight policy network to generate multi-granularity and data-dependent compression strategies for different model parts. Finally, we evaluate ACF on four DMFNs across three embedded platforms. Compared with the best results of the existing schemes, ACF obtains up to 2.61× latency reduction and 2.30× energy consumption reduction, with up to 3.57% accuracy improvement.

Fine-grained recognition of human manipulation activities on objects is crucial in the era of human-computer-object integration. However, there is a lack of solutions for simultaneous recognition of human identity, manipulation activities (including drawing and rotation), and manipulated objects. Therefore, we propose an RF-Camera system that combines RFID and computer vision techniques to address this challenge in multi-person and multi-object scenarios. In RF-Camera, we employ a skeleton-assisted method to extract facial images of target individuals, enabling precise recognition of their identities. To identify manipulation activities, we analyze the 3D hand trajectory and fingertip vector angle, differentiating drawing and rotation manipulation activities. Additionally, we model target person?s hand movements to predict phase data of the target tag, enabling the determination of person-object relationships. Implementing RF-Camera using COTS RFID and Kinect devices involves overcoming challenges such as extracting effective data from noisy streams, predicting virtual phase data considering hand-tag offset, and ensuring high tag reading rates in tag-dense scenarios. We conducted experiments involving six participants performing object manipulation activities, including drawing letters/symbols and rotating movements. Extensive experimental results show that RF-Camera achieves over 90% accuracy in recognizing person identity, manipulation activities, and person-object matching in most conditions.

The coexistence of RDMA and TCP is prevalent in the datacenter. Despite the sound isolation at the end hosts, they share the same switches in the network. Their different networking behaviors (E.g., in hardware demand and transport protocols) lead to huge differentiated buffer demand for switches. However, existing buffer management schemes ignore these dissimilarities and simply treat such RDMA/TCP mix-flows as the typical multi-class traffic, resulting in inferior isolation and degrading networking performances. This paper presents BRT, a first systematic solution for the buffer management of RDMA/TCP mix-flows in the DCN. BRT’s key insight is to allocate buffer with the awareness of traffic’s networking characteristics while minimally impacting the other’s performance. Guided by this insight, it first employs a traffic characteristics-based window to detect whether queues are in the state of persistent long queues. Then, it adjusts the total allocated buffer for each traffic type based on the number of persistent long queues and the normalized dequeue rates to reduce the buffer occupancy of meaningless queuing. Last, it calculates the buffer threshold for RDMA/TCP queues separately and uses a simple yet effective approach to prioritize the absorption of small flows. Our large-scale packet-level evaluations show that BRT can effectively optimize the networking performances for RDMA/TCP mix-flows. For example, compared to current practice, BRT achieves up to 53.5%, 46.7%, and 48.5% lower average FCT for incast flows, RDMA small flows, and TCP small flows, respectively, without sacrificing the overall throughput.

Cloud service providers improve resource utilization by co-locating latency-critical (LC) workloads with best-effort batch (BE) jobs in datacenters. However, they usually treat multi-component LCs as monolithic applications and treat BEs as “second-class citizens” when allocating resources to them. Neglecting the inconsistent interference tolerance abilities of LC components and the inconsistent preemption loss of BE workloads can result in missed co-location opportunities for higher throughput. We present Rhythm , a co-location controller that deploys workloads and reclaims resources rhythmically for maximizing the system throughput while guaranteeing LC service’s tail latency requirement. The key idea is to differentiate the BE throughput launched with each LC component, that is, components with higher interference tolerance can be deployed together with more BE jobs. It also assigns different reclamation priority values to BEs by evaluating their preemption losses into a multi-level reclamation queue. We implement and evaluate Rhythm using workloads in the form of containerized processes and microservices. Experimental results show that it can improve the system throughput by 47.3%, CPU utilization by 38.6%, and memory bandwidth utilization by 45.4% while guaranteeing the tail latency requirement.

A wide range of scenarios, such as warehousing, and smart manufacturing, have used RFID mobile robots for the localization of tagged objects. The state-of-the-art RFID-robot based localization works are based on the premise of stable speed. However, in reality this assumption can hardly be guaranteed because Commercial-Off-The-Shelf (COTS) robots typically have inconsistent moving speeds, and a small speed inconsistency will cause a large localization error. To this end, we propose a Speed Inconsistency-Immune approach to mobile RFID robot Localization (SILoc) system, which accurately locates targets when the robot moving speed varies or is even unknown. We propose an optimized unwrapping method to maximize the use of data, and a lightweight algorithm to calculate the locations in both 2D and 3D spaces. By utilizing the characteristics of tag-antenna distance and combining the phase data from multiple antennas, SILoc can effectively eliminate the side effects of speed inconsistency. To increase the flexibility, we further optimize the system and propose SILoc$+$, which enables the system to achieve localization with part of the data, keeping speed inconsistency-immune. Extensive experiments demonstrate that SILoc and SILoc$+$can achieve a centimeter-level localization accuracy in the scenario with an inconsistent or unknown robot moving speed.

Caching function is a promising way to mitigate coldstart overhead in serverless computing. However, as caching also increases the resource cost significantly, how to make caching decisions is still challenging. We find that the prior "local cache control" designs are insufficient to achieve high cache efficiency due to the workload skewness across servers.

With the advent of network function virtualization (NFV) and mobile edge computing (MEC), outsourcing network functions (NFs, i.e., firewall) to the MEC is becoming popular among network service providers. Notably, NF outsourcing raises an essential security concern about whether these outsourced NFs and associated service function chains (SFCs) are correctly implemented according to enterprises’ specifications. In particular, SFC with parallel VNFs, which take advantage of parallelism, have been conducted to reduce the traffic delay of traditional sequential SFC, called the hybrid SFC in this paper. Nevertheless, how to ensure correct behaviors and discover runtime mistakes for hybrid SFC remains an open problem.In this paper, we propose vHSFC, a verification scheme for hybrid SFC, enabling enterprises to verify the correctness of SFC enforcement in real-time. vHSFC achieves its goal with a lightweight verified routing protocol, which detects various hybrid SFC violations and attacks, i.e., packet modification, incompliant forwarding path, etc. To demonstrate the feasibility and performance of vHSFC, we have implemented the prototype on top of several containers and conducted extensive experiments with real traffic. The experimental results show that our vHSFC can continuously ensure proper enforcement and discover unexpected violations while incurring sensible overhead.

Incast happens in data center networks (DCNs) when many senders simultaneously send flows to one receiver. Incast occurs frequently and compromises flow performance severely. However, widely deployed end-to-end congestion control (CC) protocols are inefficient at handling incast, as they rely on delayed congestion signals and can not isolate incast flows. Recently, researchers have favored per-hop flow control protocols since they could achieve fast reaction and incast isolation. Nevertheless, these impressive advantages are impractical because of hardware resource limitations.In this paper, we present a new per-hop flow control protocol dBFC (Destination-based Backpressure Flow Control), which achieves these advantages within limited hardware resources. Our key insight is that the number of destinations reached by queued packets in a port is minuscule. dBFC is compatible with existing CC protocols. We use large-scale NS3 simulations to evaluate dBFC. In our evaluation, compared with deployed CC protocols and the state-of-the-art per-hop flow control protocol, dBFC reduces the maximum buffer occupancy by 5 − 23× and thus provides up to 21× lower flow completion time of small flows in limited hardware resources.

Recently, asynchronous byzantine fault tolerance (BFT) consensus has made progress in linearized communication complexity, allowing nodes to broadcast a small number of transactions at their own pace in networks with variable bandwidth. However, this means that network propagation delay poses a more severe bottleneck to the system compared to transmission delay affected by block size and bandwidth. Particularly in networks with high propagation delays, nodes need to wait more for messages to propagate through the network, which severely hinders system performance. We propose Swarm, an asynchronous BFT protocol that is suitable for networks with tight bandwidth and propagation delay. The new protocol structure supports highly concurrent propagation-delay-intensive and propagationdelay-tolerant stages by separating the control and data layers and ensures that nodes achieve an ordered and consistent log for parallel blocks. We also design a group broadcasting strategy that allows nodes to not broadcast proposals from other nodes without compromising safety and liveness while reducing communication cost. Importantly, our work is orthogonal to existing improvements based on communication complexity and bandwidth and does not sacrifice existing results. We built a complete system prototype and conducted a thorough evaluation. The results show that Swarm performs better in networks with higher propagation delays, achieving 2-5x TPS and reducing latency by 50% compared to existing asynchronous protocols for different scales.

Reinforcement Learning (RL) algorithms require large amounts of computational resources and time to train due to the simultaneous model training and real-time interactions with simulation environments. This results in an RL algorithm's training time potentially exceeding days to months. We present HRL, a comprehensive optimization system designed to improve the system throughput of RL algorithms. HRL addresses the challenges of discovering and identifying bottlenecks in the sampling, training, or communication stages and promptly improving their performance. We propose a group-parallel pipeline method to improve the sampling efficiency and apply data quantization and multi-learner training to resolve the network and learning bottlenecks. The HRL system has been fully implemented and integrated with XingTian. The results of the experiments show that the HRL system can improve the throughput by 18.6%-90.6%.

In recent years, multimodal underwater wireless sensor networks (M-UWSNs) have attracted widespread concern in academia. Due to the complex underwater communication environment and the increasing marine applications, it is a crucial issue for M-UWSNs to design an energy-efficient routing strategy that can satisfy multiple transmission latency requirements of different marine applications. Reinforcement learning (RL) approaches with distributed dynamic optimization ability provide a prospective way to solve the aforementioned problem. Therefore, we propose an improved RL framework and then design an energy-efficient multilevel routing strategy (MLRS-RL) based on this framework for multiple transmission latency requirements. In MLRS-RL, a method of model knowledge collection based on the time backoff principle is proposed to preliminarily learn the network environment information before network operates. The convergence speed of the RL framework can be accelerated by utilizing the model knowledge. Then, underwater nodes use the improved RL model to calculate the transmission rewards that data packets with different transmission latency requirements are sent to different candidate relay nodes. Finally, a cooperative transmission strategy using multiple relay nodes is designed to further improve the reliability of data transmission. We verify the effectiveness of the MLRS-RL strategy in terms of packet delivery ratio, transmission latency, energy efficiency, network lifetime, and delivery quantity.

This article focuses on the topology optimization problem to decrease transmission delay and prolong network lifetime on the premise of reliable transmission in underwater wireless sensor networks (UWSNs). This is extremely challenging owing to dynamic ocean current movement, harsh underwater communication channel, and increasingly demanding application requirements. With the development of software-defined network architectures, the centralized topology control (TC) strategy with a global perspective in UWSNs is expected to become a more effective way to tackle the above challenges compared with the existing distributed and heuristic TC strategies involving local network state information. Therefore, we first transform the topology optimization problem of UWSNs into an integer nonlinear programming (INLP) model and design a centralized TC framework to solve the INLP model. In this framework, a TC center is built to periodically generate the network topology for UWSNs according to the current network state information. Further, an efficient topology generation algorithm based on deep reinforcement learning (TGA-DRL) is proposed in the TC center. In TGA-DRL, to reduce computing overhead and improve operational efficiency, we formulate an action-space narrowed Markov decision process suitable for network topology generation and solve it with the aid of the rainbow algorithm which is a deep reinforcement learning model. Finally, the performance of our centralized TC framework is verified in terms of the node out-degree, algorithm convergence, and optimization effect.

Owing to the attractive properties of decentralization, unforgeability, transparency, and traceability, blockchain is increasingly being used in various scenarios such as supply chain and public services, where massive Spatial-Temporal-Keywords (STK) transactions need to be packaged. However, due to the multi-dimensionality and randomness of STK transactions, existing solutions fail to enable queries in a verifiable and efficient way for blockchains storing multidimensional transactions. To this end, this article takes the first step to propose an authenticated and efficient query approach in hybrid blockchain systems consisting of on-chain and off-chain parts. We first design a data structure named MRK-Tree in the block body, which organizes STK transactions for efficient nodes pruning of both kNN and range queries. Then we propose an improved block header, which improves the efficient pruning of blocks on the basis of ensuring the authentication of query results. Also, we design a cross-block searching algorithm named Efficient Block Pruning (EBP) and intra-block searching algorithms named Authenticated kNN/Range Query (AKQ/ARQ) to accelerate authenticated queries for multiple MRK-Trees in the hybrid blockchain systems. Authentication mechanisms are proposed to ensure the soundness and completeness of query results. Rigorous security analysis validates the practicability of the proposed approach. We build a blockchain prototype to comprehensively evaluate the performance of proposed query schemes. Extensive evaluation results with real datasets reveal that our approach can ensure authenticated queries, meanwhile improving the time efficiency by up to 36.45x and space efficiency by up to 4 orders of magnitude compared with the well-known benchmark query schemes.

Domain generation algorithms (DGAs) can be categorized into three types:zero-knowledge,partial-knowledge, andfull-knowledge. While prior research merely focused onzero-knowledgeandfull-knowledgetypes, we characterize their anti-detection ability and practicality and find thatzero-knowledgeDGAs present low anti-detection ability againstdetectors, andfull-knowledgeDGAs suffer from low practicality due to the strong assumption that they are fullydetector-aware. Given these observations, we proposePKDGA, a partial knowledge-based domain generation algorithm with high anti-detection ability and high practicality.PKDGAemploys the reinforcement learning architecture, which makes it evolve automatically based only on the easily-observable feedback from detectors. We evaluatePKDGAusing a comprehensive set of real-world datasets, and the results demonstrate that it reduces the detection performance of existingdetectorsfrom 91.7% to 52.5%. We further applyPKDGAto theMiraimalware, and the evaluations show that the proposed method is quite lightweight and time-efficient.

Mobile edge computing (MEC) is becoming increasingly popular due to its remarkable computing capacities in close proximity to end users or devices. With the widespread use of Industrial Internet of Things, more and more cloud service providers move their services to the edge of the network for a better quality of service and become MEC providers. These MEC providers require to rent wide area network (WAN) connections to transfer industrial data, which is a considerable expense. In this article, we propose a framework calledSublessorto reduce the WAN transmission cost for a group of cooperative MEC providers. The key idea ofSublessoris allowing some specific MEC providers to act as Internet transit brokers, transmitting not only their own network traffic but also the traffic of their partners under a reasonable reselling price. This article formulates the problem as a mixed-integer programming and finds the most suitable broker number and corresponding reselling price without damaging the profit of both brokers and partners by a deep-reinforcement-learning-based algorithm. Experimental results show that our algorithm can significantly reduce the traffic transmission cost by up to 35%.

Major manufacturers and retailers are increasingly using RFID systems in supply-chain scenarios, where theft of goods during transport typically causes significant economic losses for the consumer. This paper studies how to achieve time-efficient and secure integrity authentication problems in RFID systems. We start with a straightforward solution called SecAuth, which uses a secure identity stored on reserved memory to authenticate tags in a secure way. We then propose a time efficient KTAuth protocol, which design a verification chain mechanism to efficiently verify a small set of key tags using limited on-tag memory. We point out that the limitation of KTAuth is that it takes too much overhead to write a large block of data to tag memory, which leads to the proposed group selection mechanism. The KTAuth with group selection (KTAuth-GS) enables you to select key tags with a single select command, which helps to quickly check the existence of key tags and reduces the data writes on the tags. Experiments and simulation results demonstrate that the proposed KTAuth-GS can defend against counterfeiting attacks by providing more reliable results and reducing the execution time by as much as a factor of 5 when compared with a baseline tag identification protocol.

Mobile edge computing (MEC) is a key enabler for ultra-low latency in heterogeneous dense cellular networks in the 5G era and beyond, by deploying services at the network edge. Due to high user mobility, the services are usually migrated to follow the users by predicting the user trajectory to achieve a balance between energy consumption and service latency. However, service migration for multi-user heterogeneous dense cellular networks is challenging because (1) the user trajectory prediction, which is crucial for service migration, becomes intractable with a large number of users, and (2) making service migration decisions for each user independently is subjected to interference among the users. Therefore, in this study, we formulated the service migration of all the users in MEC-enabled heterogeneous dense cellular networks as an optimization problem, with the objective of minimizing the average energy consumption while satisfying the service latency requirements, taking into account the interference among different users. Next, we developed an efficient energy-efficient online algorithm based on the Lyapunov and particle swarm optimizations, called EGO, to resolve the original problem without predicting the trajectories of the users. Finally, a series of simulations based on real-world mobility traces of vehicles in Bologna were conducted to establish the superiority of the EGO algorithm over state-of-the-art solutions.

Container overlay network, though being widely adopted to enable communication between containers on different hosts, is a key downside for latency-sensitive applications. The state-of-the-art solution seeks to shorten the data path in packet processing by replacing overlay connection file descriptors with host namespace ones. While promising, it must block each overlay connection until the relevant host connection is set up, thus heavily influencing the request latency. In this paper, we present ShuntFlow, a systematic data delivery framework that seamlessly integrates the host and overlay networks to reduce the application's request-response latency. ShuntFlow first lets all connections flow in the overlay network directly. Then, it adopts a simple-yet-effective syscall-threshold-based mechanism to pick appropriate connections and switches their data delivery to the host network in a blocking-free way using a multi-threading technique. As such, unnecessary connection switches are prevented; yet, the pre-setup phase dilemma is eliminated. We have implemented a ShuntFlow prototype based on Linux and Docker and evaluated it extensively on a 40 Gbps testbed. The results show that ShuntFlow achieves 13%/72% and 19%/69% reductions, in average/tail request-response latency of a web server and an in-memory key-value store, respectively, while incurring less CPU overhead, compared to Slim.

Decentralized federated learning (DFL) is a promising technology to implement distributed machine learning in Internet of Vehicles (IoV), which enables vehicles to share and aggregate models with their neighbors in a vehicle-to-vehicle (V2V) network. However, due to the high mobility of vehicles, model sharing via V2V links may fail as the topology of the V2V network is time-varying, which greatly reduces the efficiency of model aggregating and the speed of model training. To address this problem, in this paper, we propose a graph convolution network (GCN)-based topology design method, named G-DFL, to improve the training efficiency of DFL in IoV by properly selecting a subgraph of the underlay V2V network, which is referred to as overlay network, in each round of model sharing. First, by encoding the state of vehicles, we utilize a GCN to extract the features of V2V network topology to predict the effective V2V links for model sharing. In addition, to further reduce the delay of model training, we use Christofides' Algorithm to find the Hamiltonian circuit with the least delay as the overlay network. Simulation results validate that the proposed method significantly improves the model training performance in DFL compared with the other baseline methods.

Modern websites increasingly rely on machine learning (ML) to improve their business efficiency. Developing and maintaining ML services incurs high costs for developers. Although serverless systems are a promising solution to reduce costs, we find that the current general purpose serverless systems cannot meet the low latency, high throughput demands of ML services.

Federated Mobile CrowdSensing is applied to collect massive sensory data and exploits the computing power of mobile devices brought by their embedded specialized computing engines (e.g., Neural Engine in iPhone) to train machine learning (ML) models. However, the heterogeneity of mobile devices includes significant differences in the size and quality of datasets, different computing power, and some unreliable clients using unreliable data for training. The heterogeneity of mobile devices reduces FL’s performance. Therefore, selecting high-quality clients for Federated learning (FL) is vital. This study proposes a client selection scheme based on the reverse auction. First, each client’s training time is predicted, the total FL time threshold is optimized, and the reputation value is calculated based on the historical performance of each client. Then, each client’s current computing power and dataset size are converted into an efficiency value. Finally, the selection value of each client is calculated based on the efficiency value and reputation value. The results of the experiments show that our scheme can select high-quality clients. Compared with FedRep, our scheme can reduce training time by 91.5%. Compared with FedEff, our scheme can reduce communication rounds by 87.5%. In the same communication rounds (5000), our scheme has higher accuracy than RandomFL, and the average accuracy is improved by about 4.4%.

In mobile video streaming, the ever-increasing user expectations for Quality of Experience (QoE) have prompted the integration of video super-resolution and adaptive bitrate techniques on either the mobile device or the edge server. By reconstructing high-resolution frames from low-resolution frames that have been downloaded, both high video quality and a short rebuffer time can be enjoyed. However, the exiting methods merely leverage the computing resources of the edge server or mobile device, leaving significant room for further QoE improvement. In this paper, we present an adaptive Video Streaming system with Edge-Client collaborative Super-resolution, named VSECS, to enhance users' QoE by simultaneously utilizing the computing resources of both the edge server and mobile device to reconstruct high-resolution frames collaboratively. First, we deploy a large-scale super-resolution model on the edge server and a lightweight model on the mobile device. Then, we exploit the Asynchronous Advantage Actor-Critic (A3C) algorithm to make decisions regarding the download resolution, the reconstructed target resolution, and the workload share of the mobile device, considering the network bandwidth, computing resources, and reconstruction complexity of video tiles. Furthermore, we utilize the branching actor network to enable the agent to converge to good policy stably. Trace-driven simulations on real-world bandwidth traces demonstrate that our approach can improve QoE by up to 10% compared to the state-of-the-art video streaming solutions.

Proactive transport has been attractive in modern high-speed and shallow buffered datacenter networks. At its heart, the link capacity is proactively allocated as credit, and then following scheduled packets are triggered by active senders according to credits, providing (near) zero loss rate and extremely low latency. Despite being promising, when waiting for credits in the first RTT, called the “pre-credit phase a substantial amount of spare bandwidth is being underutilized. To bridge this gap, current practices send a Bandwidth-Delay-Product (BDP) worth of unscheduled packets with line rate in the pre-credit phase, however, degrading throughput or tail latency. One key insight is that they consider the spare bandwidth in the pre-credit phase as a fixed value but actually changes across time and space. In this paper, we present Schef, a novel switch-based bandwidth calculation mechanism to address the pre-credit phase challenge without introducing network congestion or throughput degradation. The main idea is to calculate the spare bandwidth and apply efficient control of unscheduled packets at switches. Schef is compatible with existing proactive transports. We integrate it into a representative proposal NDP and evaluate its performance through large-scale simulations. Compared with the state-of-art, Schef reduces the average and 99th flow completion time (FCT) of short flows by up to 23% and 31%, and achieves 13% higher goodput simultaneously.

Computing systems always face a “resource allocation dilemma” that shows the great difficulties in trading off resource efficiency for tail latency, due to the internal uncertainty of cluster status and execution behavior. Inspired by the imaginary “Maxwell’s demon” in thermodynamics who can reduce the uncertainty through a per-gas molecule-level control policy, we consider the “one-to-one mapping” feature of serverless computing and build a novel resource allocator, named Maxwell, that can achieve low tail latency and high resource efficiency in serverless simultaneously. Like the “Maxwell’s demon Maxwell is able to optimize the resource allocation for every request. It observes the state of each request and makes decisions about the minimum resource allocation through a reinforcement learning predictor. As the per-request-grained control incurs significant overhead, we further design a pipeline for avoiding the accumulated effect on a workflow. Experimental results show that Maxwell not only saves up to 31% CPU resources but also reduces the standard deviation of latency by 1.9×. Its time overhead is negligible and the resource overhead is also limited when the query per second $\leq$500.

