Federated fine-tuning of Mixture-of-Experts (MoE)-based large language models (LLMs) is challenging due to their massive computational requirements and the resource constraints of participants. Existing works attempt to fill this gap through model quantization, computation offloading, or expert pruning. However, they cannot achieve desired performance due to impractical system assumptions and a lack of consideration for MoE-specific characteristics. In this paper, we propose Flux, a system designed to enable federated fine-tuning of MoE-based LLMs across participants with constrained computing resources (e.g., consumer-grade GPUs), aiming to minimize time-to-accuracy. Flux introduces three key innovations: (1) quantization-based local profiling to estimate expert activation with minimal overhead, (2) adaptive layer-aware expert merging to reduce resource consumption while preserving accuracy, and (3) dynamic expert role assignment using an exploration-exploitation strategy to balance tuning and non-tuning experts. Extensive experiments on LLaMA-MoE and DeepSeek-MoE with multiple benchmark datasets demonstrate that Flux significantly outperforms existing methods, achieving up to 4.75× speedup in time-to-accuracy.

Multimodal learning enables models to process and reason over diverse information sources, unlocking human-like perceptual and cognitive capabilities. As such models gain adoption, efficiently serving them on GPUs has become increasingly important. However, the modular architecture of multimodal models poses significant challenges to existing unimodal serving systems, which treat models as monolithic and overlook inter-module heterogeneity. This results in severe GPU underutilization. To address this, we propose Eevee, a multimodal serving system based on a new scheduling paradigm we call module multiplexing. Unlike prior approaches that execute all modules sequentially with uniform batch sizes, Eevee schedules modality-specific modules concurrently on the same GPU with independently tuned batching and resource allocation. This design enables fine-grained GPU sharing, boosting intra-GPU parallelism and improving request-level throughput. We implement a prototype of Eevee and evaluate it on several representative multimodal models (e.g., CLIP, BLIP, LLaVA, InternVL). Our results show that Eevee significantly outperforms state-of-the-art serving systems in both throughput and GPU utilization.

Expert parallelism has become the prevailing paradigm to serve Mixture-of-Experts (MoE) models. Its efficiency depends on the communication and computation latencies of the GPUs, which are linked to the placement of experts in the GPUs. Existing works for optimizing expert placement focus on leveraging past requests' expert activation patterns. However, they demonstrate deficiencies facing diverse and rapidly changing request patterns, calling for an online, proactive approach. Implementing such an approach requires addressing several challenges: the uncertainty associated with incoming requests' expert activation, the cost of expert migration, and the NP-hard complexity in optimization. Therefore, we present Director, a new distributed MoE serving system that minimizes end-to-end latency via prediction-driven, online expert placement. Director uses either a lightweight cascaded predictor or a low-bit quantized replica for expert activation patterns of incoming requests. An online migration module then enacts the changes with near-zero downtime by executing migrations in compute-bound phases, keeping disruption bounded. At its core, a relaxation-based expert placement optimizer operates under capacity constraints, runs in polynomial time, and achieves a (1+ ϵ) approximation ratio. Finally, we implement a prototype and demonstrate, through extensive experiments, a reduction in end-to-end latency of 11 ~ 55% for popular MoE models (e.g., Mistral, DeepSeek and Qwen) compared to existing work.

Federated learning, as a compelling machine learning framework, enables collaborative model training without exposing private data. However, the excessive communication overhead remains a major challenge. To tackle this challenge, hierarchical federated edge learning (HFEL) framework has been proposed for reducing the communication load via migrating the model aggregation partially from cloud to edge servers. Although HFEL has significant potential, it is still constrained by end-devices with limited computational capabilities and unfavorable network conditions. A common approach to reduce this effect is to involve only the fastest end-devices in the training process. But because only parts of end-devices' data samples can be selected by such means, it damages the diversity of training data, and hence greatly affects the model's quality. In addition, for further improving the training performance, a proper bandwidth allocation strategy is also needed to make full use of the shared network resource of edge servers. To this end, we proposeDAHFF, aDiversity-AwareHierarchicalFastFederated learning framework consisting ofVirtual Queue based Device Selectionphase andBinary Search based Bandwidth Allocation, which are responsible for selecting participated end-devices and allocating bandwidth for selected devices, respectively. Extensive experiments on different deep learning models show that our proposed framework can averagely speed up the training performance by$2.07\times$in comparison with state-of-the-art approaches.

The decentralized nature of federated learning (FL) makes it difficult to verify the trustworthiness of participating clients, creating an opportunity for backdoor attacks. This paper addresses a general backdoor-resilient decentralized FL problem without any prior knowledge of the type of backdoor attacks or information about malicious clients. After an in-depth investigation of how backdoor attacks are conducted in FL, we introduce a multi-armed bandit-based knowledge distillation approach to help benign clients learn useful knowledge from other clients while rejecting potential backdoors hidden in shared updates. Unlike most previous works that rely on identifying and removing malicious updates—an approach limited to scenarios with fewer than 50% attackers—our knowledge distillation technique enables benign clients to reject backdoored knowledge while preserving useful information, maintaining effective defense even when malicious clients exceed 50% of the population. Additionally, to handle the various updates from clients with Non-IID dataset, a multi-armed bandit scheme is designed for each benign client to select the most appropriate teachers for knowledge distillation, resulting in high accuracy and fast convergence. Extensive experiments demonstrate that our multi-armed bandit-based knowledge distillation approach achieves high accuracy and general backdoor resilience. Comparisons with previous works show that our approach can reduce the attack success rate by 14.71%∼96.78% on average.

This paper studies an efficient federated learning (FL) problem involving multiple edge-based clients with heterogeneous constrained resources. Compared with numerous training parameters, the computing and communication resources of clients in edge scenarios are usually insufficient for fast local training and real-time knowledge sharing. Besides, training on clients with heterogeneous resources may result in the straggler problem, which delays the convergence of FL. To address these issues, we proposeFed-RAA: aResource-AdaptiveAsynchronousFederated learning algorithm. Different from vanilla FL methods, where all parameters are trained by each participating client regardless of resource diversity, Fed-RAA adaptively allocates submodels of the global model to clients based on their computing and communication capabilities. Each client then individually trains its assigned submodel and asynchronously uploads the updated result. Theoretical analysis confirms the convergence of our approach. Additionally, an online greedy-based algorithm is designed for asynchronous submodel assignment in Fed-RAA, improving the convergence of Fed-RAA by optimal minimization on the training delay bound of submodels. Compared to state-of-the-art methods, our Fed-RAA algorithm reduces the time required to achieve the target accuracy by an average of$ 30.89\%$, demonstrating its superior efficiency on heterogeneous constrained computing and communication resources. To the best of our knowledge, this paper is the first resource-adaptive asynchronous method for submodel-based FL with guaranteed theoretical convergence.

The growing resource demands of large-scale transformer models pose significant challenges for resource-constrained users, particularly in distributed environments. To address this issue, we propose a federated learning framework called Fed-Grow, which enables multiple participants to collaboratively learn a lightweight scaling operation that transfers knowledge from pretrained small models to a large transformer model. In Fed-Grow, we introduce the Dual-LiGO (Dual Linear Growth Operator) architecture, consisting of Local-LiGO and Global-LiGO components. Local-LiGO addresses model heterogeneity by adapting each participant's pre-trained model to a common intermediate form, while Global-LiGO facilitates knowledge sharing across participants without sharing local models or raw data, ensuring privacy preservation. This federated approach offers a scalable solution for growing large transformers in a distributed manner, where only the Global-LiGO is shared, significantly reducing communication overhead while maintaining comparable model performance under the same communication constraints. Experimental results demonstrate that Fed-Grow outperforms state-of-the-art methods in terms of accuracy and precision, while reducing the number of trainable parameters by 59.25% and communication costs by 73.01%. These improvements allow for higher efficiency in training large models in distributed environments, without sacrificing performance. To the best of our knowledge, Fed-Grow is the first method that enables cooperative transformer scaling in a distributed setting, making it a practical solution for resource-constrained users.

With the widespread popularity of Large Language Models (LLMs), the mixture of experts (MoE) has not only emerged as a key enabler for scaling up model capacity by significantly reducing computational demands, but also for giving rise to edge-computing empowered distributed LLMs with better prices, low latency, and regional privacy. Nonetheless, besides constraints in computational capability, edge-based LLM deployments also face challenges such as unreliable environments due to the limited security of edge devices. In this paper, we propose REMIS, an inference task scheduling scheme designed to enable MoE-based LLM services under untrustworthy computation conditions. Specifically, after an LLM is properly partitioned into shards and deployed across edge devices, REMIS dynamically schedules and activates experts on devices with lower loads and higher reliability. This strategy is effectively achieved through a deep reinforcement learning procedure that optimizes both servicing latency and inference credibility. Unlike most existing MoE-based schemes with fixed Top-K routing, REMIS operates in a novel plug-in manner, intelligently selecting experts to improve task adaptability. Numerical evaluations under various untrustworthy setups validate the superiority of our proposed scheme in both servicing latency and inference credibility.

Abstract Machine unlearning in the context of cybersecurity and privacy protection facilitates the removal of specific training data impacts from deep learning (DL) models, adhering to security, privacy, or compliance demands. However, traditional methods can only handle short-term, independent unlearning tasks. Conversely, real-world scenarios often involve extensive unlearning demands from users. Current methods fail to adequately address these demands due to substantial computational overhead and adverse impacts on inference accuracy, leaving the security and privacy of many users at risk. To navigate these challenges adeptly, we introduce the Multi-Agent Reinforcement Learning Data Lifecycle Management (MADLM) strategy. MADLM intricately examines the interactions between unlearning and continuous learning processes, enabling the postponement of certain tasks for combined execution to optimize computational resources. Concurrently, it employs strategic data management to maintain and enhance inference accuracy. Furthermore, by utilizing Multi-Agent Reinforcement Learning (MARL), MADLM dynamically orchestrates task scheduling to minimize computational demands, improve task response times, and bolster inference reliability, crucial for upholding stringent cybersecurity and privacy standards. Our evaluations of MADLM reveal substantial enhancements, including a 6% uplift in inference accuracy and a dramatic reduction in computational overhead to merely 12% of the original demands, effectively expanding the data security protections.

Presents corrections to the paper, (Corrections to “Giant Could Be Tiny: Efficient Inference of Giant Models on Resource-Constrained UAVs”).

Within the Internet of Things (IoT) landscape, Encoder as a Service (EaaS) is a cloud-based service for many AI empowered scenarios, such as autopilot and face-scan payment, enabling IoT devices to keep a light classifier model at local while remotely accessing powerful encoding models. However, adversarial examples like an image with imperceptible tiny perturbations that can lead to incorrect classifying results, make it challenging to achieve a robust EaaS-based classifier. Certified radius (R) emerges as a measure against such imperceptible tiny perturbations, and guarantees the trustworthiness of any input with perturbations smaller than R. Despite offering R related services, the substantial computational overhead from traditional methods, stemming from the extensive searches of R for each request, poses a barrier to their practical deployment. In this article, we identify the repetitive computations in the EaaS framework due to the fixed encoding model and propose a novel search cache scheme to speed up the R-related computation. Therefore, we propose large-scale efficient robust EaaS (ScaleRES) to address different R-related services: First, ScaleRES strategically stores previous computation results of R, to enable the reuse and refining of R across users. Then, ScaleRES obtains the average certified radius (ACR) efficiently with selective cached R. Finally, ScaleRES performs R filtering to enhance adversarial training for a robust EaaS-based classifier. Comprehensive evaluations demonstrate that in three distinct computations, ScaleRES offers significant savings in computational overhead compared to the conventional approach—70% for R computation as the number of clients increases, 35% for ACR of the entire test set, and 40% for adversarial training with robustness comparable to other works.