Mobile robot-assisted book inventory such as book identification and book order detection has become increasingly popular in smart library, replacing the manual book inventory which is time-consuming and error-prone. The existing systems are either computer vision (CV)-based or RFID-based, however several limitations are inevitable. CV-based systems may not be able to identify books effectively due to low accuracy of detecting texts on book spine. RFID tags attached to books can be used to identify a book uniquely. However, in high tag density scenarios such as library, tag coupling effects of adjacent tags may seriously affect the accuracy of tag reading. To overcome these limitations, this paper presents a novel RFID and CV fusion system for Book Inventory using mobile robot (RC-BI). RFID and CV are first used individually to obtain book order, then the information will be fused by the sequence based matching algorithm to remove ambiguity and improve overall accuracy. Specifically, we address three technical challenges. We design a deep neural network (DNN) model with multiple inputs and mixed data to filter out interference of RFID tags on other tiers, and propose a video information extracting schema to extract book spine information accurately, and use strong link to align and match RFID- and CV-based timestamp vs. book-name sequences to avoid errors during fusion. Extensive experiments indicate that our system achieves an average accuracy of 98.4% for tier filtering and an average accuracy of 98.9% for book order, significantly outperforming the state-of-the-arts.

The consensus mechanism plays a pivotal role in guaranteeing the security and consistency of blockchain systems and substantially affects system performance. However, an increasing number of blockchain nodes degrade the consensus performance dramatically because of the high communication complexity in traditional consensus mechanisms. In this paper, we propose NS-consensus, a secure node-sampling blockchain consensus mechanism reducing the communication complexity significantly. The key novelty lies in the sampling of blockchain nodes so that the leader only needs to interact with the sampling nodes in each consensus epoch. However, NS-consensus imposes two challenges in determining an optimal sample size and denying malicious proposals. To address the challenges, we determine the sample size under the constraints of a confidence level and a margin of error to enhance communication efficiency without compromising system security. Furthermore, we design a mechanism to enable the leader to interact with all blockchain nodes in the last consensus phase, ensuring the denial of malicious proposals. The extensive experimental results indicate that NS-consensus outperforms the state-of-the-art with up to 175.1% higher system throughput and 79.9% lower time overhead in the sampling phases.

With the wide deployment of RFID in various scenarios such as warehouse management, freight transportation, and manufacturing, tag authentication is increasingly important due to the threat of counterfeit tags. Recent physical-layer authentication approaches have demonstrated that the subtle differences in the hardware features offer a unique fingerprint to authenticate a tag. Although the state-of-the-art approaches are effective in laboratory environments, they are difficult for practical deployment because they either require complex analysis of the raw signal propagation or restrict the geometrical positions of the tags. In this paper, we propose an RFID tag authentication based on frequency- and orientation-related phase fingerprints, called FopPrint, which does not require raw signal analysis or complex geometric relationship. FopPrint uses the phase values of tags at different frequencies and orientations to construct feature matrices as physical-layer fingerprints and uses a pair of adjacent tags as identifiers of each object. FopPrint can effectively eliminate the influence of environmental factors by using the feature matrix constructed by the phase difference. We implement a prototype of FopPrint using Commercial-Off-The-Shelf (COTS) RFID devices. Extensive experimental results show that FopPrint achieves high authentication accuracy of 94% in various experimental settings.

Mobile Crowd Sensing (MCS) typically assigns sensing tasks in the same target area to many participants considering data quality and the diversity of sensing devices. However, participant selection is based on the individual in many research. The efficiency of individual recruitment is low. Individuals need higher transportation costs to go to the task location alone, and the data quality perceived by individuals is difficult to guarantee. This paper proposes a team-based multitask data acquisition scheme under time constraints to address these challenges. The scheme optimised the number of participants, traffic cost, and data quality and designed four team-based multitask allocation algorithms under time constraints in the MCS: T-RandomTeam, T-MostTeam, T-RandomMITeam, and T-MostMITeam. The team size is associated with the number of participants required for the first task or the vehicle capacity to perform the task. We conducted extensive experiments based on a real large-scale dataset to evaluate the four algorithms' performances compared to two baseline algorithms (T-Random and T-most). The efficiency of the four algorithms has been significantly improved by team recruitment. The transportation cost can be multiplicatively reduced by carpooling. Data quality can be improved by at least 2% through reputation screening and team members' communication.

Federated learning (FL) plays an important role in the development of smart cities. With the evolution of big data and artificial intelligence, issues related to data privacy and protection have emerged, which can be solved by FL. In this paper, the current developments in FL and its applications in various fields are reviewed. With a comprehensive investigation, the latest research on the application of FL is discussed for various fields in smart cities. We explain the current developments in FL in fields, such as the Internet of Things (IoT), transportation, communications, finance, and medicine. First, we introduce the background, definition, and key technologies of FL. Then, we review key applications and the latest results. Finally, we discuss the future applications and research directions of FL in smart cities.

The rapid development of the blockchain leads to a blowout of on-chain transactions, contracts, and currencies, which will further accelerate the increase of data. The existing blockchain systems typically support exact transaction queries, which, however, cannot satisfy the QoS requirements with intelligent adjustment in the blockchain systems. To this end, this paper takes the first step to define and address the practically important problem of transaction cardinality estimation for QoS-adjustable intelligent blockchain systems. We first establish a mathematical relationship between the bit string and transaction cardinality. Thus, we can leverage the number of leading 1s of the obtained bit string to estimate the transaction cardinality. We then improve the block header and body with a corresponding search algorithm to access bit strings in blocks. We also propose an estimation protocol with intelligent adjustable QoS to support accuracy-guaranteed and efficiency-optimized estimation. Finally, we design an authentication scheme and guarantee the reliability of our protocol through rigorous theoretical derivation. When achieving the transaction cardinality estimation in blockchain, two technical challenges need to be addressed. (i) To ensure efficient, verifiable, and overhead-saving bit string accessing mechanism in blockchain, we propose the Merkle Cardinality Tree (MCT) and target block filtering mechanism based on Bloom Filter (BF) in off-chain and improve on-chain block header by joining the abstract of MCT and BF. (ii) To improve estimation efficiency while guaranteeing accuracy requirements in hybrid blockchain scheme, we propose a Dynamic One-round Sampling-based cardinality Estimation (DOSE) protocol and integrate BF-DOSE to intelligently accelerate estimation. We build MCT in Ethereum and store the MCT Root in the block header for estimation authentication. Extensive experiments reveal that our BF-DOSE protocol can well satisfy various accuracy and efficiency requirements of QoS-adjustable intelligent blockchain systems, and is one to two orders of magnitude faster compared with benchmark schemes.

Summary The cyber‐physical system (CPS) has become a promising direction that can enrich the interaction between people and people, people and objects, objects and objects in the physical world and the virtual world. Because of the development of CPS, the number and types of smart device connected to the Internet has rapidly increased, bringing with it the current issues of flexibility, efficiency, availability, security, and scalability of the Internet‐of‐Things (IoT) network. These problems are caused by the key mechanism of large‐scale distribution to the IoT network. In this article, we propose CPSS inspired by software‐defined networking (SDN) and service‐oriented architecture, which is a powerful software‐defined cyber‐physical system to solve the challenges of broader CPS adoption. CPSS has three complementary modules: application‐as‐a‐service (AAAS), network‐as‐a‐service (NAAS), and infrastructure‐as‐a‐service (IAAS). IAAS collects network information from a different set of controllers and devices and generates a domain‐wide network view. NAAS collects information from different domains and generates a global view, which helps data transmission between different domains and enables seamless switching between applications and infrastructure devices. AAAS provides personalized service for each user. To test CPSS, we develop a prototype system. Our experimental results show that CPSS can effectively control cyber‐physical system, and CPSS has greatly improved network latency, network throughput, and network reliability compared to traditional controller‐controlled CPS.

Modern resource management frameworks guarantee low tail latency for long-running services using the resource over-provisioning method, resulting in serious waste of resources and increasing the service costs greatly. To reduce the over-provisioning cost, we present HRSE, a hybrid resource scaling engine that enables much more efficient resource provisioning for both periodic and non-periodic workloads of long-running services while guaranteeing the tail latency Service Level Objective (SLO). HRSE employs a convolution-based time series analysis to identify periodic patterns in workloads. If periodic patterns are discovered, HRSE estimates the just-right amount of resources based on the periodic features through atop-$K$based collaborative filtering approach. Otherwise, it leverages wavelet-clustering to capture the short-term patterns in non-periodic workloads and predict the resource demands for the near future. To further enforce the tail latency SLO, HRSE uses an online reprovisioning mechanism that dynamically adjusts the resources to mitigate the performance uncertainty due to workload burstinesses. We fully implement HRSE on top of Docker and conduct extensive experiments using traces from production systems. Testbed experiments show that HRSE is able to increase the average resource utilization to 43 and 45 percent for periodic and non-periodic workloads respectively while guaranteeing the same tail latency objective.

In the era of cloud computing, network services are deployed on geographically distributed cloud platforms, which results in a large amount of inter-datacenter traffic. Multi-tier pricing schemes are widely adopted by cloud service providers (CSPs) to charge cloud users for inter-datacenter transmission services. To avoid a severe penalty associated with missing a deadline, cloud users are prone to selecting a sufficiently high service level. However, they are usually unaware of the total traffic volume before accessing the network; hence, a high transmission cost is introduced. In this paper, we propose an online cost-efficient transmission scheme for cloud users with information-agnostic traffic. The basic idea is to split a long-term transmission request into a series of short-term ones. In this scheme, we take into account the CSP’s countermeasures, and model the interactions between the cloud users and the CSP as a Stackelberg game. We show that the optimal number of short-term requests and the associated transmission service levels can be determined with an online algorithm based on Lyapunov optimization. The experimental results reveal that the CSP and the cloud users can achieve a win-win outcome, whereby the transmission cost of cloud users can be reduced by 59 percent.

Distributed machine learning (DML) is an increasingly important workload. In a DML job, each communication phase can comprise acoflow, and there are dependencies among its coflows. Thus, efficient coflow scheduling becomes critical for DML jobs. However, the majority of existing solutions focus on scheduling single-stage coflows with no dependencies. While there are a few studies schedule dependent coflows of multi-stage jobs, they suffer from either practical or theoretical issues. Motivated by this situation, we study how to schedule dependent coflows of multiple DML jobs to minimize the total JCT in a shared cluster. We present a formal mathematical formulation for this problem and prove its NP-hardness. To solve this problem without job size information, we present an online coflow-aware optimization framework calledParrot. The core idea inParrotis to infer the job with the shortest remaining processing time (SRPT) each time and dynamically control the inferred job's bandwidth based on how confident it is an SRPT job while being mindful of not starving any other job. Specifically, in the design ofParrot, we present a least per-coflow attained service (LPCAS) policy to infer the SRPT job. We further propose a dynamic job weight assignment mechanism and a linear program (LP) based weighted bandwidth scaling strategy for sharing bandwidth among DML jobs. We have proved thatParrotalgorithm has a non-trivial competitive ratio. The results from large-scale trace-driven simulations further demonstrate that ourParrotcan reduce the total JCT by up to 58.4 percent, compared to the state-of-the-art Aalo solution.

The Inter-datacenter transfer is a fundamental service for global cloud applications. Geo-distributed data centers become an essential resource for their application performance which may be destroyed by network congestion. Recent years, most inter-datacenter transfer methods focus on allocating transfers by bandwidth allocation to achieve low cost or high utilization. However, the congestion condition is rarely considered in these works. In this article, we introduce a congestion-aware traffic allocation method named CONA (CONgestion-Aware), whose target is to maximize the profit of allocation transfer among multiple data centers. On this purpose, a maximizing optimization model is proposed, and an efficient link grading strategy is presented. A matrix transformation method is also introduced to simplify the optimization problem. Furthermore, the link congestion condition is considered by the controller, as well as the prediction on link congestion. To verify our proposed method, simulation model is established and comprehensive experiments are conducted. The experimental results show that our method brings higher profit than fair share and greedy traffic allocation method.

In the era of global-scale services, analytical queries are performed on datasets that span multiple data centers (DCs). Such geo-distributed queries generate a large amount of inter-DC data transfers at run time. Due to the expensive inter-DC bandwidth, various methods have been proposed to reduce the traffic cost in geo-distributed data analytics. However, current methods do not attempt to address the throughput issue in geo-distributed analytics. In this article, we target at characterizing and optimizing a cost-throughput tradeoff problem in geo-distributed data analytics. Our objectives are two-fold: (1) we minimize the inter-DC traffic cost when serving geo-distributed analytics with uncertain query demand, and (2) we maximize the system throughput, in terms of the number of query requests that can be successfully served with guaranteed queuing delay. Specifically, we formulate a stochastic optimization problem that seamlessly combines these two objectives. To solve this problem, we take advantage of Lyapunov optimization techniques to design and analyze a two-timescale online control framework. Without prior knowledge of future query requests, this framework makes online decisions on input data placement and admission control of query requests. Rigorous theoretical analyses show that our framework can achieve a near-optimal solution and maintain system stability and robustness as well. Extensive trace-driven simulation results further demonstrate that our framework is capable of reducing inter-DC traffic cost, improving system throughput, and guaranteeing a maximum delay for each query request.

The cloud resource provision policy of a content provider in the presence of competitors on globally distributed cloud platforms plays a significant role in maximizing its profit. However, developing an optimal resource provision policy is quite challenging, due to the difficulty to capture the competition relationship between two competitive CPs and to obtain the budget of the competitors which is usually kept private. To solve this problem, in this article, we propose a learning-driven cloud resource provision policy for a CP with competitors. We formulate the competition between the CPs as alottery Colonel Blottogame in which the payoff of each region is positively related to the resource advantage achieved by the CP, formulate the budget allocation problem as a Markov decision process, and obtain the sub-optimal resource provision policy by reinforcement learning and deep reinforcement learning-based algorithms. We also prove the convergence of the sub-optimal solution. Finally, we validate our proposed method using real-world CPs statistics. The results show that the budget information is critical for a CP to make policy decisions, and it is better for CPs with smaller budget to focus their budget resources in regions with higher values.

Hashing has been widely applied to the large-scale approximate nearest neighbor search problem owing to its high efficiency and low storage requirement. Most investigations concentrate on learning hashing methods in a centralized setting. However, in existing big data systems, data is often stored across different nodes. In some situations, data is even collected in a distributed manner. A straightforward way to solve this problem is to aggregate all the data into the fusion center to obtain the search result (aggregating method). However, this strategy is not feasible because of the prohibitive communication cost. Although a few distributed hashing methods have been proposed to reduce this cost, they only focus on designing a distributed algorithm for a specific global optimization objective without considering scalability. Moreover, existing distributed hashing methods aim at finding a distributed solution to hashing, meanwhile avoiding accuracy loss, rather than improving accuracy. To address these challenges, we propose a Scalable Distributed Hashing (SDisH) model in which most existing hashing methods can be extended to process distributed data with no changes. Furthermore, to improve accuracy, we utilize the search radius as a global variable across different nodes to achieve a global optimum search result for every iteration. In addition, a voting algorithm is presented based on the results produced by multiple iterations to further reduce search errors. Theoretical analyses of communication, computation, and accuracy demonstrate the superiority of the proposed model. Numerical simulations on three large-scale and two relatively small benchmark datasets also show that the SDisH model achieves up to 44.75% and 10.23% accuracy gains compared to the aggregating method and state-of-the-art distributed hashing methods, respectively.

Tag identification is a critical operation in large-scale RFID applications. Typically, in the RFID-enabled warehouse, the reader needs to execute tag identifications to obtain the inventory information of numerous tagged items. The existing schemes usually divide the time frame into multiple slots and map each tag to one of them for replying its identity. This imposes serious tag collisions because two or more tags may be mapped to the same slot and their responses corrupt with each other. When tag collision happens, all the collided tags cannot be identified by the reader, which significantly increases the identification delay. To overcome the collision problem in the identification process, this paper proposes a Group Tag Identification (GTI) framework to identify grouped tags in both singleton and collision slots. The key novelty of GTI is in leveraging tag-correlation to identify grouped tags in the collision slots without any extra transmission overhead. The main challenge of this work is to overcome the communication and architectural limitations of RFID systems in the context of building ID and slot-correlation between tags. Extensive simulations show that GTI significantly reduces the identification delay by up to 40 percent when compared with the state-of-the-art dynamical frame slotted aloha schemes.

In Mobile crowdsensing (MCS), the platform needs an adequate user group to accomplish tasks. Its recruiting worker strategy is essential for sensor data quality. Social-network-assisted worker recruitment effectively expands task coverage. However existing studies enclose two impractical assumptions: influence between users is determined by the number of friends and the recruiting reward is fixed. To solve this problem, a novel influence and cost trade-off (ICT) algorithm is proposed to apply the worker recruitment strategy in the real world. ICT uses linear equations to estimate influence and cost iteratively under the impact of the seed set. Using the influence model based on social interaction, the algorithm selects a near-optimal set of seeds by the revenue-cost-ratio. Empirical studies on three realworld datasets verify that ICT achieves higher performance than baseline methods under various settings.

In this paper, we propose a novel super-resolution and infection edge detection co-guided learning network for COVID-19 CT segmentation (CogSeg). Our CogSeg is a coherent framework consisting of two branches. Specifically, we use image super-resolution (SR) as an auxiliary task, which assist segmentation to recover high-resolution representations. Moreover, we propose an infection edge detection guided region mutual information (RMI) loss, which uses the edge detection results of segmentation to explicitly maintain the high order consistency between segmentation prediction and ground truth around infection edge pixels. Our CogSeg network can effectively maintain high-resolution representation and leverages edge details to improve the segmentation performance. When evaluated on two publicly available COVID-19 CT datasets, our CogSeg improves 10.63 and 13.02 points than the established baseline method (i.e. U-Net) w.t.r mIoU. Moreover, our CogSeg achieves more appealing results both quantitatively and qualitatively than the state-of-the-art methods.

Eye tracking is playing an important role in many applications including human-computer interaction and behavior study. However, the existing approaches have at least one of the following limitations: (i) dedicated devices such as infrared camera and eye-tracker are required; (ii) complex calibration process is involved; (iii) substantial computing resources are consumed; (iv) users suffer from the risk of privacy leakage. To address the above limitations, we propose a H eatmap-based E ye T racking (HETrack) system. One of the key challenges in our system is to design a lightweight model for fine-grained tracking when the computing resources of device is limited. Also, it is necessary to protect user privacy in such a system. To address the above challenging issues, the proposed system consists of the following processes. First, when users randomly look at the screen of the device, HETrack obtains the raw image containing facial information. Then, we design a neural network model and train it with federated learning. The model can map the image to heatmap that implies the possibility of the user’s gaze position on the screen. Finally, HETrack can intercept the real-time video stream into frames, and employ the trained model to generate the heatmap of current frame for gaze estimation. We implement HETrack based on a Commercial-Off-The-Shelf (COTS) camera and conduct extensive experiments to evaluate its performance. Our HETrack system only requires once calibration; whereas, the state-of-the-art work proposed by Google requires 3~5 times calibration on average. Unlike previous approaches that transmit raw image data to a central server, in our HETrack system, only parameters are transmitted, thereby well protecting the user’s privacy. Experimental results demonstrate that the average distance error of estimated gaze point is 3cm, which is compatible with the state-of-the-art methods.

Localization of target tagged objects on the shelf is of great significance in RFID-enabled warehousing scenarios. Compared with the RFID localization systems that use fixed reader antennas or mobile RFID-robot, the portable reader-based methods are much more cost-effective. Hence, this paper focuses on reader-portable RFID localization. However, the existing reader-portable localization systems suffer from the following limitations: (i) reader antenna is required to pass by the target tags. Thus, the tags in the corner can never be located; (ii) many reference tags need to be deployed on the shelf in advance, which considerably increases the manpower; (iii) specialized antenna is required, which limits the promotion potential. To this end, this paper proposes a Waving action-driven RFID Localization (WRL) system, which enables tag localization with a portable camera-augmented reader. In the WRL system, a user only needs to wave the camera-augmented reader before locating the target tags. Specifically, we first use a classical camera pose estimation method named PnP to recover the antenna’s movement trajectory in a pixel coordinate system. Then, WRL constructs a gridded hologram, in which camera data and RFID phase data are jointly used to calculate a probability for each grid. Intuitively, the higher probability a grid has, the more possible the target tag lies in the corresponding grid. Based on this idea, WRL calculates the target tag’s location on the shelf. We use the Commercial-Off-The-Shelf (COTS) RFID and camera devices to implement the WRL system. Extensive experiments have been conducted, and the results demonstrate that the mean localization error of WRL is less than 20cm with a confidence of about 95%.

Major manufacturers and retailers are increasingly using RFID systems in supply-chain scenarios, where theft of goods during transport typically causes significant economic losses for the consumer. Recent sample-based authentication methods attempt to use a small set of random sample tags to authenticate the integrity of the entire tag population, which significantly reduces the authentication time at the expense of slightly reduced reliability. The problem is that it still incurs extensive initialization overhead when writing the authentication information to all of the tags. This paper presents KTAuth, a lightweight integrity authentication approach to efficiently and reliably detect missing tags and counterfeit tags caused by stolen attacks. The competitive advantage of KTAuth is that it only requires writing the authentication information to a small set of deterministic key tags, offering a significant reduction in initialization costs. In addition, KTAuth strictly follows the C1G2 specifications and thus can be deployed on Commercial-Off-The-Shelf RFID systems. Furthermore, KTAuth proposes a novel authentication chain mechanism to verify the integrity of tags exclusively based on data stored on them. To evaluate the feasibility and deployability of KTAuth, we implemented a small-scale prototype system using mainstream RFID devices. Using the parameters achieved from the real experiments, we also conducted extensive simulations to evaluate the performance of KTAuth in large-scale RFID systems.

Mobile RFID robots have been increasingly used in warehousing and intelligent manufacturing scenarios to pinpoint the locations of tagged objects. The accuracy of state-of-the-art RFID robot localization systems depends much on the stability of robot moving speed. However, in reality this assumption can hardly be guaranteed because a Commercial-Off-The-Shelf (COTS) robot typically has an inconsistent moving speed, and a small speed inconsistency will cause a large localization error. To this end, we propose a Speed Inconsistency-Immune approach to mobile RFID robot Localization (SILoc) system, which can accurately locate RFID tagged targets when the robot moving speed varies or is even unknown. SILoc employs multiple antennas fixed on the mobile robot to collect the phase data of target tags. We propose an optimized unwrapping method to maximize the use of the phase data, and a lightweight algorithm to calculate the locations in both 2D and 3D spaces based on the unwrapped phase profile. By utilizing the characteristics of tag-antenna distance and combining the phase data from multiple antennas, SILoc can effectively eliminate the side effects of moving speed inconsistency. Extensive experimental results demonstrate that SILoc can achieve a centimeter-level localization accuracy in the scenario with an inconsistent or unknown robot moving speed.

Mobile Edge Computing (MEC) is a promising cloud-network convergence paradigm which provides computational resources close to end devices at the network edge. There exist multiple Edge Infrastructure Providers (EIPs) in MEC which independently manage edges and provide services to customers. Due to the exponentially increasing data generated by end devices, it is almost impossible for a single EIP to accommodate offloaded data. Moreover, when considering that multiple EIPs provide services through federation, an urgent challenge is how to ensure the sustainability of federation. Most of the existing work improves the service provision capabilities of MEC by optimizing service placement without considering the existence of multiple EIPs. In this paper, we design the horizontal collaboration of edge federation, which integrates all edges of all EIPs. First, we model the service placement problem as a programming problem, towards the goal of maximizing social welfare. Then, we propose two dynamic pricing methods for EIPs to determine typical price for customers and insourcing price for other EIPs. The evaluation results based on two real-world data sets demonstrate that our proposed service placement model can increase the total gain of EIPs by up to 24.5% with a decrease of 35.5% in total delay.

Over the last decade, traffic engineering (TE) has always been a research hotspot in data center networks. For routing flows efficiently and practically, existing TE schemes explore experience-driven heuristics or machine learning (ML) techniques to predict/identify network flows’ size information. However, these TE schemes have significant limitations: they either identify the flow size information too late or are unaware of the ML models’ prediction errors. In this paper, we present DarkTE, a novel TE solution that can learn to predict flow size information timely for achieving better routing performance while being robust to the prediction errors. At its heart, DarkTE employs an ensemble learning technique (i.e., random forest) to classify flows into mice and elephant flows with high accuracy. It then leverages a confidence-based rate allocation and path selection scheme to mitigate the occasional classification errors. Large-scale simulations demonstrate that DarkTE classifies flows within hundreds of microseconds, and the classification accuracy is at least 86.4% over three different realistic workloads. Further, DarkTE completes flows 2.94 times faster on average and makes more links to experience over 90% bandwidth utilization than the Hedera solution.

Millimeter-wave (mmWave) communication is a promising technology for future vehicular networks, where plenty of self-driving vehicles transmit a great amount of sensing data to the edge-cloud platform for real-time processing to ensure driving safety. While beam selection has been widely investigated in single mmWave base station (mmBS) scenario to maximize the throughput between the vehicle and the mmBS, it is still quite challenging to perform optimal beam selection in mmWave vehicular networks with multiple mmBSs. On the one hand, performing beam selection at a central controller with global information of the networks is infeasible due to the exponentially increased complexity. On the other hand, a distributed solution may suffer from the interference between overlapping beams among mmBSs which leads to severe throughput degradation. To fill this gap, in this paper, we propose a Multi-Agent Reinforcement Learning based cooperative Beam Selection (MARL-BS) algorithm for mmWave vehicular networks. Specifically, we model the beam selection problem as a multi-agent multi-armed bandit problem and then adopt Q-learning to learn how to coordinate the beam selection decisions. In the proposed approach, each mmBS acts as an agent and learns the Q-values of its own actions in conjunction with those of the other mmBSs. We further propose a modified combinatorial upper confidence bound (CUCB) algorithm to take advantage of exploring and exploiting all the candidate beams to avoid falling into local optimum. Finally, our simulations validate the proposed MARL-BS algorithm and confirm its higher performance compared with the other benchmark algorithms.

Recognition of human-object interactions is practically important in various human-centric sensing scenarios such as smart supermarket, factory, and home. This paper proposes an RF-Camera system by fusing RFID and Computer Vision (CV) techniques, which is the first work to recognize the human gestural interactions with physical objects in multi-subject and multi-object scenarios. In RF-Camera, we first propose a dimension reduction method to transform the subject's 3D hand trajectory captured by depth camera to a 2D image, using which the subject's gesture can be recognized. We also propose a method to extract the facial image of target subject from an image that may contain irrelevant subjects, thereby further recognizing his/her identity. Finally, we model the physical movements of the held object's tag and further predict the tag phase data, by comparing which with real phase data of each tag human-object matching can be discovered. When implementing RF-Camera, three technical challenges need to be addressed. (i) To remove noisy data corresponding to irrelevant actions from raw sensing data, we propose a state transition diagram to determine the boundary of effective data. (ii) To predict phase data of the held target tag with unknown hand-tag offset, we quantify target tag trajectory by adding a variable hand-tag vector to captured hand trajectory. (iii) To ensure high reading rates of target tags in tag-dense scenarios, we propose a CV-assisted RFID scheduling method, in which analytics on CV data can help schedule RFID readings. We conduct extensive experiments to evaluate the performance of RF-Camera. Experimental results demonstrate that RF-Camera can recognize the gestural actions, human identity and human-object matching with an average accuracy higher than 90% in most cases.

In the tile-based 360-degree video streaming, it is essential to predict future viewport and to allocate higher bitrates to tiles inside the predicted viewport to optimize the Quality of Experience (QoE) of the users. However, the majority of existing work focuses on short-term viewport prediction, which is prone to rebuffering in dynamic network conditions. On the other hand, the recently developed on-policy Deep Reinforcement Learning (DRL)-based bitrate allocation approaches suffer from poor sample efficiency. To address these issues, in this paper we present a tile-based adaptive 360-degree video streaming system, named LS360, which consists of long-term viewport prediction and adaptive bitrate allocation. First, we propose a Long Short-Term Memory (LSTM)-based viewport prediction model to make use of the heatmap feature from all users’ previous movement information and the target user’s fixation movement feature to improve prediction accuracy. Next, we employ the off-policy Soft Actor-Critic (SAC) algorithm to make optimal tile bitrate allocation decisions by taking the predicted long-term viewport, playback buffer, and bandwidth-related information into account. Experiments on real-world datasets demonstrate that LS360 outperforms state-of-the-art streaming algorithms in terms of long-term viewport prediction accuracy and QoE under different bandwidth conditions.

Interference among distributed cloud applications can be classified into three types: full, partial and zero. While prior research merely focused on full interference, the partial interference that occurs at parts of applications is far more common yet still lacks in-depth study. Serverless computing that structures applications into small-sized, short-lived functions further exacerbate partial interference. We characterize the features of partial interference in serverless as exhibiting high volatility, spatial-temporal variation, and propagation. Given these observations, we propose an incremental learning predictor, named Gsight, which can achieve high precision by harnessing the spatial-temporal overlap codes and profiles of functions via an end-to-end call path. Experimental results show that Gsight can achieve an average error of 1.71%. Its convergence speed is at least 3X faster than that in a serverful system. A scheduling case study shows that the proposed method can improve function density by ≥ 18.79% while guaranteeing the quality of service (QoS).

Driven by the high profit, Portable Executable (PE) malware has been consistently evolving in terms of both volume and sophistication. PE malware family classification has gained great attention and a large number of approaches have been proposed. With the rapid development of machine learning techniques and the exciting results they achieved on various tasks, machine learning algorithms have also gained popularity in the PE malware family classification task. Three mainstream approaches that use learning based algorithms, as categorized by the input format the methods take, are image-based, binary-based and disassembly-based approaches. Although a large number of approaches are published, there is no consistent comparisons on those approaches, especially from the practical industry adoption perspective. Moreover, there is no comparison in the scenario of concept drift, which is a fact for the malware classification task due to the fast evolving nature of malware. In this work, we conduct a thorough empirical study on learning-based PE malware classification approaches on 4 different datasets and consistent experiment settings. Based on the experiment results and an interview with our industry partners, we find that (1) there is no individual class of methods that significantly outperforms the others; (2) All classes of methods show performance degradation on concept drift (by an average F1-score of 32.23%); and (3) the prediction time and high memory consumption hinder existing approaches from being adopted for industry usage.

Hashing methods have been widely used in Approximate Nearest Neighbor (ANN) search for big data due to low storage requirements and high search efficiency. These methods usually map the ANN search for big data into the k -Nearest Neighbor ( k NN) search problem in Hamming space. However, Hamming distance calculation ignores the bit-level distinction, leading to confusing ranking. In order to further increase search accuracy, various bit-level weights have been proposed to rank hash codes in weighted Hamming space. Nevertheless, existing ranking methods in weighted Hamming space are almost based on exhaustive linear scan, which is time consuming and not suitable for large datasets. Although Multi-Index hashing that is a sub-linear search method has been proposed, it relies on Hamming distance rather than weighted Hamming distance. To address this issue, we propose an exact k NN search approach with Multiple Tables in Weighted Hamming space named WHMT, in which the distribution of bit-level weights is incorporated into the multi-index building. By WHMT, we can get the optimal candidate set for exact k NN search in weighted Hamming space without exhaustive linear scan. Experimental results show that WHMT can achieve dramatic speedup up to 69.8 times over linear scan baseline without losing accuracy in weighted Hamming space.

This paper studies the problem of tag localization using RFID-augmented robots, which is practically important for promising warehousing applications, e.g., automatic item fetching and misplacement detection. Existing RFID localization systems suffer from one or more of following limitations: requiring specialized devices; only 2D localization is enabled; having blind zone for mobile localization; low scalability. In this paper, we use Commercial Off-The-Shelf (COTS) robot and RFID devices to implement a Mobile RF-robot Localization (MRL) system. Specifically, when the RFID-augmented robot moves along the straight aisle in a warehouse, the reader keeps reading the target tag via two vertically deployed antennas ( Z1 and Z2) and returns the tag phase data with timestamps to the server. We take three points in the phase profile of antenna Z1 and leverage the spatial and temporal changes inherent in this phase triad to construct an equation set. By solving it, we achieve the location of target tag relative to the trajectory of antenna Z1. Based on different phase triads, we can have candidate locations of the target tag with different accuracy. Then, we propose theoretical analysis to quantify the deviation of each localization result. A fine-grained localization result can be achieved by assigning larger weights to the localization results with smaller deviations. Similarly, we can also calculate the relative location of target tag with respect to the trajectory of antenna Z2. Leveraging the geometric relationships among target tag and antenna trajectories, we eventually calculate the location of target tag in 3D space. We perform various experiments to evaluate the performance of the MRL system and results show that the proposed MRL system can achieve high accuracy in both 2D and 3D localization.

The last few years have witnessed the rise of the big data era in which approximate nearest neighbor search is a fundamental problem in many applications, such as large-scale image retrieval. Recently, many research results have demonstrated that hashing can achieve promising performance due to its appealing storage and search efficiency. Since complex optimization problems for loss functions are difficult to solve, most hashing methods decompose the hash code learning problem into two steps: projection and quantization. In the quantization step, binary codes are widely used because ranking them by the Hamming distance is very efficient. However, the massive information loss produced by the quantization step should be reduced in applications where high search accuracy is required, such as in image retrieval. Since many two-step hashing methods produce uneven projected dimensions in the projection step, in this paper, we propose a novel dimension analysis-based quantization (DAQ) on two-step hashing methods for image retrieval. We first perform an importance analysis of the projected dimensions and select a subset of them that are more informative than others, and then we divide the selected projected dimensions into several regions with our quantizer. Every region is quantized with its corresponding codebook. Finally, the similarity between two hash codes is estimated by the Manhattan distance between their corresponding codebooks, which is also efficient. We conduct experiments on three public benchmarks containing up to one million descriptors and show that the proposed DAQ method consistently leads to significant accuracy improvements over state-of-the-art quantization methods.

Detecting abnormal service performance is significant for Internet-based service management and operation. Recent advances in anomaly detection methods prefer unsupervised learning algorithms since they can work without manually labelled data. However, existing unsupervised methods converge into suboptimal solutions due to their heuristic-based objectives. Moreover, they frequently rely on the strong assumption that noise follows a Gaussian distribution, and their detection accuracy is also highly sensitive to threshold settings. To detect anomalies precisely and robustly, we presentAdran, an unsupervised anomaly detection model that introduces adversarial learning into a reconstructive model, generating a reconstructive adversarial network with an anomaly detection-based training objective. It tolerates non-Gaussian noise by activating the discriminator with a non-smooth function. Our experimental results demonstrate thatAdranachieves an improvement of$\geq 32\%$over the state-of-the-art methods in terms ofF-score. Moreover, the robustness analysis demonstrates that it is reasonably easy and straightforward to set an appropriate threshold usingAdran.

Data-parallel applications generate a mix of coflows with and without deadlines. Deadline coflows are mission-critical and must be completed within deadlines, while the non-deadline coflows desire to be completed as soon as possible. Scheduling such mix-coflows is an important problem in modern datacenters. However, existing solutions only focus on one of the two types of coflows: they either solely concentrate on meeting the deadlines of deadline-aware coflows or reducing the coflow completion times (CCTs) of non-deadline coflows. In this article, we study the problem of optimizing deadline and non-deadline coflows simultaneously. To this end, we present a new optimization framework,mixCoflow, to schedule deadline coflows to minimize and balance their bandwidth footprint, such that non-deadline coflows can be scheduled as early as possible. Specifically, we develop the mathematical model and formulate the scheduling problem for deadline coflows as a lexicographical min-max integer linear programming (ILP) problem. Through rigorous theoretical analysis, this ILP problem has been proved to be equivalent to a linear programming (LP) problem that can be solved with standard LP solvers. By solving this LP,mixCoflowis able to balance the bandwidth footprint of deadline coflows while guaranteeing their deadlines. As a result, non-deadline coflows can be scheduled as soon as possible whenever they arrive. To demonstrate the effectiveness of our work, we have conducted extensive simulations based on a widely used Facebook data trace. The simulation results verify thatmixCoflowcan achieve significant improvement on the average CCT of non-deadline coflows, at no expense of increasing the deadline miss rates of deadline coflows, when compared to the state-of-art solutions.

Low latency stream processing on large clusters consisting of hundreds to thousands of servers is an increasingly important challenge. A crucial barrier to tackling this challenge is stragglers, i.e., tasks that are significantly straggling behind others in processing the stream data. However, prior straggler mitigation solutions have significant limitations. They balance streaming workloads among tasks but may incur imbalanced backlogs when the workloads exhibit variance, causing stragglers as well. Fortunately, we observe that carefully scheduling the outgoing tuples of different tasks can yield benefits for balancing backlogs, and thus avoids stragglers. To this end, we present Hone, a tuple scheduler that aims to minimize the maximum queue backlog of all tasks over time. Hone leverages an online Largest-Backlog-First (LBF) algorithm with a provable good competitive ratio to perform efficient tuple scheduling. We have implemented Hone based on Apache Storm and evaluated it extensively via both simulations and testbed experiments. Our results show that under the same workload balancing strategy-shuffle grouping, Hone outperforms the original Storm significantly, with the end-to-end tuple processing latency reduced by 78.7 percent on average.

Cloud service providers improve resource utilization by co-locating latency-critical (LC) workloads with best-effort batch (BE) jobs in datacenters. However, they usually treat an LC workload as a whole when allocating resources to BE jobs and neglect the different features of components of an LC workload. This kind of coarse-grained co-location method leaves a significant room for improvement in resource utilization.

Detecting anomaly in images is challenging due to the high dimension nature of image data. While the previous learning-based anomaly detection approaches can detect a particular type of anomaly precisely, they often fail in detecting multiple types of abnormal samples simultaneously.We identify the two specific types of anomalies that can be precisely detected by either compress-based or reconstruction-based anomaly detection approaches, named global anomaly and local anomaly. We then propose Glad, an anomaly detector that can precisely detect both of them at the same time. Glad adopts a joint approach combining the density estimation and auto-encoder. Firstly, it designs a multimodal density estimation model to derive the latent representation probability for identifying the global anomaly. Then, it uses structural similarity to measure the reconstruction loss for characterizing local anomaly. Finally, both anomalies can be diagnosed according to the joint density of latent representation and reconstruction loss. Experimental results on public benchmark datasets demonstrate that Glad outperforms the state-of-the-art methods significantly.

High-resolution (HR) medical magnetic resonance images or computer tomography (CT) images can provide clearer anatomical details of human body, which facilitates early diagnosis of the diseases. However, due to the limitations of imaging environments, imaging systems and human factors, it is not easy to obtain clear HR medical images. In this paper, we propose a novel medical image super-resolution (SR) network, namely granular multi-scale network (GMSN), in the non-subsampled contourlet transform (NSCT) domain. GMSN mainly consists of a series of cascaded Res2Net blocks to exploit the multi-scale potential features at a granular level from medical images and increase the range of receptive fields for each network layer. The architecture possesses stronger multi-scale feature extraction ability, while maintaining a low computational load. In addition, most previous methods predict the HR images in the spatial domain, producing over-smoothed outputs while losing texture details. Thus, we formulate the medical image SR problem as the prediction of NSCT coefficients, which is able to further GMSN preserve richer structure details than that in spatial domain. The experimental results on our constructed medical image database show that our proposed method is capable of obtaining higher PSNR/SSIM values and preserving global topological structure and local texture detail better than other state-of-the-art methods.

Accurate and quick failure localization is critical for automatic network troubleshooting. While it is particularly difficult to solve the problem in the traditional network due to the uncertain routing, Software Defined Networking (SDN) enables the deterministic routing for packet transmission through the traffic engineering algorithm in the centralized controller.

Geographically distributed cloud is a promising technique to achieve high performance for service providers. For inter-datacenter transfers, deadline guarantee and fairness are the two most important requirements. On the one hand, to ensure more transfers finish before their deadlines, preemptive scheduling policies are widely used, leading to the transfer starvation problem and is hence unfair. On the other hand, to ensure fairness, inter-datacenter bandwidth is fairly shared among transfers with per-flow bandwidth allocation, which leads to deadline missing problem. A mechanism that achieves these two seemingly conflicting objectives simultaneously is still missing. In this paper, we propose TINA to schedule network transfers fairly while providing deadline guarantees. TINA allows each transfer to compete freely with each other for bandwidth. More specifically, each transfer is assigned a probability to indicate whether to transmit or not. We formulate the competition among the transfers as an El Farol game while keeping the traffic load under a threshold to avoid congestion. We then prove that the Nash Equilibrium is the optimal strategy and propose a light-weight algorithm to derive it. Finally, both simulations and testbed experiments results show that TINA achieves superior performance than state-of-art methods in terms of fairness and deadline guarantee rate.

Individual activity recognition is crucial for Human-Computer Interaction (HCI) applications, especially in multi-person scenarios. Current approaches, based on wearable sensors or wireless signals (e.g., WiFi and RFID), however, are often focused on single person scenario only, due to the limitation of existing wireless sensing technologies. In order to address the issue, we design a DEeper Exercise Monitoring system, called DEEM, in which we introduce computer vision techniques to facilitate RFID devices to provide exercise estimation support, as well as identifying the users and the objects users hold. We implement this design with COTS Kinect camera and RFID devices in a smart gym application. To the best of our knowledge, it is the first system for estimating multiple people behavior in a complicated gym environment. We conduct extensive experiments to evaluate the performance of the DEEM system. The experimental results show that the matching accuracy can reach 95%, and the exercise estimation accuracy can reach 94% on average.

Internet of Things (IoT) includes numerous sensing nodes that constitute a large scale-free network. Optimizing the network topology for increased resistance against malicious attacks is an NP-hard problem. Heuristic algorithms, particularly genetic algorithms, can effectively cope with such problems. However, conventional genetic algorithms are prone to falling into premature convergence owing to the lack of global search ability caused by the loss of population diversity during evolution. Although this can be alleviated by increasing population size, additional computational overhead will be incurred. Moreover, after crossover and mutation operations, individual changes in the population are mixed, and loss of optimal individuals may occur, which will slow down the evolution of the population. Therefore, we combine the population state with the evolutionary process and propose an Adaptive Robustness Evolution Algorithm (AREA) with self-competition for scale-free IoT topologies. In AREA, the crossover and mutation operations are dynamically adjusted according to population diversity to ensure global search ability. Moreover, a self-competitive mechanism is used to ensure convergence. The simulation results demonstrate that AREA is more effective in improving the robustness of scale-free IoT networks than several existing methods.

This paper studies the practically important problem of data collection for sensor-augmented RFID systems. However, existing RFID data collection protocols suffer from two common limitations: execution time is naturally in proportion to the number of tags, thus they cannot satisfy time-stringent application scenarios; none of them is complaint with the C1G2 standard, thus they cannot be implemented using Commercial-Off-The-Shelf (COTS) RFID tags. To overcome these two limitations, this paper proposes the Geographical correlation-based RF-data Collection (GRC) protocol. GRC is fast because it is able to approximately capture the sensing data of all tags by only actually gathering data from a small set of sampled tags. This is based on the observation from the real-world data set that sensing data has a strong geographical correlation, i.e., data gathered from nearby RFID tags has similar values. In GRC, we use a greedy approach to find the minimum sampling tag set to cover the whole monitoring region such that each un-sampled tag has at least one sampled tag nearby. Then, RFID reader runs the Framed Slotted Aloha (FSA) protocol specified in C1G2 standard to collect sensing data from the sampled tags. For each un-sampled tag, we approximate its sensing data by calculating weight-average of the data collected from its nearby sampled tags, where a faraway sampled tag should be given a small weight, and vice versa. Compared with existing RFID data collection schemes, the advantages of GRC are two-fold: (1) Extensive simulation results demonstrate that the time cost of our GRC scheme is only 1/28~1/3 of the state-of-the-art data collection scheme; (2) GRC is totally complaint with C1G2 standard, thus it can be easily deployed on the COTS RFID tags.

Tag inventory is one of the most fundamental tasks for RFID systems. However, the Framed Slotted Aloha (FSA) protocol specified in the C1G2 standard is of low time-efficiency, because it needs to collect all tags in the system. To improve time-efficiency, research communities proposed a batch of sampling-based approaches, in which the reader only needs to collect a small set of sampled tags instead of all. Although time-efficiency has been improved, existing sampling-based approaches still have two common limitations. First, all tags in the system are assumed to have the same sampling probability. It is unfair that tags attached to differential items (e.g., different values) have the same chance to be sampled and collected. Second, all existing sampling-based approaches stay in theory level and cannot be deployed on Commercial Off-The-Shelf (COTS) RFID devices, because the C1G2 standard does not support the sampling function at all. To deal with the above two limitations, this paper studies the new problem of differential tag sampling-letting each RFID tag be identified with a given sampling probability. In this paper, we use the COTS RFID devices including Impinj Speedway R420 reader and Monza 4QT tags to implement the Differential Tag Sampling (DTS) operation. Then, we apply probabilistic analytics on the collected tag data to address some practically important problems such as Multi-category Tag Cardinality Estimation (MTCE), and Value-based Missing Tag Detection (VMTD). Although the analytics results are not 100 percent accurate, the deviation in the results can be controlled below a small threshold and DTS can significantly improve the time-efficiency. DTS can be easily deployed on the COTS RFID systems, because it is totally compliant with the C1G2 standard. Extensive experiments demonstrate that DTS is able to let each tag take the given sampling probability to be sampled and identified. Moreover, the proposed DTS protocol can significantly reduce the execution time of MTCE and VMTD by nearly 70 percent than the FSA protocol.

Over the last decade, we have witnessed growing data volumes generated and stored across geographically distributed datacenters. Processing such geo-distributed datasets may suffer from significant slowdown as the underlying network flows have to go through the inter-datacenter networks with relatively low and highly heterogeneous available link bandwidth. Thus, optimizing the transmissions of inter-datacenter flows, especially coflows that capture application-level semantics, is important for improving the communication performance of such geo-distributed applications. However, prior solutions on coflow scheduling have significant limitations: they schedule coflows with already-fixed endpoints of flows, making them insufficient to optimize the coflow completion time (CCT). In this article, we focus on the problem of jointly considering endpoint placement and coflow scheduling to minimize the average CCT of coflows across geo-distributed datacenters. To solve this problem without any prior knowledge of coflow arrivals, we present a coflow-aware optimization framework called SmartCoflow. In SmartCoflow, we first apply an approximate algorithm to obtain the endpoint placement and scheduling decisions for a single coflow. Based on the single-coflow solution, we then develop an efficient online algorithm to handle the dynamically arrived coflows. Through rigorous theoretical analysis, we prove that SmartCoflow has a non-trivial competitive ratio. We also extend SmartCoflow to incorporate various design choices or requirements of applications and operators, such as enforcing an inter-datacenter bandwidth usage budget and considering coflow deadline. Through experimental results from testbed implementation and trace-driven simulations, we demonstrate that SmartCoflow can reduce the average CCT, lower bandwidth usage, and improve coflow deadline meet rate, when compared to the state-of-the-art scheduling-only method.