The explosion of cameras embedded in IoT devices—from mobile phones to autonomous vehicles—has positioned video analytics as a transformative AI tool across healthcare, smart cities, and beyond. Yet, the substantial computing and bandwidth demands of these applications outstrip what IoT devices alone can handle, particularly when low latency is required. Computing Power Networking (CPN) is an emerging solution that unifies cloud, edge, and device resources, enabling seamless, efficient task distribution for real-time analytics. While recent advances in cloud-edge frameworks show promise, current approaches often neglect the economic incentives that drive resource availability. To address this, we present a novel, privacy-enabled dual-layer Stackelberg game model that establishes a dynamic pricing strategy for video analytics in CPN. Our model introduces a two-stage negotiation: IoT devices contract with edge servers for computational and bandwidth resources, while edge servers may offload tasks to the cloud for enhanced service. Using game theory, we derive optimal pricing and offloading strategies under both complete and incomplete information, proving a Nash equilibrium. Comprehensive simulations validate our approach, showing improvements in resource efficiency, reduced latency, and incentivized resource-sharing across all CPN tiers. Specifically, our hybrid offloading strategy significantly reduces latency compared to edge-only and cloud-only computation models. For varying IoT device quantities, the average latency reduction across all scenarios is approximately 30.5%. This work provides an economically sustainable, privacy-conscious solution to the computational challenges of video analytics in an interconnected, resource-sharing ecosystem.

Graph neural networks (GNNs) have shown great success in IoT applications, but many IoT scenarios further involve evolving graph data over time. Streaming graph learning (SGL) tackles the issue by updating GNN models continuously, incorporating significant historical node data for tasks like node classification. However, existing research on SGL often neglects memory limitations during historical node selection, and efficient distributed training is challenging due to the coupling of node selection, placement, and parallelization. This article proposes an efficient distributed SGL system that optimizes node selection and storage in multi-GPU environments, while reducing communication overhead to accelerate training. Extensive evaluations demonstrate the effectiveness of the approach.

Hierarchical federated learning (HFL) extends traditional federated learning by introducing a cloud-edge-device framework to enhance scalability. However, the challenge of determining when devices and edges should aggregate models remains unresolved, making the design of an effective synchronization scheme crucial. Additionally, the heterogeneity in computing and communication capabilities, coupled with non-independent and identically distributed ( non-IID) data distributions, makes synchronization particularly complex. In this paper, we proposeRobin, a learning-based synchronization scheme for HFL systems. By collecting data such as models' parameters, CPU usage, communication time,etc., we design a deep reinforcement learning-based approach to decide the frequencies of cloud aggregation and edge aggregation, respectively. The proposed scheme well considers device heterogeneity, non-IID data and device mobility, to maximize the training model accuracy while minimizing the energy overhead. Meanwhile, we prove the convergence ofRobin's synchronization scheme. And we build an HFL testbed and conduct the experiments with real data obtained from Raspberry Pi and Alibaba Cloud. Extensive experiments under various settings are conducted to confirm the effectiveness ofRobin, which can improve 31.2% in model accuracy while reducing energy consumption by 36.4%.

Mixture-of-Experts (MoE) is an emerging technique for scaling large models with sparse activation. MoE models are typically trained in a distributed manner with anexpert parallelismscheme, where experts in each MoE layer are distributed across multiple GPUs. However, the default expert parallelism suffers from the heavy network burden due to the all-to-all intermediate data exchange among GPUs before and after the expert run. Some existing works have proposed to reduce intermediate data exchanges by transferring experts to reduce the network loads, however, which would decrease parallelism level of expert execution and make computation inefficient. The weaknesses of existing works motivate us to explore whether it is possible to reduce inter-GPU traffic while maintaining a high degree of expert parallelism. This paper gives a positive response by presentingLuffy, a communication-efficient distributed MoE training system with two new techniques. First,Luffymigrates sequences among GPUs to hide heavy token pulling paths within GPUs and avoid copying experts over GPUs. Second, we propose token condensation that identifies similar tokens and then eliminates redundant transmissions. We implementLuffybased on PyTorch and evaluate its performance on a testbed of 16 V100 GPUs.Luffysystem can achieve a speedup of up to$2.73\times $compared to state-of-the-art MoE training systems.

The need for an adaptive congestion control (CC) service is crucial due to the heterogeneity of systems and the diversity of applications. Traditional CC methods often fail to adaptively balance throughput and delay, struggling to meet the varied demands of different network applications. In this work, we introduceAuto, a novel CC service that employs Multi-Objective Reinforcement Learning (MORL) to transcend these limitations. Unlike conventional approaches,Autooptimizes policies within a single model to cater to all potential preferences for balancing throughput and delay, making it ideal for diverse and heterogeneous network environments. To enhance operational transparency, we developed an interpretation algorithm that translates MORL into a human- readable decision tree, essential for service computing where clarity and interpretability are crucial. Furthermore,Autoallows users to explicitly set flow priorities and target sending rates, meeting varied application demands. Our extensive evaluations show thatAutonot only consistently outperforms existing CC methods in diverse network conditions but also exhibits robustness to stochastic packet loss and rapid network changes. These capabilities establishAutoas a pioneering solution for next-generation congestion control in networking services.

Transformers have dominated the field of natural language processing, attributed to their capability to handle sequential input data. There is a surge of work on computational and networking optimizations, aimed at improving the training efficiency of Transformers. However, transformer inference, a cornerstone of myriad AI services, remains relatively underexplored. With the challenge of variable-length inputs, conventional methods adopt padding schemes, resulting in computational waste. Moreover, works on transformer inference often overlook the integration between request scheduling and batching, which play pivotal roles in inference systems. To address these challenges, we introduce TCB, a comprehensiveTransformer inference system that integrates aConcatBatching scheme to reduce computational redundancy by concatenating requests. In addition, we present an online request batching algorithm, designed to augment the throughput of scheduled requests. Consider a muiti-server case, we further introduce a joint request assignment and batching scheduling policy to fully utilize resources on servers while ensuring quality-of-service of inference. Extensive experiments demonstrate that our proposed methods can significantly outperform existing works.

Cloud-edge collaborative computing paradigm is a promising solution to high-resolution video analytics systems. The key lies in reducing redundant data and managing fluctuating inference workloads effectively. Previous work has focused on extracting regions of interest (RoIs) from videos and transmitting them to the cloud for processing. However, a naive Infrastructure as a Service (IaaS) resource configuration falls short in handling highly fluctuating workloads, leading to violations of Service Level Objectives (SLOs) and inefficient resource utilization. Besides, these methods neglect the potential benefits of RoIs batching to leverage parallel processing. In this work, we introduce Tangram, an efficient serverless cloud-edge video analytics system fully optimized for both communication and computation. Tangram adaptively aligns the RoIs into patches and transmits them to the scheduler in the cloud. The system employs a unique “stitching” method to batch the patches with various sizes from the edge cameras. Additionally, we develop an online SLO-aware batching algorithm that judiciously determines the optimal invoking time of the serverless function. Experiments on our prototype reveal that Tangram can reduce bandwidth consumption and computation cost up to 74.30 % and 66.35 %, respectively, while maintaining SLO violations within 5 % and the accuracy loss negligible.

Due to its decentralized framework and outdoor environments, federated edge learning (FEEL) faces significant vulnerability to malicious attacks within edge networks. Prevailing FEEL approaches typically hinge on a dependable parameter server (PS) to contend with the adversarial updates from Byzantine clients. Recognizing the inherent unreliability of PSs in edge networks, this paper delves into the security challenges of FEEL, specifically addressing Byzantine PSs. We present a Byzantine fault-tolerant FEEL algorithm, named Fed-MS, in which a multi-server technique along with a newly designed trimmed-mean-based model filter is employed. This combination ensures that each client can obtain a feasible global model for its local training, closely approximating a true model aggregated by benign PSs. Furthermore, we propose a sparse uploading strategy in Fed-MS to enhance communication efficiency for model aggregation to multiple PSs. Theoretical analysis demonstrates that, when Byzantine PSs are a minority, Fed-MS achieves an expected convergence speed of$O(1/T)$with$T$defined as the number of training rounds, akin to state-of-the-art works under non-Byzantine settings. Extensive experiments are conducted on the CIFAR-10 dataset with MobileNet V2 as the training model. The numerical results show that our Fed-MS can improve the model accuracy from 10% to at least 76% under the malicious attacks from Byzantine PSs. Our code is released at https://github.com/haoma2772/Fed-MS.

Crowdsourcing has become a popular paradigm for collecting large-scale labeled datasets by leveraging numerous annotators. However, these annotators often provide noisy labels due to varying expertise. Truth inference aims to infer accurate consensus labels from noisy crowdsourced annotations. Existing approaches rely heavily on hand-engineered assumptions or ground truth data, limiting their applicability. To address this, we propose GOVERN, a graph contrastive learning framework for truth inference without such external supervision. GOVERN employs a novel graph data augmentation strategy to generate views capturing worker coordination patterns. A contrastive objective then encourages invariant representations across views, enabling the discovery of features related to the hidden consensus. Further, a label correction method based on k-nearest neighbors refines noisy pseudo-labels to supervise model training. Comprehensive experiments on 9 real-world datasets demonstrate that GOVERN outperforms state-of-the-art truth inference techniques.

Fine-tuning the learnable prompt for a pre-trained vision-language model (VLM), such as CLIP, has demonstrated exceptional efficiency in adapting to a broad range of downstream tasks. Existing prompt tuning methods for VLMs do not distinguish spurious features introduced by biased training data from invariant features, and employ a uniform alignment process when adapting to unseen target domains. This can impair the cross-modal feature alignment when the testing data significantly deviate from the distribution of the training data, resulting in a poor out-of-distribution (OOD) generalization performance. In this paper, we reveal that the prompt tuning failure in such OOD scenarios can be attribute to the undesired alignment between the textual and the spurious feature. As a solution, we propose **CoOPood**, a fine-grained prompt tuning method that can discern the causal features and deliberately align the text modality with the invariant feature. Specifically, we design two independent contrastive phases using two lightweight projection layers during the alignment, each with different objectives: 1) pulling the text embedding closer to invariant image embedding and 2) pushing text embedding away from spurious image embedding. We have illustrated that **CoOPood** can serve as a general framework for VLMs and can be seamlessly integrated with existing prompt tuning methods. Extensive experiments on various OOD datasets demonstrate the performance superiority over state-of-the-art methods.

With the quick evolution of giant models, the paradigm of pre-training models and then fine-tuning them for downstream tasks has become increasingly popular. The adapter has been recognized as an efficient fine-tuning technique and attracts much research attention. However, adapter-based fine-tuning still faces the challenge of lacking sufficient data. Federated fine-tuning has been recently proposed to fill this gap, but existing solutions suffer from a serious scalability issue, and they are inflexible in handling dynamic edge environments. In this paper, we propose Tomtit, a hierarchical federated fine-tuning system that can significantly accelerate fine-tuning and improve the energy efficiency of devices. Via extensive empirical study, we find that model synchronization schemes (i.e., when edge servers and devices should synchronize their models) play a critical role in federated fine-tuning. The core of Tomtit is a distributed design that allows each edge and device to have a unique synchronization scheme with respect to their heterogeneity in model structure, data distribution and computing capability. Furthermore, we provide a theoretical guarantee about the convergence of Tomtit. Finally, we develop a prototype of Tomtit and evaluate it on a testbed. Experimental results show that it can significantly outperform the state-of-the-art.

To address the online learning problem in distributed systems, Streaming Federated learning (SFL) enables immediate model training by clients upon collecting new data, finding wide applications in AI-enabled Internet-of-Things and sensor networks. Given the variability in data distribution across different historical periods, the ability to recall and rapidly apply previously encountered data distributions significantly enhances the efficiency and accuracy of model training. In this paper, a demo based on the real-world temperature datasets is presented to demonstrate the importance of history knowledge in local training and the federating process of SFL, which also shows that vanilla federated learning without considering the history knowledge may even be harmful to model training. Observing this, we propose Fed-HIST, a Federated learning framework that enables the clients to learn from the HISTory knowledge of the whole distributed learning system. Unlike direct raw data storage, Fed-HIST employs model architectures to capture the data distributions, offering a more space-efficient and privacy-preserving method of knowledge storage on a server pool. Additionally, a model similarity comparison scheme is designed to retrieve beneficial knowledge from the pool uploaded by the clients in the past. Such a history-aware federation can enhance the efficiency of training each client, only requiring the recurrence of similar data distributions among SFL participants. We validate our framework through extensive simulations on MNIST, Fashion-MINST, CIFAR10, and CIFAR100 datasets, benchmarking against 9 baselines and highlighting the importance of federating from history in SFL problem through necessary ablation studies.