As more and more applications are deployed in data centers, they rely on high performance data center networks (DCNs) to meet users' increasing quality of the experience (QoE) requirements. Hence, minimizing the average flow completion time (FCT) has been one of the most important goals for DCNs. However, existing traffic scheduling methods assume either prior knowledge of flows (i.e., sizes and deadlines) or symmetric topologies (i.e., Fat-Tree or Bcube). In practice, it is difficult to obtain the information of flows. Moreover, even with symmetric topology design, the DCNs will become asymmetric due to the inevitable link failure and congestion. In this case, it is a great challenge to minimize the average FCT in DCNs. In this paper, we propose a flowlet based information-agnostic traffic scheduling mechanism. The key idea of our method is leveraging multiple priority queues in switches to demote the priority of flows dynamically on the flowlet level. More specifically, the priority of a flow will be demoted according to the number of flowlets it has sent, which follows the shortest job first discipline. We formulate the average FCT minimization problem as a nonlinear Sum-of-Ratios problem and design two heuristic methods to derive the sub-optimal demotion thresholds. Experiment results show that our method can reduce the average FCT by up to 15.35% with a realistic workload, as compared to the state-of-the-art traffic scheduling methods.

In mobile edge computing (MEC) systems with energy harvesting, the mobile devices are empowered with the energy that harvested from renewable energy sources. On the other hand, mobile devices can offload their computation-intensive tasks to the MEC server to further save energy and reduce the task execution latency. However, the energy harvested is unstable and the mobile devices have to make sure that the energy should not be run out. Moreover, the wireless channel condition between the mobile device and the MEC server is dynamically changing, leading to unstable communication delay. Considering the energy constraints and unstable communication delay, the benefit of computation offloading is limited. In this paper, we investigate D2D-assisted computation offloading for mobile edge computing systems with energy harvesting. In our method, the mobile device is allowed to offload its tasks to the MEC server with the help of its neighbor node. More Specifically, the neighbor node acts as a relay to help the mobile device to communicate with the MEC server. Our goal is to minimize the average task execution time by selecting an optimal execution strategy for each task, i.e., whether to execute the task locally, or offload it to the MEC server directly, or offload it to the MEC server with the help of the most suitable neighbor node, or just to drop it. We propose a low-complexity online algorithm, which stem from Lyapunov Optimization-based Dynamic Computation Offloading (LODCO) algorithm, to solve this problem. Extensive simulations verified the effectiveness of the proposed algorithm, where the average task execution time is reduced around 50% as compared to that of the original LODCO algorithm.

In software-defined network, the use of distributed controllers to control forwarding devices has been proposed to solve the issues of scalability and load balance. However, the forwarding devices are statically assigned to the controllers in these distributed systems, which can overload some controllers while others are underutilized. In this paper, we propose an architecture named ASLB (active controller selection load balance), which proactively selects appropriate controllers for load balancing and minimize packet processing delays. We also present a novel active controller selection algorithm (ACS) for ASLB that efficiently schedules traffic from the switch to the controller and designs an intermediate coordinator for actively selecting a controller to serve a request. We built a system and evaluated it on a physical platform. The results show that ASLB is much better than the static allocation scheme in terms of minimizing latency, bandwidth utilization, and throughput.

Hashing methods have attracted much attention due to their superior time and storage properties for image retrieval. To learn similarity-preserving hash function, most existing methods are designed for the centralized setting. However, the current data storage systems are distributed to increase scalability. Obviously, it is infeasible to aggregate all the data into a fusion center because of the prohibitively expensive communication and computation overhead. Motivated by this, some methods are proposed to achieve hashing for distributed data. However, these methods mostly focus on extending one specific hashing to a distributed model without considering the generality. In this letter, we propose a novel general distributed hash learning model, which can be viewed as an effective distributed model of most hashing methods. The proposed model can achieve up to 15.2% accuracy gains over state-of-the-art distributed hashing methods, while the communication cost is independent on the data size.

The emerging deployment of large-scale cloud applications incurs significant inter-datacenter traffic, which makes the scarce wide-area bandwidth across data centers become the performance bottleneck. To achieve the desirable network performance, bandwidth guarantee should be provided for the resulting inter-datacenter traffic. However, the existing bandwidth allocation methods mainly focus on intra-datacenter traffic, and cannot achieve the cost-minimizing bandwidth guarantee for inter-datacenter traffic. In this paper, we focus on the bandwidth guarantee problem for inter-datacenter traffic and present a novel bandwidth allocation model. Our model can ensure the bandwidth guarantee, minimize the resulting network cost, and efficiently avoid the potential traffic overload on low cost links. To solve the large-scale optimization problem in our model, we are motivated to develop a distributed algorithm by blending the advantages of alternating direction method of multipliers (ADMM) and the auxiliary variable method. Specifically, we efficiently decompose the optimization problem into many small sub-problems, which are allowed to be processed in a large-scale computing environment, where each server solves a few small sub-problems. We further present a theoretically proved globally, asymptotically stable algorithm to solve these sub-problems. Extensive evaluation results demonstrate that our bandwidth allocation method can effectively realize the bandwidth guarantee for inter-datacenter traffic with reduced network cost and outperforms the prior method PS-L. In particular, the total network cost is reduced by 59.57 percent on average.

An industrial mobile network is crucial for industrial production in the Internet of Things. It guarantees the normal function of machines and the normalization of industrial production. However, this characteristic can be utilized by spammers to attack others and influence industrial production. Users who only share spams, such as links to viruses and advertisements, are called spammers. With the growth of mobile network membership, spammers have organized into groups for the purpose of benefit maximization, which has caused confusion and heavy losses to industrial production. It is difficult to distinguish spammers from normal users owing to the characteristics of multidimensional data. To address this problem, this paper proposes a spammer identification scheme based on Gaussian mixture model (SIGMM) that utilizes machine learning for industrial mobile networks. It provides intelligent identification of spammers without relying on flexible and unreliable relationships. SIGMM combines the presentation of data, where each user node is classified into one class in the construction process of the model. We validate the SIGMM by comparing it with the reality mining algorithm and hybrid fuzzy c-means (FCM) clustering algorithm using a mobile network dataset from a cloud server. Simulation results show that SIGMM outperforms these previous schemes in terms of recall, precision, and time complexity.

Currently, parallel data transmissions in large-scale datacenter networks are becoming increasingly crucial to application performance. Despite fine-grained control by SDN-enabled networks, some transmission errors, such as misconfigurations, will inevitably occur, resulting in high-level forwarding policies that cannot be conformed to at the data plane. Therefore, flow trajectory detection is very important for allowing datacenter network operators to troubleshoot problems and ensure that all traffic flows are running on the correct paths. However, existing solutions detect flow trajectories by recording the entire path of each packet. These methods are prone to imposing significant overheads in terms of both the number of switch entries and the amount of packet header space required. To considerably reduce this overhead, we present FlowTracer, an efficient flow trajectory detection solution, which can sample a path one link at a time instead of recording the entire path. FlowTracer consists of a method of probabilistic packet tagging and a method of trajectory reconstruction. In this paper, we first introduce the method of probabilistic packet tagging, which is performed in OpenFlow-enabled switches with very few switch entries and limited packet header space by means of double VLAN tags. Then, we explore the topological structure of datacenter networks and propose our method of trajectory reconstruction, which is performed at end hosts and achieves rapid convergence. Finally, we evaluate FlowTracer on a 48-ary fat-tree topology. The results show that FlowTracer can detect trajectories quickly while placing far smaller demands on both switch entries and packet header space than state-of-the-art techniques.

In data centers, a lot of cluster computing applications follow the coflow working pattern. That is, a collection of flows between two groups of machines is semantically related. On the other hand, network function virtualization sufficiently improves the performance of data center networks. It however complicates the network environment by introducing many multi-function middleboxes each with multiple resources. Coflows encounter extremely different processing delays under diverse network functions. Prior coflow scheduling schemes are insufficient to guarantee the coflow completion time in the multi-resource environment. In this paper, we propose, model, and analyze the coflow scheduling problem in the multi-resource environment. We present a dedicated method, data rate guarantee for coflow (DRGC), to guarantee the data rate requirements of coflows in this situation. DRGC prioritizes the coflow scheduling sequence, assigns precise data rates for coflows, and deploys a packet scheduling algorithm at middleboxes to guarantee their transmissions. In our experiments, DRGC efficiently guarantees the completion times of coflows and supports 15% more workload, compared with other scheduling schemes.

This paper studies the important sensory data collection problem in the sensor-augmented RFID systems, which is to quickly and accurately collect sensory data from a predefined set of target tags with the coexistence of unexpected tags. The existing RFID data collection schemes suffer from either low time-efficiency due to tag-collisions or serious data corruption issue due to interference of unexpected tags. To overcome these limitations, we propose the hierarchical-hashing data collection (HDC) protocol, which can not only significantly improve the utilization of RFID wireless communication channel by establishing bijective mapping between k target tags and the first k slots in time frame, but also effectively filter out the serious interference of unexpected tags. Although HDC has attractive advantages, the theoretical analysis reveals that the computation cost involved in it is as huge as O(k2k), where k is normally large in practice. By making some modifications to the basic HDC protocol, we propose the multi-framed hierarchical-hashing data collection (MHDC) protocol to effectively reduce the involved computation complexity. Unlike HDC that only issues a single time frame, MHDC uses multiple time frames to collaboratively collect sensory data from the k target tags. It can be understood as that a big computation task is disintegrated into multiple small pieces and then shared by multiple time frames. As a result, the computation cost involved in MHDC is reduced to O(k2n), where n ≪ k is the expected number of target tags that each time frame handles. Theoretical analysis is given to jointly consider the communication cost and computation cost thereby maximizing the overall time-efficiency of MHDC. Extensive simulation results reveal that the proposed MHDC protocol can correctly collect all sensory data and is always about more than 2× faster than the state-of-the-art RFID sensory data collection protocols.

This paper studies the practically important problem of range query for sensor-augmented RFID systems, which is to classify the target tags according to the ranges specified by the user. The existing RFID protocols that seem to address this problem suffer from either low time-efficiency or the information corruption issue. To overcome their limitations, we first propose a basic classification protocol called Range Query (RQ), in which each tag pseudo-randomly chooses a slot from the time frame and uses the ON-OFF Keying modulation to reply its range identifier. Then, RQ employs a collaborative decoding method to extract the tag range information from singleton and even collision slots. The numerical results reveal that the number of queried ranges significantly affects the performance of RQ. To optimize the number of queried ranges, we further propose the Partition&Mergence (PM) approach that consists of two steps, i.e., top-down partitioning and bottom-up merging. Sufficient theoretical analyses are proposed to optimize the involved parameters, thereby minimizing the time cost of RQ+PM or minimizing its energy cost. We can trade off between time cost and energy cost by adjusting the related parameters. The prominent advantages of the RQ+PM protocol over previous protocols are two-fold: (i) it is able to make use of the collision slots, which are treated as useless in previous protocols. Thus, frame utilization can be significantly improved; (ii) it is immune to the interference from unexpected tags, and does not suffer information corruption issue. We use USRP and WISP tags to conduct a set of experiments, which demonstrate the feasibility of RQ+PM. Extensive simulation results reveal that RQ+PM can ensure 100% query accuracy, and reduce the time cost as much as 40% when comparing with the state-of-the-art protocols.

Modern resource management frameworks guarantee low tail latency for long-running services using the resource over-provisioning method, resulting in serious waste of resource and increasing the service costs greatly. To reduce the over-provisioning cost, we present EFRA, an elastic and fine-grained resource allocator that enables much more efficient resource provisioning while guaranteeing the tail latency Service Level Objective (SLO). EFRA achieves this through the cooperation of three key components running on a containerized platform: The period detector identifies the period features of the workload through a convolution-based time series analysis. The resource reservation component estimates the just-right amount of resources based on the period analysis through a top-K based collaborative filtering approach. The online reprovisioning component dynamically adjusts the resources for further enforcing the tail latency SLO. Testbed experiments show that EFRA is able to increase the average resource utilization to 43%, and save up to 66% resources while guaranteeing the same tail latency objective.

Coflow scheduling is crucial to improve the communication performance of data-parallel jobs, especially when these jobs running in the inter-datacenter networks with limited and heterogeneous link bandwidth. However, prior solutions on coflow scheduling assume the endpoints of flows in a coflow to be fixed, making them insufficient to optimize the coflow completion time (CCT). In this paper, we focus on the problem of jointly considering endpoint placement and coflow scheduling to minimize the average CCT of coflows across geo-distributed datacenters. We first develop the mathematical model and formulate a mixed integer linear programming (MILP) problem to characterize the intertwined relationship between endpoint placement and coflow scheduling, and reveal their impact on the average CCT. Then, we present SmartCoflow, a coflow-aware optimization framework, to solve the MILP problem without any prior knowledge of coflow arrivals. In SmartCoflow, we first apply an approximate algorithm to obtain the endpoint placement and scheduling decisions for a single coflow. Based on the single-coflow solution, we then develop an efficient online algorithm to handle the dynamically arrived coflows. To validate the efficiency and practical feasibility of SmartCoflow, we implement it as a real-world coflow scheduler based on the Varys open-source framework. Through experimental results from both a small-scale testbed implementation and large-scale simulations, we demonstrate that SmartCoflow can achieve significant improvement on the average CCT, when compared to the state-of-the-art scheduling-only method.

This paper takes the first step in studying the problem of range query for sensor-augmented RFID systems, which is to classify the target tags according to the range of tag information. The related schemes that seem to address this problem suffer from either low time-efficiency or the information corruption issue. To overcome their limitations, we first propose a basic classification protocol called Range Query (RQ), in which each tag pseudo-randomly chooses a slot from the time frame and uses the ON-OFF Keying modulation to reply its range identifier. Then, RQ employs a collaborative decoding method to extract the tag information range from even collision slots. The numerical results reveal that the number of queried ranges significantly affects the performance of RQ. To optimize the number of queried ranges, we further propose the Partition&Mergence (PM) approach that consists of two steps, i.e., top-down partitioning and bottom-up merging. Sufficient theoretical analyses are proposed to optimize the involved parameters, thereby minimizing the time cost of RQ+PM. The prominent advantages of RQ+PM over previous schemes are two-fold: (i) it is able to make use of the collision slots, which are treated as useless in the previous schemes; (ii) it is immune to the interference from unexpected tags. We use the USRP and WISP tags to conduct a set of experiments, which demonstrate the feasibility of RQ+PM. Moreover, extensive simulation results reveal that RQ+PM can ensure 100% query accuracy, meanwhile reducing the time cost as much as 40% comparing with the existing schemes.

Traditional schemes achieving load balancing in asymmetric topology, which need to maintain global or local congestion information, turn out to be complicated to implement. One recent research has verified that flowlet switching is more simple and efficient to achieve adaptive load balancing in asymmetric topology. Nevertheless, one tricky problem lies in determining the flowlet timeout value, δ. Setting it too small would risk reordering issue while setting it too large would reduce flowlet opportunities. In this paper, by formulating the timeout setting problem with a stationary distribution of Markov chain, we give a theoretical reference for setting an appropriate timeout value in flowlet switching based load balancing scheme. Then, we implement a flowlet switching based load balancing scheme, called EasyLB, by extending OpenFlow protocol. Experiment results show that, by setting timeout value following the preceding theoretical reference, EasyLB is adaptive to asymmetric topology and achieves fast convergence of load balancing after link failures.

Data-parallel applications generate a mix of coflows with and without deadlines. Deadline coflows are mission-critical and must be completed within deadlines, while non-deadline coflows desire to be completed as soon as possible. Scheduling such mix-coflows is an important problem in modern datacenters. However, existing solutions only focus on one of the two types of coflows: they either solely focus on meeting the deadlines of deadline-aware coflows or reducing the coflow completion times (CCTs) of non-deadline coflows. In this paper, we study the problem of optimizing deadline and non-deadline coflows simultaneously. To this end, we present a new optimization framework, mixCoflow, to schedule deadline coflows with the objective of minimizing and balancing their bandwidth footprint, such that non-deadline coflows can be scheduled as early as possible. Specifically, we develop the mathematical model and formulate the scheduling problem for deadline coflows as a lexicographical min-max integer linear programming (ILP) problem. Through rigorous theoretical analysis, this ILP problem has been proved to be equivalent to a linear programming (LP) problem that can be solved with standard LP solvers. By solving this LP, mixCoflow is able to balance the bandwidth footprint of deadline coflows while guaranteeing their deadlines. As a result, non-deadline coflows can be scheduled as soon as possible whenever they arrive. To demonstrate the effectiveness of our work, we have conducted extensive simulations based on a widely used Facebook data trace. The simulation results verify that mixCoflow can achieve significant improvement on the average CCT of non-deadline coflows, at no expense of increasing the deadline miss rates of deadline coflows, when compared to the state-of-art solutions.

Summary Enterprise IT departments must decide whether to build a private cloud themselves or lease space in a cloud instead. Actually, most enterprise cloud systems are heterogeneous simultaneously consisting partly of private cloud and partly of public cloud. An enterprise needs to know the optimal mixing ratio to achieve the lowest cost. However, the optimal ratio is a dynamic value because the popularity of data changes constantly and the data increases rapidly. In this paper, we formulate maximizing the net present value (NPV) of the investment revenue as a dynamic decision‐making problem and propose a model to simplify the decision‐making problem. We use the K‐means algorithm to cluster the devices into groups. In addition, we also provide a solution with the kNN algorithm for locating data. Furthermore, our system provides a cost‐saving solution that requires no additional devices or investment. Subsequently, as the system grows larger, the decision‐making solution can be employed again to determine how to best extend the cloud system at the lowest cost.

Summary Virtual cluster has recently emerged as a common abstraction for cloud applications or tenants to specify and reserve resources. Such virtual cluster brings valuable insights into the cloud elasticity when scaling up or down the number of resources on demand. Unfortunately, for global‐scale applications running on geo‐distributed datacenters, it is always a challenge to scale the virtual cluster. Due to the fact that the inter‐datacenter bandwidth is an expensive and scarce resource, it is increasingly important yet typically hard to achieve cost‐minimizing bandwidth guarantees when scaling. However, existing approaches mainly focus on the scaling within intra‐datacenter networks and cannot be simply extended to the inter‐datacenter scenario. In this paper, we study the problem of scaling up a virtual cluster with consideration of both bandwidth cost minimization and bandwidth guarantees fulfillment targeting inter‐datacenter networks. Specifically, we first propose an efficient algorithm to scale up the virtual cluster without changing its original VM placement. With the observation that such VM placement can hinder the cluster scalability, we further present an optimized algorithm, which exploits VM migration when scaling. Finally, we conduct extensive simulations to demonstrate the effectiveness of our algorithms, in terms of both bandwidth cost and the acceptance rate of scaling requests with bandwidth guarantees.

Nowadays, service function chaining (SFC) becomes more and more widespread and profound to implement flexible and economical virtual network infrastructures for the Internet of Things (IoT). With the benefits of SFC, the IoT service providers can steer massive traffic through a sequence of heterogeneous virtual network function instances based on their business logic. SFC is viewed as an attractive solution for building virtualized IoT-dedicated network. However, the SFC orchestration in IoT is still a challenge problem. Existing work usually focuses on the performance guarantee while ignoring the issue of resource idleness. To meet the sharp increase in IoT traffic amounts and the diversification of IoT traffic requirements, it is necessary to implement the performance and resource aware SFC orchestration system. Motivated by this, we propose a novel linear programming model and an effective approximation optimization algorithm for SFC orchestration, in order to achieve performance guarantee while avoiding resource idleness. Based on the proposed model and algorithm, a new prototype system named performance and resource aware orchestration system of SFC for IoT (PRSFC-IoT) is built upon OpenStack for online SFC orchestration. A large number of simulation experiments show that the PRSFC-IoT outperforms existing solutions for SFC orchestration in IoT.

Cascading failures in the command and control networks (C2 networks) could substantially affect the network invulnerability to some extent. In particular, without considering the characteristics of hierarchy structure, it is quite misleading to employ the existing cascading failure models and effectively analyze the invulnerability of C2 networks. Therefore, a novel cascading failure model for command and control networks with hierarchy structure is proposed in this paper. Firstly, a method of defining the node’s initial load in C2 networks based on hierarchy-degree is proposed. By applying the method, the impact of organizational positions and the degree of the node on its initial load could be highlighted. Secondly, a nonuniform adjustable load redistribution strategy (NALR strategy) is put forward in this paper. More specifically, adjusting the redistribution coefficient could allocate the load from failure nodes to the higher and the same level neighboring nodes according to different proportions. It could be demonstrated by simulation results that the robustness of C2 networks against cascading failures could be dramatically improved by adjusting the initial load adjustment coefficient, the tolerance parameter, and the load redistribution coefficient. And finally, comparisons with other relational models are provided to verify the rationality and effectiveness of the model proposed in this paper. Subsequently, the invulnerability of C2 networks could be enhanced.

The widely used RFID systems are vulnerable to the denial-of-service (DoS) attacks launched by malicious blocker tags. This paper studies how to quickly and completely identify the valid RFID tags that are blocked. The existing work that can seemingly address this problem suffers from either low time-efficiency or serious false positives. This paper proposes a hybrid approach that consists of two complementary component protocols, namelyAloha Filtering(AF) andPoll&Listen(PL).AFis fast but inaccurate, whilePLis accurate but slow. Taking the merit of each protocol, our hybrid approach is to first repeat the fastAFfor multiple rounds to quickly filter out the target tags that are definitely not blocked. Then, on the size-reduced remaining set that just contains a small number of suspicious tags, we invoke the accuratePLto verify the intactness of each suspicious tag with 100 percent confidence. We optimize the round count ofAFthat trades off between the time costs ofAFandPLto minimize the total time ofAF+PL. As required in the optimization process, we need to know the size of the blocked tag set and that of the unknown tag set, which, however, are not known in advance. To estimate these two set sizes, we propose a supplementary protocol calledSimultaneous Estimation of the Blocked tag size and the Unknown tag size(SEBU). The key advantages of our approach over the prior art are four-fold. First, unlike the detection protocol that just discovers the existence of blocking attacks, our approach exactly identifies all the blocked target tags. Second, our approach is compliant with the C1G2 standard, and does not require any modifications to be made to the commercial RFID tags. It only needs to be installed on readers as a software module. Third, our approach does not involve any false positives. Finally, our approach significantly reduces the execution time when compared with the state-of-the-art schemes that can completely identify the blocked tags.

Inefficient bandwidth sharing in a datacenter network, between different application frameworks, e.g., MapReduce and Spark, can lead to inelastic and skewed usage of link bandwidth and increased completion times for the applications. Existing work, however, either solely focuses on managing computation and storage resources or controlling only sending/receiving rate at hosts. In this paper, we present CoMan, a solution that provides global in-network bandwidth management in multiplexed data centers, with two goals: improving bandwidth utilization and reducing application completion time. CoMan first designs a novel abstraction of virtual link groups (VLGs) to establish a shared bandwidth resource pool. Based on this pool, CoMan implements a three-level bandwidth allocation model, which enables elastic bandwidth sharing among computing frameworks as well as guarantees network performance for the applications. CoMan further improves the bandwidth utilization by devising a VLG dependency graph and solves an optimization problem to guide the path selection using a 32-approximation algorithm. We conduct comprehensive trace-driven simulations as well as small-scale testbed experiments to evaluate the performance of CoMan. Extensive simulation results show that CoMan improves the bandwidth utilization and speeds up the application completion time by up to 2.83× and 6.68×, respectively, compared to the ECMP + ElasticSwitch solution. Our implementation also verifies that CoMan can realistically speed up the application completion times by 2.32× on average.

Increasing number of Internet-scale applications, such as video streaming, incur huge amount of wide area traffic. Such traffic over the unreliable Internet without bandwidth guarantee suffers unpredictable network performance. This result, however, is unappealing to the application providers. Fortunately, Internet giants like Google and Microsoft are increasingly deploying their private wide area networks (WANs) to connect their global datacenters. Such high-speed private WANs are reliable, and can provide predictable network performance. In this paper, we propose a new type of service-inter-datacenter network as a service (iDaaS), where traditional application providers can reserve bandwidth from those Internet giants to guarantee their wide area traffic. Specifically, we design a bandwidth trading market among multiple iDaaS providers and application providers, and concentrate on the essentialbandwidth pricingproblem. The involved challenging issue is that the bandwidth price of each iDaaS provider is not only influenced by other iDaaS providers, but also affected by the application providers. To address this issue, we characterize the interaction between iDaaS providers and application providers using a Stackelberg game model, and analyze the existence and uniqueness of the equilibrium. We further present an efficient bandwidth pricing algorithm by blending the advantage of a geometrical Nash bargaining solution and the demand segmentation method. For comparison, we present two bandwidth reservation algorithms, where each iDaaS provider's bandwidth is reserved in a weighted fair manner and a max-min fair manner, respectively. Finally, we conduct comprehensive trace-driven experiments. The evaluation results show that our proposed algorithms not only ensure the revenue of iDaaS providers, but also provide bandwidth guarantee for application providers with lower bandwidth price per unit.

Users’ preferences, and consequently their ratings and reviews to items, change over time. Likewise, characteristics of items are also time-varying. By dividing data into time periods, temporal Recommender Systems (RSs) improve recommendation accuracy by exploring the temporal dynamics in user rating data. However, temporal RSs have to cope with rating sparsity in each time period. Meanwhile, reviews generated by users contain rich information about their preferences, which can be exploited to address rating sparsity and further improve the performance of temporal RSs. In this article, we develop a temporal rating model with topics that jointly mines the temporal dynamics of both user-item ratings and reviews. Studying temporal drifts in reviews helps us understand item rating evolutions and user interest changes over time. Our model also automatically splits the review text in each time period into interim words and intrinsic words. By linking interim words and intrinsic words to short-term and long-term item features, respectively, we jointly mine the temporal changes in user and item latent features together with the associated review text in a single learning stage. Through experiments on 28 real-world datasets collected from Amazon , we show that the rating prediction accuracy of our model significantly outperforms the existing state-of-art RS models. And our model can automatically identify representative interim words in each time period as well as intrinsic words across all time periods. This can be very useful in understanding the time evolution of users’ preferences and items’ characteristics.

To address the challenge arising from mobile users' increasing demands for multimedia services, applying content caching in heterogeneous networks (HetNets) is regarded as an effective way to offload traffic and improve the capacity of mobile networks. In this paper, we aim at designing novel content caching strategies in HetNets to offload the network traffic and support users' requests locally. Specifically, based on some practical network constraints (i.e., patterns of user requests, link capacity and heterogeneous cache sizes) and the derived network topology, we propose a low-complexity and practicable distributed collaborative hierarchical caching framework by decomposing the formulated large-scale optimization problem into a series of simpler subproblems. Trace-based simulation results demonstrate the effectiveness of the proposed framework.

Recently many studies demonstrate that exploiting the Device-to-Device (D2D) content sharing in offline Mobile Social Networks (MSNs) is a promising solution to offload cellular data to local connectivities in proximity to reduce the duplicated cellular transmissions via the backbone network demanded by nearby users and hence to improve users' quality of service (QoS). However, related D2D-based social sharing and offloading proposals are based on either assumptions and theoretical models, or limited data-driven analysis caused by small scale of data sets (e.g. hundreds of MSN users) or single-dimensional feature (e.g. human mobility only), which severely restricts applications in practice. In this paper, we perform the world-first large-scale measurement and analytics on D2D-based content sharing groups from the perspective of social networks via the platform of Xender, a leading global D2D sharing platform in Asia. We analyze the behaviours of about 30 million users with 443 million D2D transmissions of 17 million files in 884 thousand social groups, and unveil the details of social structure properties, network motifs, cascade trees of friendship and propagation, which are helpful for improving the service of social D2D sharing. Finally we discuss challenges and opportunities for improving social D2D sharing services.

The emerging deployment of geographically distributed data centers (DCs) incurs a significant amount of data transfers over the Internet. Such transfers are typically charged by Internet Service Providers (ISPs) with the widely adopted q-th percentile charging model. In such charging model, the time slots with top 100-q percent of data transmission do not affect the total transmission cost, and can be viewed as free. This brings the opportunity to optimize the scheduling of inter-DC transfers to minimize the entire transmission cost. However, very little work has been done to exploit those free time slots for scheduling inter-DC transfers. The crux is that existing work either lacks a mechanism to accumulate traffic to free time slots, or inevitably relies on prior knowledge of traffic arrival patterns. In this paper, we attempt to exploit those free time slots by leveraging diverse time-sensitivities among inter-DC transfers, so as to reduce or even minimize the transmission cost. Specifically, we advocate that a simple principle should be followed: more traffic peaks should be scheduled in free time slots, while less traffic differentiation should be maintained among the remaining time slots. To this end, we take advantage of the Lyapunov optimization techniques to design a pricing-aware control framework. This framework efficiently makes online decisions for inter-DC transfers without requiring a prior knowledge of traffic arrivals. To verify our proposed framework, we conduct small-scale testbed implementation. The results show that our framework can realistically reduce the transmission cost by up to 19.38%.

In the era of global-scale services, analytical queries are performed on datasets that span multiple data centers (DCs). Due to the scarce and expensive inter-DC bandwidth, various methods have been proposed to reduce either the traffic cost or the completion time for those analytics queries. However, current methods make no attempt to maximize the number of successfully served query requests. Moreover, most of them rely on unrealistic assumptions - such as analytical queries are repeated or known in advance. In this paper, we target at characterizing and optimizing the cost-performance tradeoff for geo-distributed data analytics. Our objectives are two-fold: (1) we minimize the inter-DC traffic cost when serving geo-distributed analytics with uncertain query demand, and (2) we maximize the system throughput, in terms of the number of query requests that can be successfully served with guaranteed queuing delay. To achieve these objectives, we take advantage of Lyapunov optimization techniques to design a two-timescale online control framework. Without prior knowledge of future query requests, this framework makes online decisions on input data placement and admission control of query requests. Extensive trace-driven simulation results demonstrate that our framework is capable of reducing inter-DC traffic cost, improving system throughput and guaranteeing a maximum delay for each query request.

The heavy multimedia traffic produced by mobile users poses great challenges for the mobile network operators, especially in the areas with large user densities but limited cellular network capacities (e.g. India). Recently, many studies demonstrate that exploiting the device-to- device(D2D) content sharing in offline Mobile Social Networks is a promising solution to cellular data offloading. However, such approaches are based on either unrealistic assumptions, or limited data analytics caused by small data size (e.g. hundreds of MSN users) or single-dimensional feature (e.g. human mobility only), which severely restricts their applications in practice. To address this issue, this paper performs the first large-scale data measurement and multi-feature analytics of D2D content sharing. Specifically, by using Apache Spark over a 20-server cluster, we analyze the behaviors of 30 million users (with 40 billion D2D transmissions and 16 million content files) of Xender, a leading global D2D sharing platform. Several important features are studied, including performance basics, content properties, location relations, meeting dynamics, and social characteristics. Furthermore, as a proof-of-concept study of our analytics, we also develop a multi-feature learning based framework, which demonstrates the large potentials of predicting and recommending D2D sharing activities using machine learning methods.

To deal with explosively growing demands for multimedia contents from mobile users, content caching in base stations has been considered as an effective solution to improve the network performance by, e.g., offloading network traffic and improving users' Quality of Service (QoS). Moreover, the proactive caching policy in a cache- enabled system needs to be optimized taking into account of content delivery by wireless transmissions. Thus, in this paper, we investigate and propose an efficient resource allocation scheme for min-rate guaranteed content delivery in the downlink multiuser cache-enabled orthogonal frequency division multiple access small cell networks (OFDMA-SCNs). Our aim is to maximize the weighted sum of data rates in an OFDMA-SCN based on the constraints of the caching method, users' QoS, subcarrier reuse and small base stations' transmit power. We employ the alternating direction multiplier method to decompose the formulated complex nonconvex optimization problem into a series of simpler subproblems for which optimal solutions can be easily obtained, and propose corresponding low-complexity methods to solve the subproblems and then the whole problem. Numerical results demonstrate the effectiveness of the proposed resource allocation scheme.

Summary In wireless sensor networks, a lot of applications need the sensed information be transmitted to the sink node within a predefined time threshold. So end‐to‐end delay is an important performance metric in wireless sensor networks. Opportunistic routing protocols have been proposed to reduce the waiting delay. In the duty cycle networks, increasing the duty cycle ratio can also reduce the end‐to‐end delay. However, this method will consume more energy. It is obvious that there exists a trade‐off between delay and energy consumption. So adjusting the duty cycle ratio of each node can investigate this trade‐off. To the best of our knowledge, no existing work takes both of end‐to‐end delay and energy efficiency into consideration in the opportunistic routing networks. In this paper, we want to minimize the whole energy consumption while guaranteeing the expected end‐to‐end delay between the source nodes and the sink node is below the given threshold. To deal with this problem, we propose a dynamic duty cycle scheme which can significantly reduce the energy consumption and guarantee the expected end‐to‐end delay demand in the opportunistic routing network. To be specific, firstly, we formulate a new metric with the wake‐up time slots as the variable to measure the end‐to‐end delay. Secondly, for simplifying the complex problem, we decompose it into a set of single‐hop delay guarantee problems. Feedback controller has been used to solve the problem. We also analyze the influence of the multiple receivers in the same forwarding set. Finally, we conduct extensive simulations to evaluate the performance of the proposed algorithm. The experimental results reveal that our scheme can guarantee the delay requirement, meanwhile, significantly reduce the energy consumption compared with prior schemes.

Summary Because of the exponential growth of mobile users' demand for multimedia services in recent years, the increasing network traffic load gets a close attention of the mobile network operators. For the mobile traffic explosion issue to be solved, there are many efforts trying to offload the mobile traffic from infrastructure cellular links to direct local short‐range communications among groups of users, which is called device‐to‐device sharing (D2D) in mobile social networks. Although there have been a number of studies for improving the exploitation of friends, contents, and sharing performance, there is no any large‐scale measurement‐based study to analyze the realistic D2D sharing service. We focus on the empirical trace fromXender, a popular mobile application for D2D sharing, and implement an effective big data processing platform based onSparkwith customized algorithms. Extensive analysis and discussions are carried out from the perspectives of general time series statistics, content properties, and social graph basics. The trace‐driven analysis exploits a number of implications regarding power law distribution for content popularity disparity, clustering effects of user relationships, and so on. We further discuss the potentials of improvingXender's quality of service and optimizing its system resource, and hopefully, our study can offer useful guidelines for not onlyXenderbut also those growing global social D2D sharing services.

Summary Adaptive routing designing for network‐on‐chip architecture is complicated to guarantee deadlock‐free. In this paper, we present methodology to design efficient multi‐path routing algorithms for specific application wormhole network‐on‐chip system to guarantee deadlock‐free. The methodology presents an analytical model for overall average delay of wormhole network‐on‐chip by queueing theory and uses a linear programming–based methodology to minimize overall average delay. The methodology guarantees deadlock freedom by efficient branch‐and‐bound algorithm. Compared to other adaptive deadlock‐free algorithms, results obtained through simulation show that the algorithms can potentially achieve better throughput and latency.

In large-scale Internet applications running on geographically distributed datacenters, such as video streaming, it is important to efficiently allocate requests among datacenters. To the best of our knowledge, existing approaches, however, either solely focus on minimizing total cost for provider, or guaranteeing QoS for end-users. In this paper, we apply the software defined network (SDN) controller to enable the central control of the entire network, and propose a joint optimization model to consider high bandwidth utilization for provider and low delay for users. We present the Nash bargaining solution (NBS) based method to model both requirements of provider's high bandwidth utilization and end-users' low delay. Specifically, we formulate the design of request allocation under those requirements as an optimization problem, which is NP-hard. To solve such hard optimization problem, we develop an efficient algorithm blending the advantages of Logarithmic Smoothing technique and the auxiliary variable method. According to the theoretical analysis, we verify the existence and uniqueness of our solution and the convergence of our algorithm. We conduct a large amount of experiments based on real-world workload traces and demonstrate the efficiency of our algorithm compared to both greedy and locality algorithms.

For large-scale RFID systems, this paper studies the practically important problem of target tag information collection, which aims at collecting information from a specific set of target tags instead of all. However, the existing solutions are of low time-efficiency because of two reasons. First, the serious collisions among tags due to hashing randomness seriously reduce the frame utilization, whose upper bound is just 36.8 percent. Second, they cannot efficiently distinguish the target tags from the non-target tags and thus inevitably collect a lot of irrelevant information on non-target tags, which further deteriorates the effective utilization of the time frame. To overcome the above two drawbacks, this paper proposes the minimal Perfect hashing-based Information Collection (PIC) protocol, which first leverages lightweight indicator vectors to establish a one-to-one mapping between target tags and slots, thereby improving the frame utilization to nearly 100 percent; and then uses the novel data structure called Minimal Perfect Hashing based Filter (MPHF) to filter out the non-target tags, thereby preventing them from interfering with the process of collecting information from target tags. Sufficient theoretical analyses are also presented in this paper to minimize the execution time of the proposed PIC protocol. Extensive simulations are conducted to compare the proposed PIC protocol with prior works side-by-side. The simulation results demonstrate that PIC significantly outperforms the state-of-the-art protocols in terms of time-efficiency.

For categorized RFID systems, this paper studies the practically important problem of top-k queries, which is to find the top-k smallest and (or) the top-k largest categories, as well as the sizes of such categories. In this paper, we propose a Top-k Query (TKQ) protocol and two supplementary techniques called segmented perfect hashing (SPH) and switching to framed slotted aloha (STA) for optimizing TKQ. First, TKQ lets each tag choose a time slot to respond to the reader with a single-one geometric string using the ON-OFF Keying modulation. TKQ leverages the length of continuous leading 1 s in the combined signal to estimate the corresponding category size. TKQ can quickly eliminate most categories whose sizes are significantly different from the top-k boundary, and only needs to perform accurate estimation on a limited number of categories that may be within the top-k set. We conduct rigorous analysis to guarantee the predefined accuracy constraints on the query results. Second, to alleviate the low frame utilization of TKQ, we propose the SPH scheme, which improves its average frame utilization from 36.8% to nearly 100% by establishing a bijective mapping between tag categories and slots. To minimize the overall time cost, we optimize the key parameter that trades off between communication cost and computation cost. Third, we observed from the simulation traces that TKQ+SPH pays most execution time on querying a small number of remaining categories whose sizes are close to the top-k boundary, which sometimes even exceeds the time cost for precisely identifying these remaining tags. Motivated by this observation, we propose the STA scheme to dynamically determine when we should terminate TKQ+SPH and switch to use FSA to finish the rest of top-k query. Experimental results show that TKQ+SPH+STA not only achieves the required accuracy constraints, but also achieves several times faster speed than the existing protocols.

This paper concerns the practically important problem of multi-category radio frequency identification (RFID) estimation: given a set of RFID tags, we want to quickly and accurately estimate the number of tags in each category. However, almost all the existing RFID estimation protocols are dedicated to the estimation problem on a single set, regardless of tag categories. A feasible solution is to separately execute the existing estimation protocols on each category. The execution time of such a serial solution is proportional to the number of categories, and cannot satisfy the delay-stringent application scenarios. Simultaneous RIFD estimation over multiple categories is desirable, and hence, this paper proposes an approach called simultaneous estimation for multi-category RFID systems (SEM). SEM exploits the Manchester-coding mechanism, which is supported by the ISO 18000-6 RFID standard, to decode the combined signals, thereby simultaneously obtaining the reply status of tags from each category. As a result, multiple bit vectors are decoded from just one physical slotted frame. Built on our SEM, many existing excellent estimation protocols can be used to estimate the tag cardinality of each category in a simultaneous manner. To ensure the predefined accuracy, we calculate the variance of the estimate in one round, as well as the variance of the average estimate in multiple rounds. To find the optimal frame size, we propose an efficient binary search-based algorithm. To address significant variance in category sizes, we propose an adaptive partitioning (AP) strategy to group categories of similar sizes together and execute the estimation protocol for each group separately. Compared with the existing protocols, our approach is much faster, meanwhile satisfying the predefined estimation accuracy. For example, with 20 categories, the proposed SEM+AP is about seven times faster than prior estimation schemes. Moreover, our approach is the only one whose normalized estimation time (i.e., time per category) decreases as the number of categories increases.

With the increasing popularization of radio frequency identification (RFID) technology in the retail and logistics industry, RFID privacy concern has attracted much attention, because a tag responds to queries from readers no matter they are authorized or not. An effective solution is to use a commercially available blocker tag that behaves as if a set of tags with known blocking IDs are present. However, the use of blocker tags makes the classical RFID estimation problem much more challenging, as some genuine tag IDs are covered by the blocker tag and some are not. In this paper, we propose RFID estimation scheme with blocker tags (REB), the first RFID estimation scheme with the presence of blocker tags. REB uses the framed slotted Aloha protocol specified in the EPC C1G2 standard. For each round of the Aloha protocol, REB first executes the protocol on the genuine tags and the blocker tag, and then virtually executes the protocol on the known blocking IDs using the same Aloha protocol parameters. REB conducts statistical inference from the two sets of responses and estimates the number of genuine tags. Rigorous theoretical analysis of parameter settings is proposed to guarantee the required estimation accuracy, meanwhile minimizing the time cost and energy cost of REB. We also reveal a fundamental tradeoff between the time cost and energy cost of REB, which can be flexibly adjusted by the users according to the practical requirements. Extensive experimental results reveal that REB significantly outperforms the state-of-the-art identification protocols in terms of both time efficiency and energy efficiency.

In large-scale radio frequency identification (RFID)-enabled applications, we sometimes only pay attention to a small set of key tags, instead of all. This paper studies the problem of key tag population tracking, which aims at estimating how many key tags in a given set exist in the current RFID system and how many of them are absent. Previous work is slow to solve this problem due to the serious interference replies from a large number of ordinary (i.e., non-key) tags. However, time-efficiency is a crucial metric to the studied key tag tracking problem. In this paper, we propose a singleton slot-based estimator, which is time-efficient, because the RFID reader only needs to observe the status change of expected singleton slots corresponding to key tags instead of the whole time frame. In practice, the ratio of key tags to all current tags is small, because key members are usually rare. As a result, even when the whole time frame is long, the number of expected singleton slots is limited and the running of our protocol is very fast. To obtain good scalability in large-scale RFID systems, we exploit the sampling idea in the estimation process. A rigorous theoretical analysis shows that the proposed protocol can provide guaranteed estimation accuracy to end users. Extensive simulation results demonstrate that our scheme outperforms the prior protocols by significantly reducing the time cost.

To deal with the explosive growth in multimedia service requests in mobile networks, caching contents at the cells (base stations) is regarded as an effective emerging technique to reduce the duplicated transmissions of content downloads, while heterogeneous networks (HetNets) are regarded as an effective technique to increase the network capacity. Yet, the combination of content caching and HetNets for future networks (i.e., 5G) is still not well explored. In this paper, we propose an efficient collaborative multi-tier caching framework in Het-Nets. In particular, based on patterns of user requests, link capacities, heterogenous cache sizes, and the derived system topology, we focus on exploring the maximum capacity of the network infrastructure so as to offload the network traffic and support users' content requests locally. Due to the NP-hardness of the complex multi-tier caching problem, we approximately decompose it into some subproblems that focus on the caching cooperation at different tiers by utilizing the derived system topology. Our proposed framework is low-complexity and distributed, and can be used for practical engineering implementation. Trace-based simulation results demonstrate the effectiveness of the proposed framework.

This paper studies the practically important problem of top-k queries, which is to find the top k largest categories and their corresponding sizes. In this paper, we propose a Top-k Query (TKQ) protocol and a technique that we call Segmented Perfect Hashing (SPH) for optimizing TKQ. Specifically, TKQ is based on the framed slotted Aloha protocol. Each tag responds to the reader with a Single-One Geometric (SOG) string using the ON-OFF Keying modulation. TKQ leverages the length of continuous leading 1s in the combined signal to estimate the corresponding category size. TKQ can quickly eliminate the sufficiently small categories, and only needs to focus on a limited number of large-size categories that require more accurate estimation. We conduct rigorous analysis to guarantee the predefined accuracy constraints. To further improve time-efficiency, we propose the SPH scheme, which improves the average frame utilization of TKQ from 36.8% to nearly 100% by establishing a bijective mapping between tag categories and slots. To minimize the overall time cost, we optimize the key parameter that trades off between communication cost and computation cost. Experimental results show that our TKQ+SPH protocol not only achieves the required accuracy constraints, but also achieves a 2.6~7x faster speed than the existing protocols.

In data centers, a lot of cluster computing applications follow a coflow pattern. On the other hand, network function virtualization (NFV) sufficiently improves the performance of the data center network. However, coflows encounter extremely different processing delays under diverse network functions. Traditional coflow scheduling schemes become insufficient in this situation. Based on the observation that the benefit of coflows is closely related to the data rates of flows, we propose DRGC (Data Rate Guarantee for Coflow) to guarantee the data rate requirements of coflows in the NFV environment. We prioritize the scheduling sequence of coflows, precisely allocate data rates for individual flows, and design an efficient scheduling algorithm. DRGC maintains the desired data rates of coflows with higher priorities at middleboxes and leaves more scheduling opportunities to the ones with lower priorities. In the large-scale trace-driven experiment, DRGC efficiently guarantees the data rate requirements of coflows and supports more 15% workload, compared with other scheduling schemes.