Giant models, characterized by their billions or even trillions of parameters, has demonstrated unprecedented capabilities in handling complex tasks on Artificial intelligence (AI)-driven UAVs, such as disaster relief, aerial navigation, and manipulation. However, there is an open challenge about the mismatching between the massive computation and memory requirements of giant models and the limited resources on UAVs. Existing works either pose privacy concerns with offloading methods or compromise model accuracy with various model compression techniques. In this paper, we fill the gap by exploiting the Mixture-of-Expert (MoE) model architecture that decouples giant models into multiple tiny experts, so that UAVs can dynamically load a few experts that best match their current input. We consider a general scenario of several edge servers feeding experts to multiple UVAs and formulate a core problem of expert selection and UAV-edge association. Due to the high complexity of this problem, we propose a solution, termed GESolver, based on graph learning, which automatically solves the problem by learning the complicated interaction between edge servers, UAVs, as well as their required experts. We evaluate our proposed method with three popular MoE-based models under various problem settings. The experiments demonstrate that our proposed method can significantly outperform other baselines.

Thanks to the capability of fine-grained resource allocation and fast task scheduling, serverless computing has been adopted into edge cloud to accommodate various applications, e.g., deep neural network (DNN) inference for Artificial Intelligence of Things (AIoT). In serverless edge cloud, the servers are started up on-demand. However, as a container-based architecture, the inherent sequential startup feature of container imposes high affection on the DNN inference performance in serverless edge clouds. In this article, we investigate the distributed DNN inference problem in serverless edge cloud with the consideration of such characteristics, aiming to eliminate the extra container startup time cost to minimize the DNN inference latency. We formulate this problem into a nonlinear optimization form and then linearize it into an integer programming problem, which is proved as NP-hard. To tackle the computation complexity, we propose a priority-based layer scheduling (PLAYS) algorithm. Extensive experiment results verify the effectiveness and the adaptability of our PLAYS algorithm in comparison with other state-of-art algorithms under several well known DNN models.

Computing Power Networking (CPN) represents a transformative paradigm in distributed computing, harnessing the collective capabilities of edge servers dispersed across diverse geographical locations. CPN’s core strengths lie in its ability to accelerate data processing, diminish latency, and scale efficiently, rendering it particularly apt for real-time applications and the Internet of Things. When coupled with blockchain technology, CPN extends its potential by facilitating secure and transparent allocation and trading of computing resources, bolstering data integrity and reliability. However, current research at the intersection of CPN and blockchain primarily focuses on framework development and technology integration, often overlooking the challenge of delivering dependable computing services, especially in the presence of potentially unreliable nodes. To tackle this issue, we introduce a reputation-enhanced resource trading framework, designed to ensure equitable and trustworthy computing power transactions. We establish a decentralized reputation model, capable of accurately assessing node behavior over extended periods. Additionally, we present three optimization mechanisms for reputation updates, accounting for transaction history, quality of service, and transaction amount. Furthermore, our work introduces a reputation-enhanced consensus mechanism within the trading system, strategically employing incentives to motivate participants to deliver high-quality services, thereby increasing their rewards. Simultaneously, it effectively mitigates wealth inequality among resource providers of varying sizes. To validate our approach, we develop a prototype system and conduct performance evaluations, which affirm the superiority of our system in enhancing reputation and delivering robust economic features.

Metaverse, a hypothetical digital environment linking the cyber world and the physical world, is expected to revolutionize the way people interact. In the metaverse, people interact with objects, the environment, and each other through digital representations of themselves or avatars across time and space. For example, in the metaverse, people can have meetings with colleagues hundreds of miles away. They can also walk through the aisles of a store, find the best fit and have it delivered to their doorstep. It is also possible to simulate the optimal process manufacturing line to adjust for product variation and minimize bottlenecks, or test an innovative aircraft wing design without building expensive prototypes.

Metaverse, a hypothetical digital environment linking the cyber world and the physical world, is expected to revolutionize the way people interact. In the metaverse, people interact with objects, the environment, and each other through digital representations of themselves or avatars across time and space. For example, in the metaverse, people can have meetings with colleagues hundreds of miles away. They can also walk through the aisles of a store, find the best fit, and have it delivered to their doorstep. It is also possible to simulate the optimal process manufacturing line to adjust for product variation and minimize bottlenecks, or test an innovative aircraft wing design without building expensive prototypes.

Federated edge learning (FEEL) is a novel paradigm that enables privacy-preserving and distributed machine learning on end devices. However, FEEL faces challenges from data/system heterogeneity among the participating clients and resource constraints of edge networks, which affect the efficiency and accuracy of the learning process. In this paper, we propose a comprehensive framework for client selection in FEEL based on the concept of Value-of-Information (VoI), which measures how valuable a client is for the global model aggregation. Our framework consists of two independent components: a VoI estimator that uses reinforcement learning to learn the relationship between VoI and various heterogeneous factors of clients; and a greedy client selector that chooses the most valuable clients under network resource constraints. Compared with most of the previous works that use concrete criteria to evaluate and select heterogeneous clients, our VoI-based approach is more comprehensive. Extensive experiments on different datasets and learning tasks are conducted, which show that our framework outperforms several state-of-the-art methods in terms of accuracy.

Distributed machine learning (DML) has shown great promise in accelerating model training on multiple GPUs. To increase GPU utilization, a common practice is to let multiple learning jobs share GPU clusters, where the most fundamental and critical challenge is how to efficiently schedule these jobs on GPUs. However, existing works about DML job scheduling are constrained to settings with homogeneous GPUs. GPU heterogeneity is common in practice, but its influence on multiple DML job scheduling has been seldom studied. Moreover, DML jobs have internal structures that contain great parallelism potentials, which have not yet been fully exploited in the heterogeneous computing environment. In this paper, we proposeHare, a DML job scheduler that exploits both inter-job and intra-job parallelism in a heterogeneous GPU cluster.Hareadopts a relaxed fixed-scale synchronization scheme that allows independent tasks to be flexibly scheduled within a training round. Given full knowledge of job arrival time and sizes, we propose a fast heuristic algorithm to minimize the average job completion time and derive its theoretical bound is derived. Without prior knowledge of jobs, we propose an online algorithm based on the Heterogeneity-aware Least-Attained Service (HLAS) policy. We evaluateHareusing a small-scale testbed and a trace-driven simulator. The results show that it can outperform the state-of-the-art, achieving a performance improvement of about 2.94×.

Graph neural networks (GNNs) have attracted significant research attention because of their impressive capability in dealing with graph-structure data, such as energy networks, that are crucial for sustainable computing. We find that the communication of data loading from main memory to GPUs is the main bottleneck of GNN inference because of redundant data loading. In this paper, we propose RAIN, an efficient GNN inference system for graph learning. There are two key designs. First, we explore the opportunity of conducting similar inference batches sequentially and reusing repeated nodes among adjacent batches to reduce redundant data loading. This method requires reordering the batches based on their similarity. However, comparing the similarity across a large number of inference batches is a difficult task with a high computational cost. Thus, we propose a local sensitive hash (LSH)-based clustering scheme to group similar batches together quickly without pair-wise comparison. Second, RAIN contains an efficient adaptive sampling strategy, allowing users to sample target nodes’ neighbors according to their degree. The number of sampled neighbors is proportional to the size of the node's degree. We conduct extensive experiments with various baselines. RAIN can achieve up to 6.8X acceleration, and the accuracy decrease is smaller than 0.1%.

With the widespread utilization of intelligent devices, massive mobile users' needs for rich multimedia services bring serious challenges in the aspects of network traffic, energy consumption and carbon emission. How to realize green content distribution by optimizing resource allocation is an urgent problem to solve in complex and dynamic networks. In this paper, we design a cross-layer cooperative scheme to promote energy efficiency in non-orthogonal multiple access (NOMA)-assisted cloud-edge-side environments. To be specific, we formulate the joint optimization issue of computation, caching and communication resources as an energy minimization model while considering request aggregation. Next, we propose a new deep reinforcement learning (DRL)-based task offloading strategy to minimize energy consumption by making optimal resource allocation decisions according to content request history and resource availability. Simulation results show that the proposed solution has better performance than current typical strategies in cloud-edge-end collaboration environments.

The abundance of data provided by mobile edge devices enables a wide range of mobile edge computing (MEC) applications. Numerous studies have investigated efficient offloading methods for bandwidth savings in MEC. However, they focus on trading the device's computational cost for a reduction in communication, while edge devices can be rather resource-limited and must handle several jobs simultaneously. In this paper, the computation overhead on the device is pushed to its absolute minimum (almost no overhead), and consideration is given to enhancing the accuracy of the image recognition task within the constraints of the transmission volume limitation. We propose a mask-reconstruct system called MOT to mask images on the device side and recover images with the Masked Autoencoders (MAE)-based model on the server side. We further design a feedback-driven scheme to achieve content-aware transmission. Extensive experiments have been conducted to verify the effectiveness of the MOT.

Federated learning (FL) enables collaborative model training among distributed devices without data sharing, but existing FL suffers from poor scalability because of global model synchronization. To address this issue, hierarchical federated learning (HFL) has been recently proposed to let edge servers aggregate models of devices in proximity, while synchronizing via the cloud periodically. However, a critical open challenge about how to design a good synchronization scheme (when devices and edges should be synchronized) is still unsolved. Devices are heterogeneous in computing and communication capability, and their data could be non-IID. No existing work can well synchronize various roles (e.g., devices and edge) in HFL to guarantee high learning efficiency and accuracy. In this paper, we propose a learning-based synchronization scheme for HFL systems. By collecting data such as edge models, CPU usage, communication time, etc., we design a deep reinforcement learning-based approach to decide the frequencies of cloud aggregation and edge aggregation, respectively. The proposed scheme well considers device heterogeneity, non-IID data and device mobility, to maximize the training model accuracy while minimizing the energy overhead. We build an HFL testbed and conduct experiments using real data obtained from Raspberry Pi and Alibaba Cloud. Extensive experimental results have confirmed the effectiveness of Hwamei.

Connected autonomous vehicles (CAVs) are promising to improve road safety, thanks to various on-board sensors, such as LiDAR, radars, and stereo cameras. However, perception view could be significantly limited due to occlusions, extreme weather, and far objects. To address these challenges, in this paper, we propose an efficient edge-assisted perception sharing scheme, which enables vehicles to exchange the information about their sensed environment to improve road safety. We formulate perception sharing as an online optimization problem, with the objective of maximizing the total weighted utility, where utility indicates the quality of collected sensor data while weight means the intensity of the vehicle's demand for information in a certain area. To solve this problem, we propose an efficient online heuristic algorithm, which decouples the original problem into multiple sub-problems and solves them alternatively to find the optimal solution. Extensive simulations demonstrate that our proposed method can significantly improve the perception sharing performance.

Dynamic Graph Neural Network (DGNN) has shown a strong capability of learning dynamic graphs by exploiting both spatial and temporal features. Although DGNN has recently received considerable attention by AI community and various DGNN models have been proposed, building a distributed system for efficient DGNN training is still challenging. It has been well recognized that how to partition the dynamic graph and assign workloads to multiple GPUs plays a critical role in training acceleration. Existing works partition a dynamic graph into snapshots or temporal sequences, which only work well when the graph has uniform spatio-temporal structures. However, dynamic graphs in practice are not uniformly structured, with some snapshots being very dense while others are sparse. To address this issue, we propose DGC, a distributed DGNN training system that achieves a 1.25× - 7.52× speedup over the state-of-the-art in our testbed. DGC's success stems from a new graph partitioning method that partitions dynamic graphs into chunks, which are essentially subgraphs with modest training workloads and few inter connections. This partitioning algorithm is based on graph coarsening, which can run very fast on large graphs. In addition, DGC has a highly efficient run-time, powered by the proposed chunk fusion and adaptive stale aggregation techniques. Extensive experimental results on 3 typical DGNN models and 4 popular dynamic graph datasets are presented to show the effectiveness of DGC.

By leveraging standard IT virtualization technology and Commercial-Off-The-Shelf (COTS) servers, Network Function Virtualization (NFV) decouples network functions from proprietary hardware devices for flexible service provisioning. But the potential of NFV is significantly limited by its performance inefficiency. With the unparalleled advantages of multi-core parallelism and high memory bandwidth, Graphics Processing Units (GPUs) are regarded as a promising way to accelerate Virtualized Network Functions (VNF). However, the special architecture of GPU brings new challenges to task scheduling and resource allocation. To this end, we propose aGPUorientedspatio-temporal sharing framework for NFV calledGost, aiming for GPU based VNF performance promotion. The execution order and GPU resource allocation (i.e., the number of threads) are considered in task scheduling to minimize the end-to-end latency for VNF flows. First, we formulate the task scheduling problem into a nonlinear programming form, and then transform it into an equivalent Integer Linear Programming (ILP) form. The problem is proved as NP-hard. We customize the classical list scheduling algorithm and propose a List Scheduling based Spatio-Temporal GPU sharing strategy (LSSTG), whose achievable worst-case performance is also formally analyzed. We practically implementGostprototype, based on which extensive experiments verify the high performance efficiency of LSSTG compared to state-of-the-art in terms of latency and throughput.

As a rising star of social apps, short video apps, e.g., TikTok, have attracted a large number of mobile users by providing fresh and short video contents that highly match their watching preferences. Meanwhile, the booming growth of short video apps imposes new technical challenges on the existing computation and communication infrastructure. Traditional solutions maintain all videos on the cloud and stream them to users via contend delivery networks or the Internet. However, they incur huge network traffic and long delay that seriously affects users’ watching experiences. In this article, we propose an edge-assisted short video sharing framework to address these challenges by caching some highly preferred videos at edge servers that can be accessed by users via high-speed network connections. Since edge servers have limited computation and storage resources, we design an online algorithm with provable approximation ratio to decide which videos should be cached at edge servers, without the knowledge of future network quality and watching preferences changes. Furthermore, we improve the performance by jointly considering video fetching and user-edge association. Extensive simulations are conducted to evaluate the proposed algorithms under various system settings, and the results show that our proposals outperform existing schemes.

Blockchain is an important supporting technology for various sustainable systems. It relies on a number of distributed nodes running blockchain client software, which is responsible for some critical tasks, such as communicating with other nodes and generating new blocks. However, the quick evolution of blockchain technology brings crucial challenges to blockchain client design. After carefully examining existing blockchain client software, we have identified a critical weakness: Blockchain clients are weak in supporting live upgrades, resulting in a blockchain fork that incurs security concerns and risks. In this article, we propose Phoenix, a novel blockchain client design that is live upgradable. Phoenix uses blockchain service encapsulation to decouple blockchain services. Based on service encapsulation, we propose a live upgrade scheme that packs upgrade codes into blockchain transactions and uses a Just-In-Time engine to avoid service interruption. A parallel execution engine is developed to increase service efficiency. We evaluated Phoenix on a 51-node blockchain, and experimental results show that Phoenix outperforms existing solutions in overhead and upgrade latency.

Payment channel network (PCN) is a promising off-chain technology for blockchain scalability, but it suffers from poor sustainability in practice. In other words, due to the imbalanced transfer in channels, the balance in one direction of channels gradually becomes exhausted until the PCN is rebalanced via a consensus-based rebalancing protocol, during which the involved channels must be suspended. This paper presents Cycle, the first off-chain protocol for a sustainable PCN. It not only keeps the PCN at a balanced level consistently but also avoids the channel freeze incurred by the rebalancing protocol, leading to minimum failed payments and sustained PCN service, respectively. Cycle achieves these benefits based on a novel idea of asynchronous rebalancing. During the normal off-chain running, the participants share the information about their payments and asynchronously rebalance the PCN following the principle that payments along circular channels can cancel each other out. To guarantee security, the protocol resolves the disputes resulting from network latency or malicious participants by a message mechanism for synchronization and a smart contract for arbitration. Moreover, to address the privacy concern during the information sharing, a truncated Laplace mechanism is designed to achieve differential privacy. Finally, we provide a proof-of-concept implementation in Ethereum, over which a real data-based simulation shows that Cycle satisfies 31% more payments than the state-of-the-art technique.

Distributed machine learning (DML) has shown great promise in accelerating model training on multiple GPUs. To increase GPU utilization, a common practice is to let multiple learning jobs share GPU clusters, where the most fundamental and critical challenge is how to efficiently schedule these jobs on GPUs. However, existing works about DML job scheduling are constrained to settings with homogeneous GPUs. GPU heterogeneity is common in practice, but its influence on multiple DML job scheduling has been seldom studied. Moreover, DML jobs have internal structures that contain great parallelism potentials, which have not yet been fully exploited in the heterogeneous computing environment. In this paper, we propose Hare, a DML job scheduler that exploits both inter-job and intra-job parallelism in a heterogeneous GPU cluster. Hare has three novel designs. First, Hare optimizes GPU execution environment to reduce task switching overhead by exploiting unique features of DML scheduling. Second, Hare adopts a relaxed fixed-scale synchronization scheme that allows independent tasks to be flexibly scheduled within a training round. Finally, we propose a fast heuristic algorithm to minimize the total weighted job completion time by jointly considering job features and hardware heterogeneity. Its theoretical bound is derived. We evaluate Hare using a small-scale testbed and a trace-driven simulator. The results show that it can outperform the state-of-the-art by about 2x.

Transformer has dominated the field of natural language processing because of its strong capability in learning from sequential input data. In recent years, various computing and networking optimizations have been proposed for improving transformer training efficiency. However, transformer inference, as the core of many AI services, has been seldom studied. A key challenge of transformer inference is variable-length input. In order to align these input, existing work has proposed batching schemes by padding zeros, which unfortunately introduces significant computational redundancy. Moreover, existing transformer inference studies are separated from the whole serving system, where both request batching and request scheduling are critical and they have complex interaction. To fill the research gap, we propose TCB, a Transformer inference system with a novel ConcatBatching scheme as well as a jointly designed online scheduling algorithm. ConcatBatching minimizes computational redundancy by concatenating multiple requests, so that batch rows can be aligned with reduced padded zeros. Moreover, we conduct a systemic study by designing an online request scheduling algorithm aware of ConcatBatching. This scheduling algorithm needs no future request information and has provable theoretical guarantee. Experimental results show that TCB can significantly outperform state-of-the-art.

As a promising solution to blockchain scalability, sharding divides blockchain nodes into small groups called shards, splitting the workload. Existing works for sharding, however, are limited by cross-shard transactions, since they need to split each cross-shard transaction into multiple sub-transactions, each of which costs a consensus round to commit. In this paper, we introduce PYRAMID, a novel sharding system based on the idea of layered sharding. In PYRAMID, the nodes with better hardware are allowed to participate in multiple shards and store the blockchains of these shards thus they can validate and execute the cross-shard transactions without splitting. Next, to commit the cross-shard transactions with consistency among the related shards, we design a cooperative cross-shard consensus based on collective signature-based inter-shard collaboration. Furthermore, we present an optimization framework to compute an optimal layered sharding strategy maximizing the transaction throughput with the constraint of system security and node resource. Finally, we implement a prototype for PYRAMID based on Ethereum and the experimental results reveal the efficiency of PYRAMID in terms of performance and scalability, especially in workloads with a high percentage of cross-shard transactions. PYRAMID improves the throughput by up to 3.2 times compared with the state-of-the-art works and achieves about 3821 transaction per seconds for 20 shards.

Payment channel network (PCN) is one of the promising solutions for scalable blockchains since it shows great potential in improving blockchain network throughput. However, the growing number of transactions and the payment-channel sharing of concurrent transactions can lead to channel congestion. Although many studies have proposed different solutions to solve this problem, they ignore a fact that applications may have different transaction rate requirements at different times. In this paper, we propose a priority-aware PCN to meet the requirements of those transactions. Senders in priority-aware PCNs can specify the priority of their transactions by paying a corresponding forwarding fee on each hop along the transaction path. However, capacity competition occurs on the shared hops. Moreover, we propose a multi-agent DQN-based priority assignment algorithm to address the competition issue and design a PCN simulator for performance evaluation. Simulation results show that our solution can guarantee a high throughput of transactions and assign priorities appropriately to balance the transaction rate and forwarding fee cost. The experimental results demonstrate that the priority scheduling scheme can achieve higher transaction throughput and success ratio than other scheduling methods in a congested PCN environment.

Data collection in big data is an effective way to aggregate information that the collector is interested in. However, there is no assurance for the data that the users provide. Since collector does not have the ability to check the authenticity of every piece of information, the trustworthiness of users participated in the collection become important. In this paper, we design an efficient approach to calculate users’ trustworthiness in data collection for big data context. We divide the trustworthiness into familiarity trustworthiness and similarity trustworthiness, and study the influences of user actions on trustworthiness. To prevent malicious users from raising their trustworthiness and providing false information that may mislead final results, we also design a security queue to record users’ historical trust information, so that we can detect malicious users with high accuracy. Simulation results show that our model can sensitively resist the malicious actions of users.

As the large-scale deployment of machine learning applications, there is much research attention on exploiting a vast amount of data stored on mobile clients. To preserve data privacy, federated learning has been proposed to enable large-scale machine learning by massive clients without exposing raw data. Existing works of federated learning struggle for accelerating the learning process, but ignore the energy efficiency that is critical for resource-constrained clients. In this article, we propose to improve the energy efficiency of federated learning by lowering CPU cycle frequencies of clients who are faster in the training group. Based on this idea, we formulate an optimization problem aiming to minimize the total system cost defined as a weighted sum of learning time and energy consumption. Due to the hardness of the formulated optimization problem and unpredictability of network quality, we propose L4L (Learning for Learning), an experience-driven computational resource control approach based on the deep reinforcement learning, which can derive the near-optimal solution with only the clients’ bandwidth information in the previous training rounds. We conduct the experiments using both real-world traces and synthetic traces to evaluate the proposed L4L approach. The results demonstrate the superiority of L4L as compared with the state-of-the-art solutions.

Federated learning has attracted much research attention due to its privacy protection in distributed machine learning. However, existing work of federated learning mainly focuses on Convolutional Neural Network (CNN), which cannot efficiently handle graph data that are popular in many applications. Graph Convolutional Network (GCN) has been proposed as one of the most promising techniques for graph learning, but its federated setting has been seldom explored. In this article, we propose FedGraph for federated graph learning among multiple computing clients, each of which holds a subgraph. FedGraph provides strong graph learning capability across clients by addressing two unique challenges. First, traditional GCN training needs feature data sharing among clients, leading to risk of privacy leakage. FedGraph solves this issue using a novel cross-client convolution operation. The second challenge is high GCN training overhead incurred by large graph size. We propose an intelligent graph sampling algorithm based on deep reinforcement learning, which can automatically converge to the optimal sampling policies that balance training speed and accuracy. We implement FedGraph based on PyTorch and deploy it on a testbed for performance evaluation. The experimental results of four popular datasets demonstrate that FedGraph significantly outperforms existing work by enabling faster convergence to higher accuracy.

Sharding can significantly improve the blockchain scalability, by dividing nodes into small groups called shards that can handle transactions in parallel. However, all existing sharding systems adopt complete sharding, i.e., shards are isolated. It raises additional overhead to guarantee the atomicity and consistency of cross-shard transactions and seriously degrades the sharding performance. In this paper, we present Pyramid, the first layered sharding blockchain system, in which some shards can store the full records of multiple shards thus the cross-shard transactions can be processed and validated in these shards internally. When committing cross-shard transactions, to achieve consistency among the related shards, a layered sharding consensus based on the collaboration among several shards is presented. Compared with complete sharding in which each cross-shard transaction is split into multiple sub-transactions and cost multiple consensus rounds to commit, the layered sharding consensus can commit cross-shard transactions in one round. Furthermore, the security, scalability, and performance of layered sharding with different sharding structures are theoretically analyzed. Finally, we implement a prototype for Pyramid and its evaluation results illustrate that compared with the state-of-the-art complete sharding systems, Pyramid can improve the transaction throughput by 2.95 times in a system with 17 shards and 3500 nodes.