This paper studies the problem of data collection in sensor-augmented RFID networks: how to quickly obtain the error-bounded data from sensor-augmented RFID tags. Existing data collection protocols require each tag to transmit the sensor data to the reader through a low-rate channel. However, in large-scale RFID system, they take too long time and block other time-sensitive operations. By exploring the correlation of sensor data, our Sampling-based Information Collection (SIC) protocol significantly reduces the number of responding tags. Specifically, SIC obtains an error bound based on the estimation model by using some randomly-sampled data. The error bound is expected to maximize the number of data within it. These data can be seen as a cluster and be approximated by one value within the error bound. Then, SIC only needs to collect the data of out this cluster, thereby significantly reducing the data transmission. It minimizes the execution time by optimizing the sample size and estimating the number of tags out of the error bound. We conduct extensive simulations to evaluate the performance of SIC and compare it with three major related work. The results demonstrate that SIC is 1 to 10 times faster than the state-of-the-art solution.

Facing the spectrum supply-demand gap, heterogeneous network (HetNet) is a promising approach to achieve drastic gains in network coverage and capacity compared with macro-only networks, thus making it especially attractive to network operators. On the other hand, software-defined networking brings a number of advantages along with many challenges. One particular concern is on the resilience for in-band fashioned control plane. Existing approaches mainly rely on a local rerouting policy when performing the routing protection for the target sessions in software-defined networks. However, such a policy would potentially bring congestions in the neighbouring links of the failed one. To this end, we study a weighted cost-minimization problem, where the traffic load balancing and control-channel setup cost are jointly considered. Because this problem is NP-hard, we first propose a near-optimal Markov approximation-based approach for in-band-fashioned software-defined HetNets. We then extend our solution to an online case that handles a single-link failure. We also conduct theoretical analysis on the performance fluctuation due to the single-link failure. We finally carry out experiments by experimental simulation. The extensive numerical results show that the proposed algorithm has fast convergence and high efficiency in resource utilization.

The metabolic network model allows for an in-depth insight into the molecular mechanism of a particular organism. Because most parameters of the metabolic network cannot be directly measured, they must be estimated by using optimization algorithms. However, three characteristics of the metabolic network model, i.e., high nonlinearity, large amount parameters, and huge variation scopes of parameters, restrict the application of many traditional optimization algorithms. As a result, there is a growing demand to develop efficient optimization approaches to address this complex problem. In this paper, a Kriging-based algorithm aiming at parameter estimation is presented for constructing the metabolic networks. In the algorithm, a new infill sampling criterion, named expected improvement and mutual information (EI&MI), is adopted to improve the modeling accuracy by selecting multiple new sample points at each cycle, and the domain decomposition strategy based on the principal component analysis is introduced to save computing time. Meanwhile, the convergence speed is accelerated by combining a single-dimensional optimization method with the dynamic coordinate perturbation strategy when determining the new sample points. Finally, the algorithm is applied to the arachidonic acid metabolic network to estimate its parameters. The obtained results demonstrate the effectiveness of the proposed algorithm in getting precise parameter values under a limited number of iterations.

Internet attacks, such as distributed denial-of-service attacks and worm attacks, are increasing in severity and frequency. Identifying and mitigating realtime attacks are an important and challenging task for network administrators. An infected host can make a large number of connections to distinct destinations during a short time. Such a host is called a superpoint. Detecting superpoints can be utilized for traffic engineering and anomaly detection. This paper proposes a novel data streaming method for detecting superpoints and proves guarantees on its accuracy with low memory requirements. The superior performance of this method comes from a new data structure, called vector bloom filter (VBF), which is a variant of standard BF. The VBF consists of six hash functions, four of which take some consecutive bits from the input string as the corresponding value, respectively. The information of superpoints is obtained by using the overlapping of hash bit strings of the VBF. Theoretical analysis and experimental results show that the proposed method can detect superpoints precisely and efficiently through comparison with other existing approaches.

In Wireless Sensor Networks (WSNs), end-to-end delay is an important metric because the sensed information is necessary to be transmitted to the sink node within a predefined time threshold. Therefore, opportunistic routing protocols are proposed to reduce the end-to-end delay. As a matter of fact, increasing the number of wake-up slots will certainly reduce the transmission delay, however, also consumes more energy. Hence, it is interesting to control the number of wake-up slots to investigate the trade-off between the end-to-end delay and the energy-efficiency. To the best of our knowledge, no existing work takes both of end-to-end delay and energy-efficiency into consideration in the opportunistic routing networks. Therefore, this paper studies how to minimize the energy-consumption while guaranteeing that the expected end-to-end delay is below a given threshold. To solve this problem, we propose an energy-based Dynamic Duty Cycle Scheme(D2CS) in opportunistic routing network. Specifically, we first present an analytical model to measure the expected end-to-end delay. Then, we decompose the studied problem into a set of single-hop delay guarantee problems and using the feedback controller to approximate the optimal solution. Finally, extensive simulations are conducted to evaluate the performance of the proposed D2CS algorithm. The experimental results reveal that our D2CS can guarantee the delay requirement, meanwhile, significantly reduce the energy consumption compared with prior schemes.

Mobile social networks have been increasingly popular with the explosive growth of mobile devices. Mobile users are allowed to interact with potential friends within a certain distance. Motivated by this feature, many exciting applications have been developed, yet the challenge of privacy protection is thus aroused. In this paper, we propose an efficient customized privacy preserving friend discovery mechanism, which not only protects the privacy of users' profile, but also establishes a verifiable secure communication channel between matched users. Besides, the initiator has the freedom to set a customized request profile by choosing the interested attributes and giving each attribute a specific value. Moreover, the request profile's privacy protection level is customized by the initiator according to his/her own privacy requirements. We also consider the collusion attacks among unmatched users. To the best of our knowledge, this is the first work to address such a security threat. Our protocol guarantees that only exactly matched users are able to communicate with the initiator securely, while little information can be obtained by other participants. To increase the matching efficiency, our design adopts the Bloom filter to efficiently exclude most unmatched users. As a result, our design effectively protects the profile privacy and efficiently decreases the computational overhead. Security analysis and performance evaluation are conducted to justify the superiority of our protocol.

Traditional networks are surprisingly fragile and difficult to manage. Software Defined Networking (SDN) gained significant attention from both academia and industry, as if simplify network management through centralized configuration. Existing work primarily focuses on networks of limited scope such as data-centers and enterprises, which makes the development of SDN hindered when it comes to large-scale network environments. One way of enabling communication between data-centers, enterprises and ISPs in a large-scale network is to establish a standard communication mechanism between these entities. In this paper, we propose Zebra, a framework for enabling communication between different SDN domains. Zebra has two modules: Heterogeneous Controller Management (HCM) module and Domain Relationships Management (DRM) module. HCM collects network information from a group of controllers with no interconnection and generate a domain-wide network view. DRM collects network information from other domains to generate a global-wide network view. Moreover, HCM supports different SDN controllers, such as floodlight, maestro and so on. To test this framework, we develop a prototype system, and give some experimental results.

The widely used RFID tags impose serious privacy concerns as a tag responds to queries from readers no matter they are authorized or not. The common solution is to use a commercially available blocker tag which behaves as if a set of tags with known blocking IDs are present. The use of blocker tags makes RFID estimation much more challenging as some genuine tag IDs are covered by the blocker tag and some are not. In this paper, we propose REB, the first RFID estimation scheme with the presence of blocker tags. REB uses the framed slotted Aloha protocol specified in the C1G2 standard. For each round of the Aloha protocol, REB first executes the protocol on the genuine tags and the blocker tag, and then virtually executes the protocol on the known blocking IDs using the same Aloha protocol parameters. The basic idea of REB is to conduct statistically inference from the two sets of responses and estimate the number of genuine tags. We conduct extensive simulations to evaluate the performance of REB, in terms of time-efficiency and estimation reliability. The experimental results reveal that our REB scheme runs tens of times faster than the fastest identification protocol with the same accuracy requirement.

As applications of computing systems have permeated in all aspects of daily life, the power of computing system has become increasingly critical, which offers many challenging problems on the area of efficiency, performance, reliability, security, and interoperability. New programming paradigms, interconnection networks, and storage systems have joined the traditional workflow and parallel computing technologies for the highest-performance systems. This special issue presents the recent advances in parallel computing and distributed network, which were selected out of the significantly extended versions of accepted papers in the 2014 World Ubiquitous Science Congress (U-Science 2014) 1, the 14th International Conference on Algorithms and Architectures for Parallel Processing (ICA3PP 2014) 2, and a large number of open submissions. The selection has been very rigorous, and only the best papers were selected. Tang et al. 3 note that there exists the correlation of volunteer or desktop failures in Desktop Grid and Volunteer Computing Systems. To achieve long-term and sustained high throughput, they present a hybrid MapReduce (HybridMR) computing environment, in which the cluster nodes and the volunteer computing nodes are integrated. HybridMR includes two innovative solutions. The first one is a hybrid distributed file system to alleviate the volatility of desktop PCs. The second innovation is a new node priority-based fair scheduling algorithm to achieve both data storage balance and job assignment balance. They also provide performance evaluation on the I/O, fault-tolerance and cost-saving of HybridMR, denoting that new model is not only able to achieve a higher throughput and efficiency, but also able to achieve the “green computing” goal. Ji et al. 4 study the efficient and scalable RNN algorithms in the distributed environment. Noted that the major downside of the existing RNN is its inherent sequential nature and using in-memory algorithm, they firstly use the inverted grid, not the R-tree or Voronoi to index the data. It is proved that the grid increases opportunities for parallelism. Furthermore, two pruning ways Lazy-Scalable Reverse Nearest Neighbor (SRNN) and Eager-SRNN are proposed to improve the performance. Finally, they perform extensive experiments on both real and synthetic datasets, demonstrating that their methods outperform the state-of-the-art algorithms in scalable RNN queries. As in future extreme-scale systems, one compute node will have multiple accelerators. Dong et al. 5 make an attempt on this kind of programming clusters that have multiple Xeon Phi coprocessors in each compute node. To increase the efficiency, they present an offload programming approach that allows each coprocessor to run an independent sub-program, while bi-directional and asynchronous coprocessor–coprocessor data transfers are directly enabled by Intel's low-level APIs of COI and SCIF. They also present a hybrid programming strategy combining techniques such as Message Passing Interface (MPI), Open Multi-Processing (OpenMP), Coprocessor Offload Infrastructure (COI) and Symmetric Communication Interface (SCIF), thus extending their work to cover clusters with multi-coprocessor nodes. They also provide performance results of the proposed COI-SCIF approach running on Tianhe-2, in terms of both bandwidth benchmark measurements and time usages of a real-world 3D application. Topology structure is usually viewed as a big issue for interconnection networks. Zhang et al. 6 detailed studied one of the typical topology structures, hyper-star graph HS(2n, n), and found some interesting and attractive properties. More specifically, they show that the surface area of HS(2n; n) is . Furthermore, they prove that HS(2n, n) is isomorphic to the well-known middle cube, thus linking the Hamiltonicity of HS(2n, n) to that of the middle cube. Finally, they study the embedding properties of HS(2n, n) by showing that full binary trees can be embedded into the network with dilation 1, and an optimal algorithm is found for performing neighborhood broadcasting on HS(2n, n). Noting that the balance of performance and hardware costs becomes quite challenging in traditional hypercube-based topology structures, Qi et al. 7 build a new interconnection topology structure named EFH. Different from EH, some complementary edges are added to link a node with its farthest node of the Hypercube. Owe to these complementary edges, the authors investigate that the network diameter of EFH is about half of the diameter of EH. Furthermore, they design a more efficient routing algorithm and a load balancing algorithm used in their EFH structure. They also give a lot of strict proofs on the properties of EFH and analyze its fault tolerance capabilities such as fault diameter and cost effectiveness factor. Aiming at handling distance fraud attacks and relay attacks in anonymous RFID applications, Yang et al. 8 presents an improved Distance-Bounding Trust Protocol (DBTP). With the proposed DBTP, a tag can defend distance fraud from the malicious reader by the output of trust values. In addition, they deploy trusted third party architecture to provide anonymity for tags in anonymous RFID systems without requiring tag identifiers. This enabled DBTP to further defend relay attacks in anonymous RFID systems. Finally, they build a prototype of DBTP using commercial RFID readers to track off-the-shelf RFID tags and evaluate the performance of DBTP not only by theoretical analysis but also by a large set of extensive simulations. Cheng et al. 9 noted that in hybrid storage systems, the two-level cache DRAM and SSD typically use independent cache replacement policies, which makes cache resource management inefficient and lower the system performance. They propose a novel AMC replacement algorithm to deal with this problem. Compared with classic multi-level exclusive caching techniques, they introduce combined selective Promote and Demote operations to dynamically determine the level and keep “hot” data blocks in DRAM and SSD caches. Furthermore, they design an online method using probabilistic Promote and Demote values that are adjusted via the usage of blocks already cached. In experiments, they show the proposed AMC algorithm reduces average response time and increases SSD lifetime, compared with the traditional multi-level cache algorithms 2C-Least Recently Used (LRU), ind-LRU, and exc-LRU. We hope that you will enjoy reading these papers in this special issue. We would like to thank the authors for contributing their papers to this issue, and thank all the reviewers for their time and constructive reviews. Finally, we would like to thank the editors of Concurrency and Computation: Practice and Experience for providing this opportunity to publish this special issue.

This special issue presents the recent advances in cloud computing and software-defined network, which \were selected out of the significantly extended versions of accepted papers in the Fifth IEEE International Conference on Big Data and Cloud Computing (BDCloud 2015) 1, the Ninth International Conference on Frontier of Computer Science and Technology 2, and a large number of open submissions. The selection has been very rigorous, and only the best papers were selected. Wei et al. 3 observe that most existing message forwarding algorithms in delay-tolerant networks prefer to deliver messages to the nodes with a higher popularity or centrality. This forwarding scheme can achieve high delivery ratio and low end-to-end delay but is prone to cause unfair load distribution and further lead to network congestion. To tackle this, they first track the evolution of communities through a novel distributed community detection approach. The second one is to develop a congestion avoidance mechanism to divert load away from congested areas and further present a congestion-aware message forwarding algorithm where messages can avoid being transmitted to the congested nodes Since the rapid growth of large-scale online services, massive amounts of the generated traffic have been seen in the data center network. Li et al. 4 study the emerging congestion problem in the software-defined data center network. Note that existing approaches are either hard to be implemented in hardware or unable to obtain the optimal solutions. The authors first propose a heuristic algorithm for efficiently compute a timeslot allocation for the coming packets. Further, they model the path selection as a bin-packing problem. By seamlessly combining the timeslot allocation and path selection, each data packet will not suffer queuing and waiting in the data center network. Anomaly detection is an effective approach to enhance availability and reliability of cloud infrastructures. Hong et al. 5 study the anomaly detection problem in cloud computing systems without the need for prior knowledge about normal or anomalous behaviors. They propose an unsupervised online anomaly detection scheme based on hidden Markov model. In order to achieve high scalability, their proposed algorithm runs in a distribution manner among multiple computing machines in the cloud. They also perform extensive experiments based on real data sets to validate the high detection accuracy for their proposed algorithm. Because of the benefits of reducing the communication overhead, distributed data-centric storage in wireless sensor networks have received considerable attention. Xu et al. 6 focus on the big data storage problem in wireless sensor network with the nonuniform node distribution. Note that most existing distribution methods can significantly consume more energy and are unable to deal with the case of nonuniform sensor nodes distribution. To address this issue, they propose an efficient storage retrieval algorithm to estimate the real distribution of the sensor nodes and the real addresses of these nodes. Based on this algorithm, they further take the data redundancy among sensor nodes into account and exploit an efficient routing mechanism. Incorporating cloud computing into vehicular networks is a promising solution to the collection, storage, and analysis of big traffic-related data but can lead to new challenges to the allocation and management for cloud resources in road-side cloudlet. Yao et al. 7 study a VM migration problem with the goal of minimizing the total network cost, by making the decisions on which VM should be migrated and where the VM shall be migrated. They further formulate an optimization for the static off-line VM placement problem and then propose a heuristic algorithm with polynomial time to solve the optimization. NoSQL systems, replicating and partitioning data over many servers for improving the performance, are widely used for storing big data. Conventional radon virtual nodes and manual configuration methods for consistent hashing can significantly lead to imbalanced data partition. Huang et al. 8 study the performance degradation problem caused by the imbalanced data partition. They first propose a novel imbalance coefficient of data distribution. They further propose a dynamic programming algorithm to compute the position of the new coming node in the consistent ring. Finally, they conduct comprehensive simulations based on a benchmark Yahoo Cloud Serving Benchmark (YCSB) to show the benefit of their proposed algorithm. Data centers are increasingly deploying the NUMA architecture. Zhu et al. 9 focus on the performance degradation problem when running multi-threaded programs on such NUMA systems. Note that the existing works mainly use the single-threaded multi-programming workloads to study the performance of NUMA on the resource contention and data locality. To solve the performance lagging problem, they propose a novel scheduler—symmetric scheduler, which can balance the number of costly remote shared data accesses for threads on NUMA systems. Finally, they perform extensive simulations on the PARSEC benchmark, and their proposed schedulers can significantly outperform Linux kernel scheduling mechanism. As the number of functionally equivalent services in the cloud grows, collaborative service QoS prediction has recently garnered increasing attention. Tang et al. 10 propose a collaborative QoS prediction method with location-based data smoothing, for addressing the data sparsity issue and improving the QoS prediction accuracy. Note that existing solutions, simply exploring the historical QoS information generated by interactions between users and services, however, can significantly suffer from the data sparsity issue. To address this issue, the authors firstly compute neighborhoods of users and services based on their locations, which provide a basis for data smoothing. They further combine user-based and service-based collaborative filtering techniques to make QoS predictions. Finally, they conduct comprehensive experiments on real service invocation dataset to validate the performance of their proposed QoS prediction method. We hope that you will enjoy reading these papers in this special issue. We would like to thank the authors for contributing their papers to this issue and thank all the reviewers for their time and constructive reviews. Finally, we would like to thank the editors of Concurrency and Computation: Practice and Experience for providing this opportunity to publish this special issue.

Summary This paper focuses on the topology structure of interconnection networks. To overcome drawbacks in the existing hypercube structure, we present an exchanged folded hypercube (EFH) structure, which is an improvement of the exchanged hypercube. Compared with the existing hypercube structures, EFH shows better performance in terms of many metrics such as smaller diameter, lower cost factor, and constant node degree. In this paper, we first introduce the structure of an EFH; then, we propose a routing algorithm and a load‐balancing algorithm for EFHs. Finally, we analyze the fault tolerance characteristics of EFHs including fault diameter and the cost effectiveness factor. Copyright © 2015 John Wiley & Sons, Ltd.

Summary With an explosive growth of the datacenter research, virtual machines migration aiming at optimization of virtual machines placement is a major technology of improving power efficiency and resource utilization in the datacenter. While recent studies have primarily focused on maximizing resource utilization or minimizing migration cost separately, there has been little attention on jointly taking these two objectives into account. In this paper, we present the optimization model taking minimum migration cost and maximum resource utilization with multi‐resources such as storage, bandwidth, CPU, and disk space into account. The optimization of our proposed model is non‐deterministic polynomial‐time hard. According to the model, we present an efficient approximate algorithm based on bin packing algorithm, called MinCost, to resolve our model and obtain a near‐optimal solution. Finally, the simulation results in this paper show that our model and algorithm is efficient. Copyright © 2015 John Wiley & Sons, Ltd.

Radio Frequency Identification (RFID) technology has been widely used in inventory management in many scenarios, e.g., warehouses, retail stores, hospitals, etc. This paper investigates a challenging problem of complete identification of missing tags in large-scale RFID systems. Although this problem has attracted extensive attention from academy and industry, the existing work can hardly satisfy the stringent real-time requirements. In this paper, a Slot Filter-based Missing Tag Identification (SFMTI) protocol is proposed to reconcile some expected collision slots into singleton slots and filter out the expected empty slots as well as the unreconcilable collision slots, thereby achieving the improved time-efficiency. The theoretical analysis is conducted to minimize the execution time of the proposed SFMTI. We then propose a cost-effective method to extend SFMTI to the multi-reader scenarios. The extensive simulation experiments and performance results demonstrate that the proposed SFMTI protocol outperforms the most promising Iterative ID-free Protocol (IIP) by reducing nearly 45% of the required execution time, and is just within a factor of 1.18 from the lower bound of the minimum execution time.

In Delay Tolerant Networks (DTNs), the connections between mobile nodes are always disrupted and constant end-to-end paths rarely exist. In order to cope with these communication challenges, most existing DTN routing algorithms favour the “multi-hop forwarding” fashion where a message can be forwarded by multiple relay nodes in the hope that one of the employed relay nodes can deliver the message to the destination node. Since aggressively employing relay nodes may incur the intolerable delivery cost in DTNs, it is meaningful to design a cost-efficient routing algorithm that can achieve a high delivery performance. In this paper, we first design a novel delivery metric to measure the forwarding capability of nodes. Then, we utilize small-world properties to design the principles of relay node selection, e.g., limiting the number of relays and finding the appropriate relay nodes, and further develop a cost-efficient social-aware forwarding algorithm called TBSF. Extensive simulations on real mobility traces are conducted to evaluate the performance of TBSF, and the results demonstrate its efficiency and usefulness.

Unknown RFID tags appear when the unread tagged objects are moved in or tagged objects are misplaced. This paper studies the practically important problem of unknown tag detection while taking both time-efficiency and energy-efficiency of battery-powered active tags into consideration. We first propose a Sampling Bloom Filter which generalizes the standard Bloom Filter. Using the new filtering technique, we propose the Sampling Bloom Filter-based Unknown tag Detection Protocol (SBF-UDP), whose detection accuracy is tunable by the end users. We present the theoretical analysis to minimize the time and energy costs. SBF-UDP can be tuned to either the time-saving mode or the energy-saving mode, according to the specific requirements. Extensive simulations are conducted to evaluate the performance of the proposed protocol. The experimental results show that SBF-UDP considerably outperforms the previous related protocols in terms of both time-efficiency and energy-efficiency. For example, when 3 or more unknown tags appear in the RFID system with 30000 known tags, the proposed SBF-UDP is able to successfully report the existence of unknown tags with a confidence more than 99%. While our protocol runs 9 times faster than the fastest existing scheme and reducing the energy consumption by more than 80%.

Large-scale data centers are the key infrastructures for hosting and running a variety of applications. Besides traditional L2/L3 devices, middleboxes are widely deployed in data centers and perform many important functions, e.g., the intrusion detection and firewall. Middleboxes are equipped with multiple kinds of resources, such as CPU and memory. Data flows undergoing different functions have heterogeneous processing time requirements on diverse resources. Researchers are in a dilemma as to how to provide fair service for flows and efficiently utilize those scarce resources. To address this problem, we propose a novel packet scheduling method, active time fairness queuing (ATFQ), for multi-resource environments. Prior packet scheduling methods usually focus on pursuing the fairness among flows, resulting in enormous waste of those scarce resources. ATFQ overcomes this essentially by redefining the fairness and can maximize the resource utilization with the guarantee of fairness. We conduct extensive simulations to evaluate the performance of ATFQ. The evaluation results demonstrate that flows get better service in many aspects under ATFQ. Meanwhile, the resource utilization rises up by about 10% than the traditional DRFQ, which is one of the mainstream involved methods.

With the problem of congestion becoming more and more serious in data centers, a great deal of attention has been paid to the congestion control technology. In this paper, we first analyze and summarize the reasons that cause congestion in data center. Second, from the point of the network congestion control algorithm, the improvement of packet loss detection, the modification for transmission completion time, adjustment of transmission speed and fairness are discussed, respectively. Finally, we discuss the future work based on the current status of congestion control algorithms.