Federated learning has been proposed as a promising distributed machine learning paradigm with strong privacy protection on training data. Existing work mainly focuses on training convolutional neural network (CNN) models good at learning on image/voice data. However, many applications generate graph data and graph learning cannot be efficiently supported by existing federated learning techniques. In this paper, we study federated graph learning (FGL) under the cross-silo setting where several servers are connected by a wide-area network, with the objective of improving the Quality-of-Service (QoS) of graph learning tasks. We find that communication becomes the main system bottleneck because of frequent information exchanges among federated severs and limited network bandwidth. To conquer this challenge, we design Glint, a decentralized federated graph learning system with two novel designs: network traffic throttling and priority-based flows scheduling. To evaluate the effectiveness of Glint, we conduct both experiments on a testbed and trace-driven simulations. The results show that Glint can significantly outperform existing federated learning solutions.

Network Function Virtualization (NFV) enables network functions to run on general-purpose servers, thus alleviates the reliance on dedicated hardware and significantly improves the scalability and flexibility in networking service provisioning. Meanwhile, it is recognized that Virtualized Network Functions (VNFs) suffer from serious performance problem. Graphics Processing Unit (GPU), with massive processing cores, has been advocated as a potential accelerator for improving the performance efficiency of VNFs. However, the special architecture of GPU makes existing CPU-oriented task scheduling strategies fail to be applied, limiting the acceleration potential of GPUs. To this end, we propose a GPU-oriented spatio-temporal sharing framework as Gost to improve the performance of GPU-accelerated VNFs. We also study how to minimize the end-to-end latency of VNF flows via careful scheduling on the execution order and the GPU resource allocation (i.e., the number of threads). We first formally describe the problem as a non-linear integer programming problem, which is then equivalently transformed into an integer linear programming (ILP) form. Considering the high computation complexity of solving ILP, we further propose a customized list scheduling based spatio-temporal GPU sharing strategy (LSSTG). We have practically implemented a prototype of Gost, based on which we also verify the high efficiency of LSSTG by extensive experiments.

Human behavior recognition which is the indispensable technology for Artificial Intelligence(AI) application like smart home and other practical applications, is very challenging as the optimal recognition generally is required to be non-invasive and easy to deploy. An increasing interest has been paid on the human behavior recognition with off-the-shelf Wi-Fi devices. However, most of existing works just limit their focus on the small-scale scene while human behavior recognition will be quite different in large areas for a larger number of antennas and correspondingly a more complex antenna layout. For example, if we want to build a complete behavior awareness system using the distributed Wi-Fi equipment of the entire building, though collecting and using all antennas’ data is feasible maybe, the overhead concerns of computing and bandwidth resources, and the operation complexity will be hard to lessen in practice. In this paper, we first present analyses of the signal performances between different antenna pairs. Then closely following these analyses, we propose a novel scheme for the large-area human behavior recognition. Finally, we conduct extensive confirmatory experiments to verify the validity of our proposed scheme.

Efficient urban traffic monitoring is a key enabler for intelligent planning and management of modern cities. Network tomography can monitor the urban traffic with a comparably small number of traffic detectors like cameras, and has become an appealing technique for urban traffic management. However, previous work on network tomography based traffic monitoring focuses primarily on developing estimators using the given end-to-end travel time measurements, while the design of data collection for efficiently distributed collecting and processing the raw monitoring videos to such measurements is often neglected. We fill this gap by exploring the vision of edge intelligence for optimal urban traffic monitoring, and tackle the following two problems in regard of limited telecommunications resources: 1) when the total number of monitoring videos that are successfully processed into the end-to-end travel time measurements is pre-bounded, we employ a Fisher Information Matrix (FIM) to help determine the best quota scheme for the monitoring videos that each traffic detector need to generate and 2) when the centralised processing of monitoring videos alone is insufficient, we make use of the computation capabilities from these edge devices, i.e., traffic detectors, and employ a multi-agent reinforcement learning approach to help them conduct intelligent computation offloading individually. Extensive simulations demonstrate that our proposed scheme effectively reduces the estimation error of network tomography compared to common approaches with either uniform or random strategy.

The parameter server (PS) paradigm has achieved great success in deploying large-scale distributed Deep Learning (DL) systems. However, these systems implicitly assume that the cluster is homogeneous and this assumption does not hold in many realworld cases. Although the previous efforts are paid to address heterogeneity, they mainly prioritize the contribution of fast workers and reduce the involvement of slow workers, resulting in the limitations of workload imbalance and computation inefficiency. We reveal that grouping workers into communities, an abstraction proposed by us, and handling parameter synchronization at the community level can conquer these limitations and accelerate the training convergence progress. The inspiration of community comes from our exploration of prior knowledge about the similarity between workers, which is often neglected by previous work. These observations motivate us to propose a new synchronization mechanism named Community-aware Synchronous Parallel (CASP), which uses the Asynchronous Advantage Actor-Critic (A3C)-based algorithm to intelligently determine community configuration and fully improve the synchronization performance. The whole idea has been implemented in a prototype system called Petrel that achieves a good balance between convergence efficiency and communication overhead. The evaluation under various benchmarks with multiple metrics and baseline comparison demonstrates the effectiveness of Petrel. Specifically, Petrel accelerates the training convergence speed by up to 1.87 x faster and reduces communication traffic by up to 26.85 percent, on average, over the non-community synchronization mechanisms.

The accurate identification of congested links is crucial for network performance monitoring. Network boolean tomography uses end-to-end path measurements to identify congested links, and appears as a significant alternative when direct link monitoring is not available. However, most of existing tomographic methods assume no correlations between links, i.e., the congestion of one link is assumed to be independent from the congestion of any others, hindering their applications in practice because links could become correlated during a joint optimization procedure of many network operations like traffic routing and balancing. In this paper, we study practical network boolean tomography without such an assumption. We elaborate on the ill-posed nature of network boolean tomography to highlight the significance of integrating link correlations, and model the congested link identification from end-to-end congestion observations of paths as a problem of Maximum A-Posteriori (MAP) estimation. To avoid the explicit acquisition of any priori knowledge of link correlations, we then propose a learning-based algorithm with Long Short-term Memory (LSTM), a special recurrent neural network that is good at learning statistical dependencies of sequence elements from historical data. Numerical results over real network topologies validate our learning-based network boolean tomography.

Recently, deep learning (DL) becomes increasingly important since its transformative effect on a wide range of applications. During inference process, the DL model is deployed on the cloud to answer online queries. One crucial issue in the progress of DL inference is energy consumption, which significantly retards computation performance. Therefore, many previous investigations decrease the energy consumption via computation reuse technique based on similarity. However, if input data consists individually from mobile devices, applying these schemes will significantly decline computation performance. Because in disordered individual inputs, similarity for reuse is difficult to exploit directly. Results of initial experimental observations show that (1) individual input data also has high similarity for reuse, and (2) the total similarity during computation process has a relation with the characteristics of input data. This motivates us to design a reordering scheme to enhance similarity for computation reuse. Our main approaches are using statistical theory to predict the similarities among input data, and determining the execution sequence. Based on these approaches, we propose an effective input reordering scheme for computation reuse to save energy consumption. The evaluation under various benchmarks demonstrates that the reordering scheme significantly outperforms the previous schemes, for instance, the computation reuse is enhanced to $1.1 \times$ and the energy consumption is minimized to 40% according to the configuration of traditional computation reuse technique.

Payment channel networks (PCN) have demonstrated its significant advantages in improving the scalability of blockchain. However, the existing work of PCN leads to serious privacy leakage problem that intermediate nodes along a payment path can collude to obtain the payment amounts and payment receivers. To address this problem, we propose to move PCN-related modules into the Trusted Execution Environment (TEE) commonly available on modern CPUs, so that adversaries cannot access the critical payment information protected by TEE, even though they compromise the software (e.g., blockchain clients or operating system) outside of TEE. An additional challenge is that adversaries can still infer payment receivers by observing the pattern of message transmissions among nodes. To hide payment receivers, we further propose to send redundant transactions to pseudo receivers to confuse adversaries. A fast algorithm with provable approximation ratio has been proposed to maximize the level of privacy protection under the constraint of communication overhead. Both experiments on a small-scale testbed and large-scale simulations are conducted to evaluate our proposal. The results show that our proposed solution outperforms existing work significantly.

As to address the impact of heterogeneity in distributed Deep Learning (DL) systems, most previous approaches focus on prioritizing the contribution of fast workers and reducing the involvement of slow workers, incurring the limitations of workload imbalance and computation inefficiency. We reveal that grouping workers into communities, an abstraction proposed by us, and handling parameter synchronization in community level can conquer these limitations and accelerate the training convergence progress. The inspiration of community comes from our exploration of prior knowledge about the similarity between workers, which is often neglected by previous work. These observations motivate us to propose a new synchronization mechanism named Community-aware Synchronous Parallel (CSP), which uses the Asynchronous Advantage Actor-Critic (A3C), a Reinforcement Learning (RL) based algorithm, to intelligently determine community configuration and fully improve the synchronization performance. The whole idea has been implemented in a system called Petrel that achieves a good balance between convergence efficiency and communication overhead. The evaluation under different benchmarks demonstrates our approach can effectively accelerate the training convergence speed and reduce synchro-nization traffic.

Off-blockchain payment channels can significantly improve blockchain scalability by enabling a large number of micro-payments between two blockchain nodes, without committing every single payment to the blockchain. Multiple payment channels form a payment network, so that two nodes without direct channel connection can still make payments. A critical challenge in payment network construction is to decide how many funds should be deposited into payment channels as initial balances, which seriously influences the performance of payment networks, but has been seldom studied by existing work. In this paper, we address this challenge by designing PnP, a balance planning service for payment networks. Given estimated payment demands among nodes, PnP can decide channel balances to satisfy these demands with a high probability. It does not rely on any trusted third-parties, and can provide strong protection from malicious attacks with low overhead. It obtains these benefits with two novel designs, the cryptographic sortition and the chance-constrained balance planning algorithm. Experimental results on a testbed of 30 nodes show that PnP can enable 30% more payments than other designs.

Federated learning is promising in enabling large-scale machine learning by massive mobile devices without exposing the raw data of users with strong privacy concerns. Existing work of federated learning struggles for accelerating the learning process, but ignores the energy efficiency that is critical for resource-constrained mobile devices. In this paper, we propose to improve the energy efficiency of federated learning by lowering CPU-cycle frequency of mobile devices who are faster in the training group. Since all the devices are synchronized by iterations, the federated learning speed is preserved as long as they complete the training before the slowest device in each iteration. Based on this idea, we formulate an optimization problem aiming to minimize the total system cost that is defined as a weighted sum of training time and energy consumption. Due to the hardness of nonlinear constraints and unawareness of network quality, we design an experience-driven algorithm based on the Deep Reinforcement Learning (DRL), which can converge to the near-optimal solution without knowledge of network quality. Experiments on a small-scale testbed and large-scale simulations are conducted to evaluate our proposed algorithm. The results show that it outperforms the start-of-the-art by 40% at most.

Internet of Things (IoT) generates large amounts of data at the network edge. Machine learning models are often built on these data, to enable the detection, classification, and prediction of the future events. Due to network bandwidth, storage, and especially privacy concerns, it is often impossible to send all the IoT data to the data center for centralized model training. To address these issues, federated learning has been proposed to let nodes use the local data to train models, which are then aggregated to synthesize a global model. Most of the existing work has focused on designing learning algorithms with provable convergence time, but other issues, such as incentive mechanism, are unexplored. Although incentive mechanisms have been extensively studied in network and computation resource allocation, yet they cannot be applied to federated learning directly due to the unique challenges of information unsharing and difficulties of contribution evaluation. In this article, we study the incentive mechanism for federated learning to motivate edge nodes to contribute model training. Specifically, a deep reinforcement learning-based (DRL) incentive mechanism has been designed to determine the optimal pricing strategy for the parameter server and the optimal training strategies for edge nodes. Finally, numerical experiments have been implemented to evaluate the efficiency of the proposed DRL-based incentive mechanism.

Edge computing is a new paradigm to provide strong computing capability at the edge of pervasive radio access networks close to users. A critical research challenge of edge computing is to design an efficient offloading strategy to decide which tasks can be offloaded to edge servers with limited resources. Although many research efforts attempt to address this challenge, they need centralized control, which is not practical because users are rational individuals with interests to maximize their benefits. In this article, we study to design a decentralized algorithm for computation offloading, so that users can independently choose their offloading decisions. Game theory has been applied in the algorithm design. Different from existing work, we address the challenge that users may refuse to expose their information about network bandwidth and preference. Therefore, it requires that our solution should make the offloading decision without such knowledge. We formulate the problem as a partially observable Markov decision process (POMDP), which is solved by a policy gradient deep reinforcement learning (DRL) based approach. Extensive simulation results show that our proposal significantly outperforms existing solutions.

A modern Network Function Virtualization (NFV) service is usually expressed in a service chain that contains a list of ordered network functions, each can run in one or multiple virtual machines. Although lots of efforts have been devoted to service chain deployment, the researchers normally consider a simple model of network functions where different service chains have their own network functions no matter whether some of the network function appliances are interdependent. In this paper, we study the service chain deployment by exploiting two types of correlations between network functions: the Coordination Effect due to information exchanges among multiple VMs running the same network function, and the Traffic-Change Effect where the volume of outgoing traffic is not necessarily equal to the volume of its incoming traffic at each network function because of packet manipulations such as compression and encryption. These two effects have not been studied simultaneously in the context of service chaining. With theobjective to maximize the profit measured by the admitted traffic minus the implementation cost, we first formulate a joint service-function deployment and traffic scheduling (SUPER) problem that is proved to be NP-hard. We then devise an approximation algorithm based on the Markov approximation technique and analyze its theoretical bound on the convergence time. Simulation results show that the proposed algorithm outperforms two existing benchmark algorithms significantly.

MapReduce plays a critical role as a leading framework for big data analytics. In this paper, we consider a geo-distributed cloud architecture that provides MapReduce services based on the big data collected from end users all over the world. Existing work handles MapReduce jobs by a traditional computation-centric approach that all input data distributed in multiple clouds are aggregated to a virtual cluster that resides in a single cloud. Its poor efficiency and high cost for big data support motivate us to propose a novel data-centric architecture with three key techniques, namely, cross-cloud virtual cluster, data-centric job placement, and network coding based traffic routing. Our design leads to an optimization framework with the objective of minimizing both computation and transmission cost for running a set of MapReduce jobs in geo-distributed clouds. We further design a parallel algorithm by decomposing the original large-scale problem into several distributively solvable subproblems that are coordinated by a high-level master problem. Finally, we conduct real-world experiments and extensive simulations to show that our proposal significantly outperforms the existing works.

Crowdsensing has been well recognized as a promising approach to enable large scale urban data collection. In a typical crowdsensing system, the task owner usually needs to provide incentives to the users (say participants) to encourage their participation. Among existing incentive mechanisms, posted pricing has been widely adopted because it is easy to implement while ensuring truthfulness and fairness. One critical challenge to the task owner is to set the right posted price to recruit a crowd with small total payment and reasonable sensing quality, i.e., posted pricing problem for robust crowdsensing. However, this fundamental problem remains largely open so far. In this paper, we model the robustness requirement over sensing data quality as chance constraints in an elegant manner, and study a series of chance constrained posted pricing problems in crowdsensing systems. Although some chance-constrained optimization techniques have been applied in the literature, they cannot provide any performance guarantees for their solutions. In this work, we propose a binary search based algorithm, and show that using this algorithm allows us to establish theoretical guarantees on its performance. Extensive numerical simulations demonstrate the effectiveness of our proposed algorithm.

Nowadays, the availability of parking spaces is far behind the quick rising number of cars. Rather than building more lots, a better way is to share private-owned parking spaces. However, this faces the challenge that users are not willing to expose their privacy to the public. To solve this problem, we propose a new architecture for parking space sharing, integrating homomorphic cryptography into the design of a secure protocol for parking space searching and booking. The proposed privacy-preserving matching scheme (PPMS) is constructed in an untrusted third-party service system including two independent entities, namely, a server and an intermediary platform. Via the participant comparison protocol (PCP), a driver can choose from the matching result and be navigated to the parking space near his destination, without knowing any information of the provider and vice versa. In the meanwhile, in order to further improve the efficiency of matching, we also propose a block algorithm based on the longitude and latitude (BABLL), which utilizes a novel partitioning scheme. The feasibility of the architecture is validated through the detailed theoretical analysis and extensive performance evaluations, including the assessment of the resilience to attacks.

Parameter Server (PS) has been widely used to train a large amount of data on multiple machines in parallel. In parameter server, a critical problem is how to effectively schedule multiple training jobs to minimize the job completion time. Some existing work has proposed methods of setting the number of concurrent workers. However, they do not effectively consider the topology of GPU placement which affects the efficiency of communication. This paper proposes a novel resource-to-time model based on the number of workers and the topology of GPU placement. According to the model, we propose an algorithm called TOPO-PS particularly for topology problem in parameter servers. The algorithm achieves the placement strategy based on graph mapping algorithm. Evaluation under various algorithms evidences the superiority of our algorithm. TOPO-PS yields shorter job completion, by up to 53.48% of that of FIFO and 88.77% of OASIS.

Constructing Radio Environment Map (REM) accurately and cost-efficiently is of great importance to realize dynamic spectrum access. Two kinds of approaches are widely investigated recently, i.e., radio propagation model based approaches and sensor monitoring based approaches. However, these existing approaches are suffering from either inaccurate spectrum availability or high deployment cost. To this end, outsourcing the spectrum sensing task to mobile users that are outfitted with spectrum sensors could greatly reduce the operator's expenditure, and meanwhile, achieve a satisfactory accuracy. The key of crowdsourced REM construction is to attract user participation. In this paper, we propose a novel online incentive mechanism for constructing a fine-grained REM with crowdsourcing in a realistic scenario, where the mobile users arrive and leave in an online manner. The proposed mechanism is proven to satisfy the truthfulness, individual rationality, computational efficiency and consumer sovereignty. Evaluation results demonstrate that the proposed mechanism outperforms the baseline schemes substantially.

Mini-batch Stochastic Gradient Descent (MGD) is one of the most widely used methods in Machine Learning (ML) model training. Typically, before a training process starts, researchers should manually set a fixed batch size, which is a hyper-parameter indicating the size of the random slice of the whole dataset that is trained in a single iteration. In this paper, we propose a light-weight dynamic batch size fitting algorithm based on online efficient evaluation, which has the ability of automatically tuning batch size during the train process to reach a best-so-far efficiency, but with little overhead. The experimental results have demonstrated that the algorithm is more effective compared with the commonly used fixed settings.

Network congestion will result in significant performance degradation or even failures of many bandwidth-hungry Internet of Things (IoT) applications. Accurate and efficient congested link identification has become a foundational issue to IoT applications like self-driving cars, digital health, smart city, and so on. However, directly monitoring the massive number of interior links often introduces high operation cost or even is infeasible in practice, giving rise to indirect monitoring techniques like network Boolean tomography. Nevertheless, in many networks, the number of their interior links is larger than their end-to-end paths, making it very challenging for network Boolean tomography to find a determined solution. To resolve this issue, most of current methods try to utilize some prerequisites, such as the link congestion probabilities. While these probabilities might be hard or even unable to be obtained accurately in dynamical networks, limiting the practical deployment. In this paper, we are motivated to design a framework of congested link identification without any prerequisite or assumption. We first novelly model the congested link identification procedures as a Markov decision processes (MDPs), and then employ a reinforcement learning technology, i.e., Q-learning, to solve this MDP. The simulation results show that our proposed scheme can autonomously and efficiently explore the unknown network environment, and is able to achieve better adaptivity and correctness, without any prior knowledge comparing to existing methods.

We are living in a world where massive end devices perform computing everywhere and everyday. However, these devices are constrained by the battery and computational resources. With the increasing number of intelligent applications (e.g., augmented reality and face recognition) that require much more computational power, they shift to perform computation offloading to the cloud, known as mobile cloud computing (MCC). Unfortunately, the cloud is usually far away from end devices, leading to a high latency as well as the bad quality of experience (QoE) for latency-sensitive applications. In this context, the emergence of edge computing is no coincidence. Edge computing extends the cloud to the edge of the network, close to end users, bringing ultra-low latency and high bandwidth. Consequently, there is a trend of computation offloading toward edge computing. In this paper, we provide a comprehensive perspective on this trend. First, we give an insight into the architecture refactoring in edge computing. Based on that insight, this paper reviews the state-of-the-art research on computation offloading in terms of application partitioning, task allocation, resource management, and distributed execution, with highlighting features for edge computing. Then, we illustrate some disruptive application scenarios that we envision as critical drivers for the flourish of edge computing, such as real-time video analytics, smart “things” (e.g., smart city and smart home), vehicle applications, and cloud gaming. Finally, we discuss the opportunities and future research directions.

Big data analytics in datacenters often involve scheduling of data-parallel jobs. Traditional scheduling techniques based on improving network resource utilization are subject to limited bandwidth in datacenter networks. To alleviate the shortage of bandwidth, some cluster frameworks employ techniques of traffic compression to reduce transmission consumption. However, they tackle scheduling in a coarse-grained manner at task level and do not perform well in terms of flow-level metrics due to high complexity. Fortunately, the abstraction of coflow pioneers a new perspective to facilitate scheduling efficiency. In this paper, we introduce a coflow compression mechanism to minimize the completion time in data-intensive applications. Due to the NP-hardness, we propose a heuristic algorithm called Fastest-Volume-Disposal-First (FVDF) to solve this problem. For online applicability, FVDF supports stage pipelining to accelerate scheduling and exploits recurrent neural networks (RNNs) to predict compression speed. Meanwhile, we build Swallow, an efficient scheduling system that implements our proposed algorithms. It minimizes coflow completion time (CCT) while guaranteeing resource conservation and starvation freedom. The results of both trace-driven simulations and real experiments show the superiority of our algorithm, over existing one. Specifically, Swallow speeds up CCT and job completion time (JCT) by up to 1.47χ and 1.66χ on average, respectively, over the SEBF in Varys, one of the most efficient coflow scheduling algorithms so far. Moreover, with coflow compression, Swallow reduces data traffic by up to 48.41 percent on average.

The emergence of edge computing has witnessed a fast-growing volume of data on edge devices belonging to different stakeholders which, however, cannot be shared among them due to the lack of the trust. By exploiting blockchain's non-repudiation and non-tampering properties that enable trust, we develop a blockchain-based big data sharing framework to support various applications across resource-limited edges. In particular, we devise a number of novel resource-efficient techniques for the framework: (1) the PoC (Proof-of-Collaboration) based consensus mechanism with low computation complexity which is especially beneficial to the edge devices with low computation capacity, (2) the blockchain transaction filtering and offloading scheme that can significantly reduce the storage overhead, and (3) new types of blockchain transaction (i.e., Express Transaction) and block (i.e., Hollow Block) to enhance the communication efficiency. Extensive experiments are conducted and the results demonstrate the superior performance of our proposal.

Emotion is a unique feature of human beings. Recent research in emotion sensing has already revealed its potentials in enhancing our living experiences through applications like emotion companion and autism treatment. However, existing solutions exploring audiovisual clues or psychological sensors have several critical concerns such as the availability (specialized hardware), reliability (illumination and line-of-sight constraints) and privacy issues (being watched). To this end, we present EmoSense, a first-of-its-kind WiFi-based emotion sensing system leveraging the temporal and frequency fingerprints on the wireless channel data induced by the physical expression of emotion. EmoSense has been prototyped with off- the-shelf WiFi devices and evaluated by comparing with the main-stream sensor-based approach in real environments. Experimental results demonstrate its effectiveness and robustness. Considering that EmoSense is compatible with existing WiFi infrastructures, it constitutes a low-cost yet promising solution for emotion sensing.