Shape retrieval is still a challenging problem. To address this problem, there is a growing interest in context-sensitive shape retrieval. In such methods, when computing the similarity between any two shape objects, the influence of their neighbors is propagated by a diffusion process on contextual space. We can re-evaluate the pairwise similarities to get better retrieval results. Although the context-sensitive shape retrieval methods are very promising in theory, but the time efficiency of these methods is a big problem in practice. Because it is a time-consuming to diffuse on the contextual space formed by all shape objects in the database. To improve the time efficiency, we propose a novel framework for context-sensitive shape retrieval. This framework consists of the off-line and the on-line stages. In the off-line stage, we conduct context-sensitive based shape clustering to build index. In the on-line stage, we quickly find candidate shape objects for the query by the index. Then, by diffusion on the context space consisting of candidate objects, we can get the final retrieval results. Compared with the state-of-the-art methods, the proposed framework achieves the same bull's-eye retrieval score with more than 80% reduction of searching time on the standard shape database.

RFID is an emerging technology that provides timely and high-value information to inventory management and object tracking, in which areas that identifying unknown tags completely is crucial. From prior researches in this area, one of the pending problem involves processing redundant time frames due to unknown tag collisions. In this paper, we propose a time-efficient unknown tag identification protocol based on coded filtering vector technique. This vector is able to efficiently separate unknown tags from known tags. It reduces the unknown-known tag collisions as well as the required time frame length. The proposed protocol can achieve the minimal execution time theoretically. And further simulations demonstrate that it performs much better than existing work by decreasing 30% of the total execution time on average.

In RFID-enabled applications, we may pay more attention to key tags instead of all tags. This paper studies the problem of key tag counting, which aims at estimating how many key tags in a given set exist in the current RFID system. Previous work is slow to solve this new problem because of the serious interference replies from the large number of ordinary (i.e., Nonkey) tags. However, time-efficiency is an important metric for the fast tag cardinality estimation in a large-scale RFID system. In this paper, we propose a singleton slot-based estimator, which is time-efficient because the RFID reader only needs to observe the status change of expected singleton slots of key tags instead of the whole time frame. In practice, the ratio of key tags to all current tags is small for "key" members should be rare. As a result, even when the whole time frame is long, the expected singleton slot number is limited and the running of our protocol is fast to achieve estimation accuracy. Rigorous theoretical analysis shows that the proposed protocol can provide guaranteed estimation accuracy to end users. We conduct simulations and implement a prototype of our protocol to verify its efficiency and deployability.

As a kind of interesting mobile application, Mobile Visual Search (MVS) has attracted extensive research efforts from both academy and industry. Most of the MVS systems adopt the client-server framework, in which transmission latency caused by the limited bandwidth in wireless network is a big problem. To address this problem, the state-of-the-art work focuses on designing low bit-rate descriptors for MVS. However, few work focuses on reducing the number of descriptors. To further reduce the latency, we propose a novel framework of MVS based on the weighted matching of dominant descriptor. Firstly, we present an affinity propagation based algorithm for dominant descriptor selection. Secondly, we propose a weighted feature matching method to consider the differences of dominant descriptors in feature matching. By the proposed framework, we not only reduce the network latency in MVS, but also avoid transmitting useless descriptors to improve the retrieval accuracy of MVS. The experimental results on Stanford MVS data set show that when using CHoG descriptors, the proposed framework outperforms the existing framework by reducing more than 40% of the amount of data transmission and increasing 5% of the average retrieval accuracy.

Owing to its superior properties, such as fast identification and relatively long interrogating range over barcode systems, Radio Frequency Identification (RFID) technology has promising application prospects in inventory management. This paper studies the problem of complete identification of missing RFID tag, which is important in practice. Time efficiency is the key performance metric of missing tag identification. However, the existing protocols are ineffective in terms of execution time and can hardly satisfy the requirements of realtime applications. In this paper, a Multi-hashing based Missing Tag Identification (MMTI) protocol is proposed, which achieves better time efficiency by improving the utilization of the time frame used for identification. Specifically, the reader recursively sends bitmaps that reflect the current slot occupation state to guide the slot selection of the next hashing process, thereby changing more empty or collision slots to the expected singleton slots. We investigate the optimal parameter settings to maximize the performance of the MMTI protocol. Furthermore, we discuss the case of channel error and propose the countermeasures to make the MMTI workable in the scenarios with imperfect communication channels. Extensive simulation experiments are conducted to evaluate the performance of MMTI, and the results demonstrate that this new protocol significantly outperforms other related protocols reported in the current literature.

Due to the bandwidth limitation in wireless networks, transmission overhead is a big problem in Mobile Visual Search (MVS). Existing work proposes transmitting the compressed local feature descriptors instead of the query image to reduce the transmission overhead. Although many kinds of compressed descriptors are proposed, designing a suitable lossless compressed descriptor has proven elusive. In this paper, we propose a novel framework for MVS with low transmission overhead rather than focusing on compressed descriptors. The key point of the proposed framework is to migrate the vector quantization in the bag of visual words model from the server to the client. In this framework, no matter what descriptors are used, the client only transmits the ID numbers of the visual words to the server, thereby reaching the minimal possible transmission overhead. To achieve this goal, we present vocabulary decomposition by which we can decompose the large vocabulary into several small ones satisfying storage constraints on mobile devices. In this paper, we first formulate vocabulary decomposition as an optimization problem. We then present Joint Product Quantization (JPQ) and Joint Optimized Product Quantization (JOPQ) to address the proposed optimization problem. Finally , we conduct a large number of simulation experiments and real experiments. The experimental results show that the proposed framework outperforms the existing framework by reducing more than 95% of the transmission overhead.

The articles in this special section focus on trust, network security, and privacy deployed in parallel and distributed systems.

Cognitive radio networking allows the unlicensed secondary users to opportunistically access the licensed spectrum as long as the performance of the licensed primary users does not degrade. This dynamic spectrum access strategy is enabled by cognitive radio coupled with spectrum sensing technologies. Due to the imperfection of wireless transmissions, collaborative spectrum sensing (CSS) has been proposed to significantly improve the probability of detecting the transmissions of primary users. Nevertheless, current CSS techniques are sensitive to malicious secondary users, leading to a high false alarm rate and low detection accuracy on the presence of the primary users. In this paper, we present several robust collaborative spectrum sensing schemes that can calculate a trust value for each secondary user to reflect its suspicious level and mitigate its harmful effect on cooperative sensing. Our approach explores the spatial and temporal correlations among the reported information of the secondary users to determine the trust values. Extensive simulation study has been performed and our results demonstrate that the proposed schemes can guarantee the accuracy of the cooperative sensing system with a low false alarm rate when a considerable number of secondary users report false information.

Owing to its attractive features such as fast identification and relatively long interrogating range over the classical barcode systems, radio-frequency identification (RFID) technology possesses a promising prospect in many practical applications such as inventory control and supply chain management. However, unknown tags appear in RFID systems when the tagged objects are misplaced or unregistered tagged objects are moved in, which often causes huge economic losses. This paper addresses an important and challenging problem of unknown tag identification in large-scale RFID systems. The existing protocols leverage the Aloha-like schemes to distinguish the unknown tags from known tags at the slot level, which are of low time-efficiency, and thus can hardly satisfy the delay-sensitive applications. To fill in this gap, two filtering-based protocols (at the bit level) are proposed in this paper to address the problem of unknown tag identification efficiently. Theoretical analysis of the protocol parameters is performed to minimize the execution time of the proposed protocols. Extensive simulation experiments are conducted to evaluate the performance of the protocols. The results demonstrate that the proposed protocols significantly outperform the currently most promising protocols.

k-means is undoubtedly one of the most popular clustering algorithms owing to its simplicity and efficiency. However, this algorithm is highly sensitive to the chosen initial centers and thus a proper initialization is crucial for obtaining an ideal solution. To address this problem, k-means++ is proposed to sequentially choose the centers so as to achieve a solution that is provably close to the optimal one. However, due to its weak scalability, k-means++ becomes inefficient as the size of data increases. To improve its scalability and efficiency, this paper presents Map Reduce k-means++ method which can drastically reduce the number of Map Reduce jobs by using only one MapReduce job to obtain k centers. The k-means++ initialization algorithm is executed in the Mapper phase and the weighted k-means++ initialization algorithm is run in the Reducer phase. As this new Map Reduce k-means++ method replaces the iterations among multiple machines with a single machine, it can reduce the communication and I/O costs significantly. We also prove that the proposed Map Reduce k-means++ method obtains O(α2)approximation to the optimal solution of k-means. To reduce the expensive distance computation of the proposed method, we further propose a pruning strategy that can greatly avoid a large number of redundant distance computations. Extensive experiments on real and synthetic data are conducted and the performance results indicate that the proposed Map Reduce k-means++ method is much more efficient and can achieve a good approximation.

Radio Frequency Identification (RFID) technology has been widely applied in many scenarios such as inventory control, supply chain management due to its superior properties including fast identification and relatively long interrogating range over barcode systems. It is critical to efficiently identify the unknown tags because these tags can appear when new tagged objects are moved in or wrongly placed. The state-of-the-art Basic Unknown tag Identification Protocol-with Collision-Fresh slot paring (BUIP-CF) protocol can first deactivate all the known tags and then collect all the unknown tags. However, BUIP-CF protocol investigates an ALOHA-like technique and causes too many tag responses, which results in low efficiency. This paper proposes a Fast Unknown tag Identification (FUI) protocol which investigates an indicator vector to label the unknown tags with a given accuracy and removes the time-consuming tag responses in the deactivation phase. FUI also adopts the classical Enhanced Dynamic Framed Slotted ALOHA (EDFSA) protocol to collect the labeled unknown tags. We then investigate the optimal parameter settings to maximize the performance of the proposed FUI protocol. Extensive simulation experiments are conducted to evaluate the performance of the proposed FUI protocol and the experimental results show that it considerably outperforms the state-of-the-art protocol.

In this paper, we consider the problem of cooperative spectrum sensing scheduling (C3S) in a cognitive radio network when there exist multiple primary channels. Deviated from the existing research our work focuses on a scenario in which each secondary user has the freedom to decide whether or not to participate in cooperative spectrum sensing; if not, the SU becomes a free rider who can eavesdrop the decision about the channel status made by others. Such a mechanism can conserve the energy for spectrum sensing at a risk of scarifying the spectrum sensing performance. To overcome this problem, we address the following two questions: “which action (contributing to spectrum sensing or not) to take?” and “which channel to sense?” To answer the first question, we model our framework as an evolutionary game in which each SU makes its decision based on its utility history, and takes an action more frequently if it brings a relatively higher utility. We also develop an entropy based coalition formation algorithm to answer the second question, where each SU always chooses the coalition (channel) that brings the most information regarding the status of the corresponding channel. All the SUs selecting the same channel to sense form a coalition. Our simulation study indicates that the proposed scheme can guarantee the detection probability at a low false alarm rate.

Radio Frequency Identification (RFID) technology is widely used in the the retail, warehouse and supply chain management. However, unknown RFID tags appear when the unregistered tagged objects are moved in or tagged objects are misplaced, which leads to huge economic losses (e.g., misplaced chilled food in a warehouse may quickly decay). This paper studies the practically important problem of unknown tag detection. To the best of our knowledge, this is the first piece of work taking both time-efficiency and energy-efficiency into consideration, where the energy-efficiency is very important when the battery-powered active tags are used. This paper proposes two efficient protocols to address the problem of unknown tag detection. Specifically, the Basic Unknown Tag Detection (B-UTD) protocol leverages a cost-effective filter vector to detect the unknown tags, based on which we then propose a Sampling based Unknown Tag Detection (SUTD) protocol by adopting the well-known sampling idea. We present theoretical analysis to optimize the performance of the proposed protocols. Extensive simulations are conducted to evaluate the performance of the proposed protocols. And the experimental results show that the proposed S-UTD protocol considerably outperforms the most related protocol by reducing more than 90% of the required execution time and energy consumption.

SUMMARY The recent advancement of wireless sensor networks (WSNs) technologies and practices is one of the major roles to the evolution and enrichment of human life with the advent of the next generation convergence network environments. A variety of usages and paradigms for WSNs is a growing interest in the design, development, and deployment of sensor systems for applications with other emerging technologies such as RFID Technology, Multimedia Based Surveillance System, Biomedical Technology, Mobile Agent based Networks, P2P technology and Business Process, and others. With rapid increases in applications, new architecture and techniques are required to support WSNs. Copyright © 2012 John Wiley & Sons, Ltd.

SUMMARY The service‐oriented distributed network requires more stable and persistent services, but the services from dishonest or unstable nodes would damage the correctness and availability of services. To efficiently obtain the services from high dependable nodes, many trust management systems have been developed. However, the previous trust models in distributed networks ignore a fact that the entity itself could provide trust information for services also. In this paper, we present a novel trust management model based on the subjective logic trust for service‐oriented distributed networks. The proposed algorithm involves passive trust of entity and combines the direct trust and recommendation trust. We also propose a novel scheme called passive trust feedback to avoid the deceit of malicious nodes and unstable nodes, and to encourage honest nodes. Our proposed model expands the trust range of service resource via passive trust of entity, which generates a flexible access path to the service resources. Simulations show that the proposed trust management model can significantly improve the feasibility of trust management as well as effectively detect malicious entities. Copyright © 2012 John Wiley & Sons, Ltd.

The topology of interconnection networks plays a key role in the performance of parallel computing systems. A new interconnection network called exchanged crossed cube (ECQ) is proposed and analyzed in this paper. We prove that ECQ has the better properties than other variations of the basic hypercube in terms of the smaller diameter, fewer links, and lower cost factor, which indicates the reduced communication overhead, lower hardware cost, and more balanced consideration among performance and cost. Furthermore, it maintains several attractive advantages including recursive structure, high partitionability, and strong connectivity. Furthermore, the optimal routing and broadcasting algorithms are proposed for this new network topology.

SUMMARY The anonymity and dynamic character of a Peer‐to‐Peer (P2P) network makes it an ideal medium for selfish and vicious action. In order to solve this problem, P2P reputation systems are proposed to evaluate the trustworthiness of peers and to prevent the selfish, dishonest, and malicious peers' behaviors, which collects local reputation scores and aggregates them into the global reputation. In this paper we propose a DHT (Distributed Hash Table) trust overlay network (DHTON) to model the network structure and the storage of reputation information. We also design a robust and distributing reputation system, DHTrust, which takes full advantage of the DHT to distribute local reputation to trade‐off the damage of fake reputation information by genuine reputation information. By using the trust evaluation towards two reputation scores, we can also distinguish and evaluate the fundamental behaviors of peers in the P2P network, i.e. providing service and issuing reputation scores. To adapt to the dynamic P2P networks, we take dynamic node mechanism into account. Our scheme can assure convergence effectiveness and robustness, when nodes enter or leave the system. We conduct extensive simulations to evaluate the performance of DHTrust. The results show that our system makes significant improvement in convergence speed and aggregation accuracy. Moreover, it is robust to malicious peers. Copyright © 2011 John Wiley & Sons, Ltd.

In object-based image retrieval, there are two important issues: an effective image representation method for representing image content and an effective image classification method for processing user feedback to find more images containing the user-desired object categories. In the image representation method, the local-based representation is the best selection for object-based image retrieval. As a kernel-based classification method, Support Vector Machine (SVM) has shown impressive performance on image classification. But SVM cannot work on the local-based representation unless there is an appropriate kernel. To address this problem, some representative kernels are proposed in literatures. However, these kernels cannot work effectively in object-based image retrieval due to ignoring the spatial context and the combination of local features. In this article, we present Adjacent Matrix (AM) and the Local Combined Features (LCF) to incorporate the spatial context and the combination of local features into the kernel. We propose the AM-LCF feature vector to represent image content and the AM-LCF kernel to measure the similarities between AM-LCF feature vectors. According to the detailed analysis, we show that the proposed kernel can overcome the deficiencies of existing kernels. Moreover, we evaluate the proposed kernel through experiments of object-based image retrieval on two public image sets. The experimental results show that the performance of object-based image retrieval can be improved by the proposed kernel.

Multicluster systems have emerged as a promising infrastructure for provisioning of cost-effective high-performance computing and communications. Analytical models of communication networks in cluster systems have been widely reported. However, for tractability and simplicity, the existing models are based on the assumptions that the network traffic follows the nonbursty Poisson arrival process and the message destinations are uniformly distributed. Recent measurement studies have shown that the traffic generated by real-world applications reveals the bursty nature in both the spatial domain (i.e., nonuniform distribution of message destinations) and temporal domain (i.e., bursty message arrival process). In order to obtain a comprehensive understanding of the system performance, a novel analytical model is developed for communication networks in multicluster systems in the presence of the spatio-temporal bursty traffic. The spatial traffic burstiness is captured by the communication locality and the temporal traffic burstiness is modeled by the Markov-modulated Poisson process. After validating its accuracy through extensive simulation experiments, the model is used to investigate the impact of bursty message arrivals and communication locality on network performance. The analytical results demonstrate that the communication locality can relieve the degrading effects of bursty message arrivals on the network performance.

Weak reliability and low energy efficiency are the inherent problems in Underwater Sensor Networks (USNs) characterized by the acoustic channels. Although multiple-path communications coupled by Forward Error Correction (FEC) can achieve high performance for USNs, the low probability of successful recovery of received packets in the destination node significantly affects the overall Packet Error Rate (PER) and the number of multiple paths required, which in turn becomes a critical factor for reliability and energy consumption. In this paper, a novel Multiple-path FEC approach (M-FEC) based on Hamming Coding is proposed for improving reliability and energy efficiency in USNs. A Markovian model is developed to formulate the probability of M-FEC and calculate the overall PER for the proposed decision and feedback scheme, which can reduce the number of the multiple paths and achieve the desirable overall PER in M-FEC. Compared to the existing multipath communication scheme, extensive simulation experiments show that the proposed approach achieves significantly lower packet delay while consuming only 20-30 percent of energy in multiple-path USNs with various Bit Error Rates (BER).

Due to the use of acoustic channels with limited available bandwidth, Underwater Sensor Networks (USNs) often suffer from significant performance restrictions such as low reliability, low energy-efficiency, and high end-to-end packet delay. The provisioning of reliable, energy-efficient, and low-delay communication in USNs has become a challenging research issue. In this paper, we take noise attenuation in deep water areas into account and propose a novel layered multipath power control (LMPC) scheme in order to reduce the energy consumption as well as enhance reliable and robust communication in USNs. To this end, we first formalize an optimization problem to manage transmission power and control data rate across the whole network. The objective is to minimize energy consumption and simultaneously guarantee the other performance metrics. After proving that this optimization problem is NP-complete, we solve the key problems of LMPC including establishment of the energy-efficient tree and management of energy distribution and further develop a heuristic algorithm to achieve the feasible solution of the optimization problem. Finally, the extensive simulation experiments are conducted to evaluate the network performance under different working conditions. The results reveal that the proposed LMPC scheme outperforms the existing mechanism significantly.

The open and anonymous nature of peer-to-peer (P2P) networks makes it an ideal medium for attackers to spread malicious contents, which in turn leads to lower quality of network services due to lack of effective trust management mechanism. To improve the quality of services (or transactions), this paper proposes a novel trust and reputation model, named as GroupTrust, based on peer group and evaluation similarity degree in P2P networks. In the proposed model, trust relationships between peers are divided into three categories: trust relationship within a peer group, trust relationship between different groups, and trust relationship between a peer in a peer group with another peer out of this peer group. The model presents the evaluation similarity degree under different context of services and gives local and global reputation computation. Experimental results demonstrate that this model can get more real trust value and deal with the malicious attacks efficiently by comparison with existing models. © 2010 Wiley Periodicals, Inc.

With the increasing amount of text data stored in relational databases, there is a demand for RDBMS to support keyword queries over text data. As a search result is often assembled from multiple relational tables, traditional IR-style ranking and query evaluation methods cannot be applied directly. In this paper, we study the effectiveness and the efficiency issues of answering top-k keyword query in relational database systems. We propose a new ranking formula by adapting existing IR techniques based on a natural notion of virtual document. We also propose several efficient query processing methods for the new ranking method. We have conducted extensive experiments on large-scale real databases using two popular RDBMSs. The experimental results demonstrate significant improvement to the alternative approaches in terms of retrieval effectiveness and efficiency.

The growing proliferation of wireless devices in contemporary wireless networks requires more spectrum usage. As a consequence, spectrum bands are pressingly getting congested. However, a number of frequency bands licensed to operators are under utilized for transmission. Cognitive Radio (CR), as a promising technique for improving spectrum utilization, can dynamically allocate spectrum. In this paper, we investigate the performance of a CR network composed of a primary/licensed user and a number of secondary/unlicensed users, which are subject to self-similar traffic flows and contend for a unique channel. An analytical model is developed to isolate the primary and secondary users from the original networks. Further, we address the queueing performance of individual secondary users by employing a decomposition approach. The comparison between analytical and simulation results validates the accuracy of the developed model.

The anonymity and dynamic characters of Peer-to-Peer (P2P) system makes it an ideal medium for selfish and vicious action. In order to solve this problem, P2P reputation systems are proposed to evaluate the trustworthiness of peers and to prevent the selfish, dishonest, and malicious peers' behaviors, which collects local reputation scores and aggregates them into the global reputation. In this paper we propose a DHT trust overlay network (DHTON) to model the network structure and the storage of reputation information. We also design a robust and distributing reputation system, DHTrust, which takes full advantage of the Distributed Hash Table (DHT) to distribute local reputation to trade off the damage of fake reputation information by genuine reputation information. By using the trust evaluation towards two reputation scores, we can also distinguish and evaluate the fundamental behaviors of peers in P2P network, i.e., providing service and issuing reputation scores. Through the simulation experiments, we find our system makes significant performance gains in convergence speed and aggregation accuracy, and the most important, is robust to malicious peers.

The provisioning of energy-efficient, reliable and low-delay communication in Underwater Sensor Networks (USNs) is a challenging research issue due to the use of acoustic channels. However, the existing mechanisms for enhancement of energy utilization and Quality-of-Service (QoS) in USNs have not considered noise attenuation in deep water which can deteriorate energy efficiency and QoS seriously. To fill this gap, this paper presents a novel scheme, namely Layered Multi-path Power Control (LMPC), to reduce the energy consumption as well as enhance reliable and robust communications in USNs. To this end, we first formalize an optimization problem to manage transmission power and control data rate across the whole network. The objective is to minimize energy consumption and simultaneously ensure other performance in terms of required packet error rate and maximum power. We then solve the key problems including establishment of the Energy-Efficient Tree (EET)and energy distribution in the tree and further provide a feasible solution for the optimization problem. Finally, the extensive simulation experiments are conducted to evaluate the network performance under different design alternatives. The simulation results show that the developed scheme outperforms the existing mechanism significantly.