Fuel cells are promising power sources for green data centers thanks to its high energy-efficiency, low greenhouse gas emissions and high reliability. However, fuel cells have a unique feature called limited load following, i.e., they are slow in adjusting power supply due to mechanical limitation of fuel delivery. When power demand of data centers suddenly grows, fuel cells would fail to provide sufficient power supply. On the other hand, fuel cells are slow to reduce its power supply when demand decreases, leading to energy waste. In this paper, we study to mitigate the impact of limited load following by associating a set of heterogeneous batteries with fuel cells. These batteries with different characteristics (e.g., capacity, charging and discharging rate) can power data centers when the energy supply of fuel cells is insufficient. They are charged by excessive power supply when demand decreases. Given future power demand, we formulate the energy management problem as a mixed-integer nonlinear programming. An online algorithm is designed to solve the problem without future knowledge. We conduct extensive simulations using real-world traces and results show that our proposed algorithm significantly outperforms existing solutions.

With the emergence of edge computing in various applications domains, end users are now surrounded by a fast growing volume of data from edge devices belonging to different stakeholders. However, these edge devices cannot cooperate to share big data because of the distrust among them. In this paper, the blockchain is deployed in collaborative edges by exploiting the non-repudiation and non-tampering properties to enable trust. First, we develop a blockchain based big data sharing framework in collaborative edges for adapting to the limited computational and storage resources in edge devices. Then, a consensus mechanism called Proof-of-Collaboration (PoC) is proposed for computational resources reduction in our proposed framework, where edge devices offer their credits of PoC to compete for the block generation. Moreover, we put forward a futile transaction filter algorithm for transaction offloading, greatly reducing the storage resources occupied by the blockchain in edges. Extensive experiments are performed to demonstrate the superior performance of our proposal.

Big data analytics in datacenters often involves scheduling of data-parallel job, which are bottlenecked by limited bandwidth of datacenter networks. To alleviate the shortage of bandwidth, some existing work has proposed traffic compression to reduce the amount of data transmitted over the network. However, their proposed traffic compression works in a coarse-grained manner at job level, leaving a large optimization space unexplored for further performance improvement. In this paper, we propose a flow-level traffic compression and scheduling system, called Swallow, to accelerate data-intensive applications. Specifically, we target on coflows, which is an elegant abstraction of parallel flows generated by big data jobs. With the objective of minimizing coflow completion time (CCT), we propose a heuristic algorithm called Fastest-Volume-Disposal-First (FVDV) and implement Swallow based on Spark. The results of both trace-driven simulations and real experiments show the superiority of our system, over existing algorithms. Swallow can reduce CCT and job completion time (JCT) by up to 1.47 × and 1.66 × on average, respectively, over the SEBF in Varys, one of the most efficient coflow scheduling algorithms so far. Moreover, with coflow compression, Swallow reduces data traffic by up to 48.41% on average.

Edge computing is an emerging technology which places computing at the edge of the network to provide an ultra-low latency. Computation offloading, a paradigm that migrates computing from mobile devices to remote servers, can now use the power of edge computing by offloading computation to cloudlets in edge-clouds. However, the task scheduling of computation offloading in edge-clouds faces a two-fold challenge. First, as cloudlets are geographically distributed, it is difficult for each cloudlet to perform load balancing without centralized control. Second, as tasks of computation offloading have a wide variety of types, to guarantee the user quality of experience (QoE) in terms of task types is challenging. In this paper, we present Petrel, a distributed and application-aware task scheduling framework for edge-clouds. Petrel implements a sample-based load balancing technology and further adopts adaptive scheduling policies according to task types. This application-aware scheduling not only provides QoE guarantee but also improves the overall scheduling performance. Trace-driven simulations show that Petrel achieves a significant improvement over existing scheduling strategies.

In spectrum sharing system, an exclusion zone is defined to protect both primary and secondary users from interference. Reducing the size of exclusion zone is critical for efficiently utilizing the fallow spectrum. In this paper, we propose a novel crowdsourcing augmented exclusion zone refinement framework. In our framework, a barter-like exchange model using spectrum access right is employed to incentivize the secondary users (SUs) to participate in the crowdsourcing. We further design a truthful auction mechanism to select the SUs and determine their access time in a computationally efficient way. We perform simulations to validate the proposed mechanism, and compare it with two baseline schemes.

When a large-scale disaster happens, efficient network connection and communication becomes difficult due to serious damage of existing network infrastructures. Meantime, people have strong demands of information sharing with each other for evacuation and disaster-relief activities in such a disaster environment. To serve these heavy communication demands, establishing local area networks (LANs) consisting of portable servers has been considered as one of the most promising solutions. Based on the established LANs, people can share disaster-related information in covered area. However, due to the lack of stable Internet connection, these LANs are isolated and cannot be synchronized in real time. To tackle this problem, in this paper, we propose an intermittent data synchronization scheme by introducing moving vehicles as relays to exchange data between isolated data servers after disasters. With the objective of maximizing the synchronized weighted data volume under the capability constraints of the mobile relay, we formulate a stochastic programming problem for trajectory planning. We leverage queueing theory and the Lyapunov-drift technique to solve this problem in an online setting, which is practical for a real disaster environment. Our theoretical analysis shows that the performance gap of our proposed online algorithm is (1/V) of the optimum. Additionally, extensive simulations and comparisons with other algorithms are conducted to show the superior performance of our proposed online algorithm.

To improve the spectrum utilization efficiency, radio propagation model based spectrum database is widely investigated recently. However, it is prone to offer inaccurate and stale spectrum availability since the empirical models do not count for local environment details. One promising solution is to incorporate real- time spectrum measurement into the quasi-static spectrum database. In this paper, we propose a novel fine-grained incentive mechanism for sensing augmented spectrum database. We first present a reverse auction framework, which minimizes the operator's total expenditure subject to the quality requirement of each spot that needs to be augmented. Then we propose a practical incentive mechanism to solve the auction problem, which is proven to be truthful, individual rational and computationally efficient. Simulation results demonstrate that the proposed mechanism could save noticeable expenditure compared to two baseline schemes.

The recent progress in big data has inspired lots of data- parallel applications deployed in the datacenters. Although how to optimize the data flow scheduling in datacenters has been extensively studied, traditional per-flow based optimizations usually do not perform well in dealing with the transferring of a collection of parallel flows, i.e., coflow. Consequently, how to schedule the coflow towards various objectives, e.g., minimizing the coflow completion time, has attracted much attention recently. We notice that existing coflow scheduling studies usually suggest a fixed destination for each coflow. Taking the advantage of virtualization technology, we argue that the destination can be flexibly placed in the cloud. Therefore, it is essential to jointly optimize the coflow scheduling and data processor placement. In this paper, we are motivated to investigate the problem of coflow completion time minimization with joint consideration of coflow scheduling and data processor placement. We first formally describe the problem into a mixed integer non-linear programming (MINLP) problem. By linearizing the MINLP, we further propose a relaxation based heuristic algorithm. Via extensive simulation studies, the high efficiency of our heuristic algorithm is validated.

Some sleep disorders, such as sleep apnea, restless legs syndromes (RLS), and periodic limb movement disorder (PLMD), require a full-night sleep monitoring for diagnosis. Conventional sleep monitoring devices are disturbing and inconvenient for daily scene applications. In this poster paper, we propose a sleep monitoring system by embedding RFID tags into bed cloth and realize two main functions: breath monitoring and body movement detection. We apply a finite impulse response low pass filter to get smooth breath signal wave and use a convolutional neural network (CNN) algorithm to identify the movement of person objects. Finally, we conduct experiments to evaluate the breath monitoring in a real world scenario. The experiment results show that our monitoring system can monitor breath with a high accuracy.

Big data analysis is usually casted into parallel jobs running on geo-distributed data centers. Different from a single data center, geo-distributed environment imposes big challenges for big data analytics due to the limited network bandwidth between data centers located in different regions. Although research efforts have been devoted to geo-distributed big data, the results are still far from being efficient because of their suboptimal performance or high complexity. In this paper, we propose a traffic-aware task placement to minimize job completion time of big data jobs. We formulate the problem as a non-convex optimization problem and design an algorithm to solve it with proved performance gap. Finally, extensive simulations are conducted to evaluate the performance of our proposal. The simulation results show that our algorithm can reduce job completion time by 40%, compared to a conventional approach that aggregates all data for centralized processing. Meanwhile, it has only 10% performance gap with the optimal solution, but its problem-solving time is extremely small.

It has been well recognized that network transmission constitutes a large portion of smartphone energy consumption, mainly because of the tail energy caused by cellular network interface. Traffic offloading has been proposed to reduce energy by letting a smartphone offload network traffic to its neighbors in vicinity via low-power direct connections (e.g., WiFi Direct or Bluetooth). Our experiments conducted in a realistic environment reveal that energy efficiency cannot be improved or even deteriorates without a carefully designed offloading strategy. In this paper, we propose a selective traffic offloading scheme implemented as a smartphone middleware in a software-defined fashion, which consists of a packet classifier and a traffic scheduler. Using a light-weight machine learning approach exploiting unique smartphone context information, the packet classifier identifies packets generated on the fly as offloadable or not with substantially improved efficiency and feasibility on resource limited smartphones compared to traditional approaches. Both testbed and simulation based experiments are conducted and the results show that our proposal always attains the superior performance on a number of comparison metrics.

Cloud gaming is promising to provide high-quality game services by outsourcing game execution to cloud so that users can access games via thin clients (e.g., smartphones or tablets). However, existing cloud gaming systems su er from low GPU utilization in the virtualized environment. Moreover, GPU resources are scheduled in units of virtual machines (VMs) and this kind of coarse-grained scheduling at the VM-level fails to fully exploit GPU processing capacity. In this paper, we present ShareRender, a cloud gaming sys- tem that o oads graphics workloads within VMs directly to GPUs, bypassing GPU virtualization. For each game running in a VM, ShareRender starts a graphics wrapper to intercept frame rendering requests and assign them to render agents responsible for frame rendering on GPUs. Thanks to the exible workload assignment among multiple render agents, ShareRender enables ne-grained resource sharing at the frame-level to signi cantly improve GPU utilization. Further more, we design an online algorithm to determine workload assignment and migration of render agents, which considers the tradeo between minimizing the number of active server and low agent migration cost. We conduct experiments on real deployment and trace-driven simulations to evaluate the performance of ShareRender under di erent system settings. The results show that ShareRender outperforms the existing video-streaming-based cloud gaming system by over 4 times.

Summary Deep packet inspection (DPI) is important for network security. In this paper, we consider a software‐defined network where several DPI proxy nodes are available for serving flows from ingress switches. These DPI proxy nodes can be implemented in either software or hardware. We study an integrated proxy allocation and routing determining problem with the objective of minimizing the total delay of flows from ingress switches to DPI proxies. This problem is formulated as an integer linear programming problem that is NP‐hard in general. To solve this problem, we design a 2‐phase algorithm that can quickly select proxy and find routing paths for incoming flows. Finally, extensive simulations are conducted to evaluate the performance of our proposed algorithm. Some useful parameter setting insights are obtained.

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.

Big data analytics has attracted close attention from both industry and academic because of its great benefits in cost reduction and better decision making. As the fast growth of various global services, there is an increasing need for big data analytics across multiple data centers (DCs) located in different countries or regions. It asks for the support of a cross-DC data processing platform optimized for the geo-distributed computing environment. Although some recent efforts have been made for geo-distributed big data analytics, they cannot guarantee predictable job completion time, and would incur excessive traffic overthe inter-DC network that is a scarce resource shared by many applications. In this paper, we study to minimize the inter-DC traffic generated by MapReduce jobs targeting on geo-distributed big data, while providing predicted job completion time. To achieve this goal, we formulate an optimization problem by jointly considering input data movement and task placement. Furthermore, we guarantee predictable job completion time by applying the chance-constrained optimization technique, such that the MapReduce job can finish within a predefined job completion time with high probability. To evaluate the performance of our proposal, we conduct extensive simulations using real traces generated by a set of queries on Hive. The results show that our proposal can reduce 55 percent inter-DC traffic compared with centralized processing by aggregating all data to a single data center.