Since the generation of detailed traffic statistics does not scale well with link speed, increasingly passive traffic measurement employs sampling at the packet or flow level. Sampling has become an attractive and scalable means to measure flow data on high-speed links. However, knowing the length distributions of traffic flows passing through a network link is useful for some applications such as inferring traffic demands, characterizing source traffic, and detecting traffic anomalies. Passive traffic measurement increasingly makes inferences from sampled network traffic. However, previous work has shown the inaccuracy of estimating flow length distributions from sampled traffic when the sampling is performed at the packet level. In this paper, we propose a novel method that uses flow statistics formed from double-sampled packet stream to infer the absolute frequencies of lengths of flows in the unsampled stream. We achieve this through statistical inference and by exploiting heavy-tailed feather. The method allow us to recover the complete flow length distribution.

With the increasing number of technology areas using Virtual Machine (VM) platforms, challenges exist in Virtual Machine migrating from one physical host to another. However, the complexity of these virtualized environments presents additional management challenges. Unfortunately, many traditional approaches may be either not effective well for reducing downtime or migration time, or not suitable well for Xen VMs platforms. This paper presents the design and implementation of a novel Slowdown Scheduling Algorithm (SSA) for Xen live VM migration. In our SSA methodology, the CPU resources which have been assigned to migration domain are decrease properly. That is, the dirtying page rate is reduced according to the decrease of CPU activity. Experimental results illustrate that our SSA approach can shorten both the total migration time and downtime obviously under high dirty page rate environment.

Abstract Given a set of multimedia objectsR={o1,o2, …,ok} each of which has a set of multiple versionsoi.v={Ai.0,Ai.1, …,Ai.m},i=1, 2, …,k, there is a problem of distributing these objects in a server system so that user requests for accessing specified multimedia objects can be fulfilled with the minimum energy consumption and without significant degrading of the system performance. This paper considers the allocation problem of multimedia objects in multi‐powermode storage systems, where the objects are distributed among multi‐powermode storages based on the access pattern to the objects. We design an underlying infrastructure of storage system and propose a dynamic multimedia object allocation policy based on the designed infrastructure, which integrate and prove the optimality of the proposed policy. Copyright © 2010 John Wiley & Sons, Ltd.

Abstract The provisioning of fairness to various network applications is an important Quality‐of‐Service demand and has thus become a challenging research issue in contemporary communication networks. Deficit Round Robin (DRR) is a promising fair scheduling mechanism owing to its low complexity and excellent ability of achieving a good degree of fairness in terms of throughput. On the other hand, self‐similar traffic was found to be pervasively present in communication networks and has great impact on the performance of scheduling systems. However, to the best of our knowledge, there has not been any analytical model reported in the open literature for the DRR mechanism in the presence of self‐similar traffic. To fill this gap, this paper analytically investigates the queueing performance of DRR and develops a new analytical model for deriving the upper and lower bounds of the queue length distributions of individual traffic flows in DRR scheduling systems subject to self‐similar traffic. Extensive comparison between simulation and analytical results validates the accuracy of the developed model. To demonstrate its applications, the analytical model is used to investigate the effects of packet size on the performance of the queueing system. The developed model is further applied to study the configuration of weights of individual traffic flows. Copyright © 2010 John Wiley & Sons, Ltd.

High-performance computing and communications has been an important and fundamental research topic over the past decade and has posed many challenging problems. Researchers and industrial professionals have been devoted to designing innovative tools and techniques to keep up with the rapid evolution and increasing complexity of large and complex scientific and engineering problems. This special issue contains seven extended and revised papers which were mainly selected from the 10th IEEE International Conference on High Performance Computing and Communications (HPCC-08) in Dalian, China, September 25–27, 2008. The purpose of the conference was to provide a forum for engineers and scientists in academia and industry to present and discuss their novel ideas, new research results, applications, experiences, work-in-progress, and state-of-the-art techniques in the area of high-performance computing and communications. The extended versions of these papers were carefully peer-reviewed according to the practice of this journal. The selected papers deal with a wide range of important aspects and challenging issues of high-performance computing and networking systems and the contents are built on analytical modeling, experimental and simulation studies. The contributions of these papers are outlined below. Dispatching a large number of jobs directly to a small number of physical servers requires much runtime information and computation in order to make the distributed computing system efficient and scalable. Choi et al. 1 presented a two-level indirect dispatching framework that dispatches requests to servers through virtual machines (VMs), called Dispatching Requests Indirectly through Virtual Environment (DRIVE). They set up an experimental environment consisting of a personal computer (PC) cluster and four benchmark suites to demonstrate the effectiveness of the DRIVE framework. The experimental results have shown that the use of VMs indeed abstracts away the client requests and hence helps to improve the overall performance of a dynamically changing computing environment. Virtualization is being widely deployed now as an emerging trend. Liao et al. 2 proposed a Lightweight Virtual Desktop (LVD) management architecture to combine the virtualization technology and inexpensive PCs. They implemented LVD in the cluster with VMs and compared the performance of LVD prototype system with those of the currently used desktop systems, including Microsoft Remote Desktop, Citrix MetaFrameXP and Sun Ray. The experimental results have shown that the LVD performs well in its functions improving the response time while decreasing the power consumptions. The workload characteristics and resource usage patterns of available applications are critical for the design and development of hardware and software stacks of future machines. Seelam et al. 3 presented a comprehensive workload performance characterization of three large-scale applications, i.e. Hybrid Coordinate Ocean Model, Parallel Ocean Program, and Lattice Boltzemann Magneto-Hydrodynamics Code, using timers and performance counters available on Power5+ chip. On the basis of the experimental results, they have described the sources of load imbalances in the applications and identified the potential impediments to the scalability of the applications under large processor counts. Communication through relay channels has received considerable research interests due to their ability of creating diversity and consequently improving the robustness of data transmission for ubiquitous computing and networking applications. Zhang et al. 4 presented the theoretical model and the implementation of two relaying algorithms, dynamic relaying and fixed relaying, in wireless sensor networks (WSNs) with a couple of relay nodes and investigated the performance of relay channels in terms of diversity gain and throughput. The experimental results have shown that the data throughput between the source node and the destination node is enhanced by the presence of the relay nodes. Moreover, the results have revealed the performance trade-offs between diversity and energy efficiency. Multipath routing protocols can potentially enhance the network robustness significantly to repudiation attacks. Yin et al. 5 proposed a novel Secure Multipath Routing Protocol with authentication mechanism for Mobile Ad hoc Networks (MANETs) to extensively defense the various attacks. Extensive analysis and experiments have demonstrated that the proposed schemes can effectively secure MANETs, yet incur very low communication and computation overhead. Attacks and defense are prevalent in the current Internet. The modeling of Internet traffic has been a critical task in the design of network architectures. Xiang et al. 6 presented a microscopic competition model to analyze the dynamics among the nodes, i.e. benign or malicious. The NS2 simulation results have proven that the model can well describe the competition behavior among normal users and attackers. According to this model, they proposed an anomaly attack detection method based on the Adaptive Reasons Theory (ART) neural network. The experiments have shown that the proposed model can effectively detect Distributed Denial of Service attacks. Target tracking is typically a thought-provoking application in WSNs. Wang et al. 7 proposed a two-level Cooperative and Energy-efficient Tracking (CET) algorithm that reduces energy consumption by requiring only a minimum number of sensor nodes to participate in communication, transaction and perform sensing for target tracking in WSNs. Performance analysis and simulation studies have demonstrated that CET improves target capturing speed and outperforms some existing protocols of target tracking with energy saving under certain ideal situations. We express our deep thanks to the Editor-in-Chief, Professor Geoffrey Fox, for providing us with the opportunity to host this special issue in Concurrency and Computation: Practice and Experience. We also thank the authors for their contributions, including those whose papers were not included. Last but not least, we thank the thoughtful work of the many reviewers who provided invaluable evaluations and recommendations.

The k-ary n-cube which has an n-dimensional grid structure with k nodes in each dimension has been a popular topology for interconnection networks. Analytical models for k-ary n-cubes have been widely reported under the assumptions that the message destinations are uniformly distributed over all network nodes and the message arrivals follow a non-bursty Poisson process. Recent studies have convincingly demonstrated that the traffic pattern in interconnection networks reveals the bursty nature in the both spatial domain (i.e., non-uniform distribution of message destinations) and temporal domain (i.e., bursty message arrival process). With the aim of capturing the characteristics of the realistic traffic pattern and obtaining a comprehensive understanding of the performance behaviour of interconnection networks, this paper presents a new analytical model for k-ary n-cubes in the presence of spatio-temporal bursty traffic. The accuracy of the model is validated through extensive simulation experiments of an actual system.

The increasing demand for the coverage of high-speed wireless local area networks (WLANs) is driving the installation of a very large number of access points. Wireless mesh networks (WMNs) have emerged as a promising technology in next generation networks to provide economical and scalable broadband access to wireless interconnection of multiple access points which manage individual WLANs in order to extend the coverage of conventional single WLANs. Due to various types of applications with particular purposes running at different WLANs, the traffic generated by stations in different WLANs possess a high degree of heterogeneity. To the best of our knowledge, there is hardly any analytical model reported in the current literature to handle heterogeneous network traffic in the integrated WMNs and WLANs. To fill this gap, we develop a new analytical model to investigate the Quality-of-Service (QoS) performance metrics in WMNs interconnecting multiple WLANs with heterogeneous stations. The Poisson process is employed to model the traffic of non-bursty data applications and the Markov modulated Poisson process (MMPP) is used to model the traffic of bursty multimedia applications. Extensive simulation experiments are conducted to validate the accuracy of the analytical model.

Cluster-based systems have emerged as a promising technology for providing cost-effectiveness in high-performance computing and communication systems. Performance studies on communication networks in cluster-based systems have been reported based on the simplified assumptions that the traffic follows the non-bursty Poisson process and the message destinations are uniformly distributed over all network nodes. However, the uniform distribution of message destinations is not always realistic in practice. Moreover, the communication locality, a typical example of the non-uniform destination distribution, has been shown to be an important phenomenon in the communication networks of cluster systems. Many recent measurement studies have revealed that the traffic generated by many real-world applications exhibits a high degree of burstiness. In order to have a comprehensive understanding of the system performance, this paper proposes a new analytical model for communication networks in multi-cluster systems under the bursty message arrivals with communication locality. The model is validated through extensive simulation experiments.

Providing a seamless handoff and quality of service (QoS) guarantees is one of the key issues in real-time services. Several IPv6 mobility management schemes have been proposed and can provide uninterrupted service. However, these schemes either have synchronization issues or introduce signaling overhead and high packet loss rate. This paper presents a scheme that reduces the handoff latency based on the location and movement pattern of a Mobile Node (MN). An MN can detect its movement actively and forecast the handoff, which alleviates the communication cost between the MN and its Access Router (AR). Furthermore, by setting the dynamic interval in an appropriate range, the MN’s burden can also be alleviated. Finally, by performance evaluation using both theoretical analysis and computer simulations, we show that the proposed scheme can lower the handoff latency efficiently.

To overcome the inefficiency of the cluster structure construction in the existing wireless sensor networks, we propose an equilibrium multi-hop cluster hierarchy (EMCH). Both the remaining energy of cluster head nodes and the cluster distribution are considered during the cluster construction. According to the signal intensity of communication between sensor nodes, the cluster structure is formed in a reasonable distribution. The experiment shows EMCH can balance the energy consumption and prolong the lifetime of WSNs.

It is challenging to deal with the tradeoff between energy efficiency and quality of service (QoS) for multimedia applications in wireless sensor networks (WSNs). The packet aggregation scheme is widely applied to reduce overhead for high throughput and energy saving in wireless sensor networks. However, it may cause unacceptable delay in WSNs. In order to solve the problem for voice delivering with energy efficiency and proper QoS in WSNs, in this paper, we propose an adaptive waiting packet aggregation (AWPA) algorithm. First, we formulate the problem as an optimization problem. Then we optimize the problem and implement the proposed algorithm. Furthermore, we analyze the performance of the proposed algorithm. Finally, we conduct simulations in the NS-2 platform to evaluate the performance. The results show that AWPA outperforms existing algorithms for the performance.

Selfish behaviors significantly affect the overall performance of mobile ad hoc networks (MANETs). Reputation systems have been proved to be an efficient way to block such behaviors in MANETs. Several reputation models based on subjective logic have been proposed to improve the reputation mechanism, in which an uncertainty value is introduced for reputation computation when the local information is not sufficient. However, these reputation models fail to utilize the recommended opinions effectively and reduce the uncertainty value while these opinions are combined. In this paper, we propose a novel reputation computation model based on subjective logic to overcome the above deficiencies. We consider not only the recommenders' trustworthiness but also the familiarities among the recommended nodes during reputation computations. This familiarity is defined as a certainty value which is used to weight opinions in reputation computation. In our model, the recommendations of nodes with low trustworthiness or high uncertainty on the recommended nodes have little impact on the recommended nodes' reputations so that nodes can reach opinions with lower uncertainty value through reputation computations. We conduct simulations to evaluate our model on its performance. The simulation results show that the proposed model achieves about 40% improvement in the time of discovering and isolating selfish nodes compared with a previous model based on subjective logic and selfish nodes' success rate is further reduced by up to 10%.

Wireless mesh networks (WMNs) are emerging as a key technology for the next-generation wireless networks owing to its attractive properties, such as dynamic self-organisation, quick deployment, easy maintenance, low cost, and high scalability. In this paper, we develop an analytical model to investigate the end-to-end delay in a random-access two-tier (i.e., the backhaul tier and access tier) WMN. In the backhaul tier, mesh routers are uniformly distributed in a grid placement. The access tier is formed by a series of wireless local area networks. After validating the accuracy of the analytical model through simulation experiments, we use the model to tune the system parameters in order to minimise the end-to-end delay.

The administration of resources and application scheduling in a large-scale distributed storage environment is a formidable task nowadays, especially in dynamic changing scenario both in storage resources and application requirements. In order to make a difference in application oriented storage service, this paper shows a new autonomic approach to match individual applications to most appropriate storage systems and storage services based on their particular characteristic matrix, such as capacity, performance, availability, dependability, recoverability, user requirements, scalability, and value to the organization etc.

In a wide range of applications, such as detecting environmental hazards, monitoring remote terrain and even monitoring customer behavior, wireless sensor networks (WSNs) are playing an important role. Due to the limited energy in each node, a lot of attention has been paid on improving energy efficiency. In this paper, we first propose an energy-efficient waiting caching (EEWC) algorithm to improve the data-centric (DC) transport control protocol by cutting down traffic flow. Second, we deploy networked predictive controller (NPC) to shorten network delay. Thus, the energy consumption can be reduced accordingly. We also analyze the proposed algorithm. Finally, we conduct extensive simulation experiments to evaluate the performance of the proposed algorithm.

A key problem in research on wireless sensor networks (WSNs) is how to improve energy efficiency and extend the lifetime of sensor nodes and therefore the WSNs. In this paper, we address this problem and present a geography-based heterogeneous hierarchy routing (GBHHR) protocol. GBHHR adopts hierarchy structure for better data congregation, which can improve the efficiency of data collection and uses heterogeneous nodes as cluster heads to solve the problem of selecting cluster head in traditional hierarchy routing algorithms. A dormancy mechanism is introduced to heterogeneous node for energy saving and a multi-jump data transmission mode is applied for remote and wide-area applications. Furthermore, a new geographical routing algorithm, named hole-forewarning routing algorithm (HFR), is proposed to solve the problem of hole during routing path. Simulation results show that GBHHR has higher energy efficiency[1]than traditional geographical routing and hierarchy routing.

Due to lack of effective trust management mechanism, there are a lot of deceptive behaviors in P2P networks, which seriously decrease the quality of network services. In order to improve the quality of services (or transactions), this paper proposes a novel trust and reputation model based on similarity evaluation in P2P environments. According to the different context of services, the model gives the similarity degree evaluation, and presents local reputation and global reputation computation. Experimental results demonstrate that this model can get more real trust value and deal with the malicious attacks.

The variable precision rough set (VPRS) model generalizes the Pawlak rough set model with variable parameters. The Bayesian rough set (BRS) model improves the VPRS model with non-parametric modification by using the prior probability as a reference. This paper presents two research results related to rough set model and rough vague sets. One is a modification to the Bayesian rough set model and the other is a modification to rough vague sets. First, the Bayesian rough set model is analyzed and discussed. Second, a modification to this model is proposed and verified. Finally, a modification to rough vague sets is presented and its related properties are discussed..

Two important factors that affect the performance of wireless sensor networks (WSNs) are data quality and network lifetime. This paper exploits the tradeoff between data quality and network lifetime to improve data collection precision while the network lifetime is adapted. The problem is to minimize the total error bound for approximate data aggregation in both single-hop and multi-hop WSNs to achieve the adaptive network lifetime. This problem is formulated as an optimization problem by combining the changing pattern of sensor readings, the residual energy of sensor nodes, and the communication cost from the sensor node to the base station. Our method is theoretically analyzed and further evaluated by conducting simulation experiments. To the best of our knowledge, this is the first study on minimizing the total error bound while achieving the adaptive network lifetime.

The main concern of the Internet user-base has shifted from what kind of information are available to how to find the desired information on the Internet thanks to the explosive growth of the WWW and the increasing amount of data available via the Internet. Since mobile agent technology is expected to be a promising technology for information retrieval, there are a number of mobile agent based-information retrieval approaches have been proposed in recent years. For a better understanding and efficiency improvement of these approaches, performance evaluation of great importance. However, most of existing evaluation results are experimental and there is a lack of theoretical performance analysis which is helpful to reveal the insight of the working mechanisms. In this paper, we further the study in W. Qu et al. (2007) in which some primary studies on the performance of several mobile agent-based information retrieval approaches are provided and provide the exact probability distributions of execution time for each approach. Our results reveal the insight of mobile agent-based information retrieval approaches and our analytical method provides a useful tool for further research to information retrieval.

Multimedia object caching, by which the same multimedia object can be adapted to diverse mobile appliances through the technique of transcoding, is an important technology for improving the scalability of Web services, especially in the environment of mobile networks. In this paper, we address the cache replacement problem for multimedia object caching by exploring the aggregate effect of caching multiple versions of the same multimedia object. First, we present an optimal solution for calculating the minimal access cost of caching multiple versions of the same multimedia object. Second, based on this solution, we propose an effective cache replacement algorithm for multimedia object caching. Finally, we evaluate the performance of the proposed solution with a set of simulation experiments for various performance metrics over a wide range of system parameters.

Transparent data replication is a promising technique for improving the system performance of a large distributed network. Transcoding is an important technology which adapts the same multimedia object to diverse mobile appliances; thus, users' requests for a specified version of a multimedia object could be served by a more detailed version cached according to transcoding. Therefore, it is particularly of theoretical and practical necessity to determine the proper version to be cached at each node such that the specified objective is achieved. In this paper, we address the problem of multimedia object placement for transparent data replication. The performance objective is to minimize the total access cost by considering both transmission cost and transcoding cost. We present optimal solutions for different cases for this problem. The performance of the proposed solutions is evaluated with a set of carefully designed simulation experiments for various performance metrics over a wide range of system parameters. The simulation results show that our solution consistently and significantly outperforms comparison solutions in terms of all the performance metrics considered

In this paper, we address present an optimal solution for the problem of multimedia object placement for hybrid transparent data replication. The performance objective is to minimize the total access cost by considering both transmission cost and transcoding cost. The performance of the proposed solution is evaluated with a set of carefully designed simulation experiments for various performance metrics over a wide range of system parameters. The simulation results show that our solution consistently and significantly outperforms comparison solutions in terms of all the performance metrics considered.

Transcoding is an important technology which adapts the same multimedia object to diverse mobile appliances; thus, users' requests for a specified version of a multimedia object could be served by a more detailed version cached according to transcoding. Therefore, it is of particularly theoretical and practical necessity to determine the proper versions to be cached at a node such that the specified objective is achieved. In this paper, we address the problem of multimedia object placement. The performance objective is to minimize the total access cost by considering both transmission cost and transcoding cost. We present an optimal dynamic programming-based solution for this problem. The performance of the proposed solutions is evaluated with a set of carefully designed simulation experiments for various performance metrics over a wide range of system parameters. The simulation results show that our solution consistently and significantly outperforms comparison solutions in terms of all the performance metrics considered.

Uncertain information will often exist in a complicated system which may result from many reasons. The accuracy of the conclusion will definitely be influenced without taking this uncertain information into consideration in the process of decision-making. By considering these uncertain factors, the solution proposed in the essay to multi-objective fuzzy optimization dynamic programming makes itself more universal.

Research on non-structural system group decision-making problems largely depends on the knowledge and experience of the experts for tactical analysis. The usual method of voting may result in a great loss of information in case of much renunciation, and the accuracy of voting can therefore be directly influenced.This paper proposes a new solution to non-structural system group decision-making problems by taking advantage of the characteristics of correlate and the information easy to lose to make an overall analysis of the ayes, blackballs and renunciation polls for an accurate result.

In this paper, we address the problem of cache replacement for transcoding proxy caching. First, an efficient cache replacement algorithm is proposed. Our algorithm considers both the aggregate effect of caching multiple versions of the same multimedia object and cache consistency. Second, a complexity analysis is presented to show the efficiency of our algorithm. Finally, some preliminary simulation experiments are conducted to compare the performance of our algorithm with some existing algorithms. The results show that our algorithm outperforms others in terms of the various performance metrics.

Web caching is an important technology for improving the scalability of Web services. One of the key problems in coordinated enroute Web caching is to compute the locations for storing copies of an object among the enroute caches so that some specified objectives are achieved. In this article, we address this problem for tree networks, and formulate it as a maximization problem. We consider this problem for both unconstrained and constrained cases. The constrained case includes constraints on the cost gain per node and on the number of object copies to be placed. We present dynamic programming-based solutions to this problem for different cases and theoretically show that the solutions are either optimal or convergent to optimal solutions. We derive efficient algorithms that produce these solutions. Based on our mathematical model, we also present a solution to coordinated enroute Web caching for autonomous systems as a natural extension of the solution for tree networks. We implement our algorithms and evaluate our model on different performance metrics through extensive simulation experiments. The implementation results show that our methods outperform the existing algorithms of either coordinated enroute Web caching for linear topology or object placement (replacement) at individual nodes only.

Transcoding is a promising technology that allows systems to effect a quality-versus-size tradeoff on multimedia objects. As audio and video applications have proliferated on the Internet, caching in transcoding proxies has become an important technique for improving network performance, especially in mobile networks. This article addresses the problem of coordinated enroute multimedia object caching in transcoding proxies for tree networks. We formulate this problem as an optimization problem based on our proposed model, in which multimedia object caching decisions are made on all enroute caches along the routing path by integrating both object placement and replacement policies and cache status information along the routing path of a request is used to determine the optimal locations for caching multiple versions of the same multimedia object. We propose an optimal solution using dynamic programming to compute the optimal locations. We also extend this solution to solve the same problem for several constrained cases, including constraints on the cost gain per node and on the number of versions to be placed. Our model is evaluated on different performance metrics through extensive simulation experiments. The implementation results show that our model significantly outperforms existing models that consider Web caching in transcoding proxies either on a single path or at individual nodes.

Web caching is an important technique for reducing web traffic, user access latency, and server load and cache replacement plays an important role in the functionality of web caching. In this paper we propose an improved GreedyDual (GD) cache document replacement algorithm, which considers update frequency as a factor in its utility function. We use both trace data and statistical data to simulate our proposed algorithm. The experimental results show that our improved GD algorithm can outperform the existing GD algorithm over the performance metrics considered.