Energy Internet (EI) is proposed as the evolution of smart grid, aiming to integrate various forms of energy into a highly flexible and efficient grid that provides energy packing and routing functions, similar to the Internet. As an essential part in EI system, a scalable and interoperable communication infrastructure is critical in system construction and operation. In this article, we survey the recent research efforts on EI communications. The motivation and key concepts of EI are first introduced, followed by the key technologies and standardizations enabling the EI communications as well as security issues. Open challenges in system complexity, efficiency, reliability are explored and recent achievements in these research topics are summarized as well.

After big disasters, a damaged area can be out of contact because of severe damage of existing network infrastructures. Meanwhile, high demands for network connections to the disaster area will arise to collect damage information and disseminate rescue instructions. In this paper, we propose to construct a network for disaster management using mobile stations equipped with sensors and network interfaces. They are controlled by a disaster management center via wide area network technology, and conduct various disaster management tasks, such as damage sensing, information collection and message dissemination. To address the challenges of unpredictable tasks and limited number of mobile stations with working capability constraints, we propose an online algorithm that schedules mobile stations for disaster management tasks with weights in each time slot, without any knowledge of future task arrivals. Our objective is to maximize the total weight of finished tasks under constraints of maximum working capability of mobile stations. We prove that the performance of proposed online algorithm is no worse than e-1/e of optimal solutions. Extensive simulations are conducted to evaluate our proposed algorithm.

Data traffic in cellular networks has dramatically increased in recent years as the emergence of various new wireless applications, which imposes an immediate requirement for large network capacity. Although many efforts have been made to enhance wireless channel capacity, they are far from solving the network capacity enhancement problem. Device-to-Device (D2D) communication is recently proposed as a promising technique to increase network capacity. However, most existing work on D2D communications focuses on optimizing throughput or energy efficiency, without considering economic issues. In this paper, we propose a truthful double auction for D2D communications (TAD) in multi-cell cellular networks for trading resources in frequency-time domain, where cellular users with D2D communication capability act as sellers, and other users waiting to access the network act as buyers. Both intra-cell and inter-cell D2D sellers are accommodated in TAD while the competitive space in each cell is extensively exploited to achieve a high auction efficiency. With a sophisticated seller-buyer matching, winner determination and pricing, TAD guarantees individual rationality, budget balance, and truthfulness. Furthermore, we extend our TAD design to handle a more general case that each seller and buyer ask/bid multiple resource units. Extensive simulation results show that TAD can achieve truthfulness as well as high performance in terms of seller/buyer sanctification ratio, auctioneer profit and network throughput.

Software-defined networking (SDN) is an emerging network paradigm that simplifies network management by decoupling the control plane and data plane, such that switches become simple data forwarding devices and network management is controlled by logically centralized servers. In SDN-enabled networks, network flow is managed by a set of associated rules that are maintained by switches in their local Ternary Content Addressable Memories (TCAMs) which support high-speed parallel lookup on wildcard patterns. Since TCAM is an expensive hardware and extremely power-hungry, each switch has only limited TCAM space and it is inefficient and even infeasible to maintain all rules at local switches. On the other hand, if we eliminate TCAM occupation by forwarding all packets to the centralized controller for processing, it results in a long delay and heavy processing burden on the controller. In this paper, we strive for the fine balance between rule caching and remote packet processing by formulating a minimum weighted flow provisioning ( MWFP) problem with an objective of minimizing the total cost of TCAM occupation and remote packet processing. We propose an efficient offline algorithm if the network traffic is given, otherwise, we propose two online algorithms with guaranteed competitive ratios. Finally, we conduct extensive experiments by simulations using real network traffic traces. The simulation results demonstrate that our proposed algorithms can significantly reduce the total cost of remote controller processing and TCAM occupation, and the solutions obtained are nearly optimal.

The MapReduce programming model simplifies large-scale data processing on commodity cluster by exploiting parallel map tasks and reduce tasks. Although many efforts have been made to improve the performance of MapReduce jobs, they ignore the network traffic generated in the shuffle phase, which plays a critical role in performance enhancement. Traditionally, a hash function is used to partition intermediate data among reduce tasks, which, however, is not traffic-efficient because network topology and data size associated with each key are not taken into consideration. In this paper, we study to reduce network traffic cost for a MapReduce job by designing a novel intermediate data partition scheme. Furthermore, we jointly consider the aggregator placement problem, where each aggregator can reduce merged traffic from multiple map tasks. A decomposition-based distributed algorithm is proposed to deal with the large-scale optimization problem for big data application and an online algorithm is also designed to adjust data partition and aggregation in a dynamic manner. Finally, extensive simulation results demonstrate that our proposals can significantly reduce network traffic cost under both offline and online cases.

Our experiences in East Japan Earthquake show that disasters will cause a large-scale network interruption due to serious damage of existing network infrastructures. To enable Internet connection before network service restoration, which is usually time-consuming, we propose to use a mobile relay to carry data for several isolated communities formed after disasters. Specifically, we consider that only one community has the Internet connection, and data from Internet need to be carried to other communities by the mobile relay. The data downloading performance of each community is evaluated by the utility of obtained data minus the penalty of corresponding latency. With the objective of maximizing the poorest performance among communities, we formulate a max-min problem to optimize the trajectory of the mobile relay and its carried data volume for each community. Due to the NP-hardness of this problem, we propose a genetic algorithm by representing the trajectories of mobile relay as chromosomes that evolve to approximate the optimal solution. The fitness of each chromosome is evaluated by optimizing the data volume carried for each community. Extensive simulations are conducted to show that our proposed algorithm significantly outperforms existing algorithms.

Delay tolerant networks (DTN) have shown its great successes in many mobile applications with intermittent connectivity. To increase contact opportunities that play an important role in routing performance and energy efficiency of DTN, the approaches of deploying static relay nodes, which are also referred to as throwboxes, has been widely adopted. In this paper, we study a relay deployment problem with the objective of minimizing communication latency among all nodes in DTN, which has been little studied so far. Upon the concept of latency graph that we propose to model the message delivery latency in DTN, the problem of relay placement for latency minimization (RPLM) is defined and proven NP-hard. To solve the RPLM problem, we propose a heuristic algorithm with low complexity, and conduct extensive simulations to show that it significantly outperforms other two existing schemes.

It has been recognized that cellular network interfaces are not energy efficient because of tail energy after each transmission. Although many research efforts have been made to reduce tail energy, they ignore the dynamic of link quality caused by user mobility or network congestion, which would lead to limited improvement without quality-of-experience guarantee. In this paper, we study to minimize energy consumption of the cellular network interface with a sequence of download/upload requests. Given accurate estimation of achievable link rate, we design a dynamic-programming (DP) based algorithm to obtain the optimal solution. Without the knowledge of dynamic link quality and future requests, an online algorithm is proposed to approximate the optimal solution. Finally, we conduct extensive simulations using real traces to evaluate the performance of our proposals, and the results show that 29% energy can be saved by using our algorithm under typical network settings.

Code offloading has been proposed to improve the performance and energy-efficiency of mobile devices by sending heavy computation tasks to resourceful cloud, instead of executing all tasks on local mobile devices. Unfortunately, current code offloading techniques are not efficient enough because of high communication cost. In this paper, we propose a novel code offloading strategy with cellular traffic aggregation that aggregates offloaded codes to several mobile devices before sending them to cloud, which can significantly reduce tail energy effect. With the objective of minimizing the total cost of computation and communication, we propose an optimization framework by jointly considering code partition and traffic aggregation. Due to the hardness of this problem, we design an efficient heuristic algorithm and evaluate its performance via extensive simulations. Simulation results demonstrate that the proposed algorithm significantly outperforms existing schemes.

Software-Defined Network (SDN) is a promising network paradigm that separates the control plane and data plane in the network. It has shown great advantages in simplifying network management such that new functions can be easily supported without physical access to the network switches. However, Ternary Content Addressable Memory (TCAM), as a critical hardware storing rules for high-speed packet processing in SDN-enabled devices, can be supplied to each device with very limited quantity because it is expensive and energy-consuming. To efficiently use TCAM resources, we propose a rule multiplexing scheme, in which the same set of rules deployed on each node apply to the whole flow of a session going through but towards different paths. Based on this scheme, we study the rule placement problem with the objective of minimizing rule space occupation for multiple unicast sessions under QoS constraints. We formulate the optimization problem jointly considering routing engineering and rule placement under both existing and our rule multiplexing schemes. Via an extensive review of the state-of-the-art work, to the best of our knowledge, we are the first to study the non-routing-rule placement problem. Finally, extensive simulations are conducted to show that our proposals significantly outperform existing solutions.

The dramatic growth of mobile multimedia communications imposes new requirements on quality-of-service and energy efficiency in wireless networks. In this paper, we study the energy- and spectrum-efficient cooperative communication (ESCC) problem by exploiting the benefits of cooperative communication (CC) for mobile multimedia applications in multi-channel wireless networks. In a static network, it is formulated as a mixed-integer nonlinear programming problem. To solve this problem, we use linearization and reformulation techniques to transform it into a mixed-integer linear programming problem that is solved by a branch-and-bound algorithm with enhanced performance. To deal with the problem in dynamic networks, we propose an online algorithm with low computational complexity and deployment overhead. Extensive simulations are conducted to show that the proposed algorithm can significantly improve the performance of energy efficiency in both static and dynamic networks.

After a decade of extensive research on application-specific wireless sensor networks (WSNs), the recent development of information and communication technologies makes it practical to realize the software-defined sensor networks (SDSNs), which are able to adapt to various application requirements and to fully explore the resources of WSNs. A sensor node in SDSN is able to conduct multiple tasks with different sensing targets simultaneously. A given sensing task usually involves multiple sensors to achieve a certain quality-of-sensing, e.g., coverage ratio. It is significant to design an energy-efficient sensor scheduling and management strategy with guaranteed quality-of-sensing for all tasks. To this end, three issues are investigated in this paper: 1) the subset of sensor nodes that shall be activated, i.e., sensor activation, 2) the task that each sensor node shall be assigned, i.e., task mapping, and 3) the sampling rate on a sensor for a target, i.e., sensing scheduling. They are jointly considered and formulated as a mixed-integer with quadratic constraints programming (MIQP) problem, which is then reformulated into a mixed-integer linear programming (MILP) formulation with low computation complexity via linearization. To deal with dynamic events such as sensor node participation and departure, during SDSN operations, an efficient online algorithm using local optimization is developed. Simulation results show that our proposed online algorithm approaches the globally optimized network energy efficiency with much lower rescheduling time and control overhead.

WiFi and cellular networks are pervasively provided for mobile Internet access. Although most existing mobile devices are equipped with both WiFi and cellular network interfaces, concurrent data transmissions over these interfaces for improved throughput are not provided. In this paper, a bandwidth aggregation prototype, named Application Layer Protocol based Aggregation (ALP-A), is developed for easy use by simply installing an application in mobile devices without modifying their operating systems or drivers. It provides desired quality-of-experience (QoE), i.e., acceptable response delay to users, learned from application characteristics and users behaviors. Furthermore, we propose an online algorithm of traffic scheduling over WiFi and cellular interfaces with the objective of minimizing energy consumption while guaranteeing the QoE. Over the prototype implemented on Andriod-based smartphones, we conduct extensive experiments to show that ALP-A outperforms existing schemes significantly.

To characterize the delivery performance of message dissemination in Disruption/Delay Tolerant Networks, various methods have been proposed. However, existing work shares a common simplification that the pairwise meeting rate between any two mobile nodes is exponentially distributed. In this paper, instead of relying on such assumption, we jointly consider the deactivation-rate over relay nodes and the number of sinks deployed in the network as the primary system parameters. Then, an ODE-based theoretical framework is proposed in a stochastic manner, which enables to describe how these two parameters affect the performance of a message delivery process using controlled epidemic routing. Via extensive experiments, the high accuracy of our analytical framework are verified.

With the surging popularity of smartphones and tablets, mobile network traffic has dramatically increased in recent years. Software defined network (SDN) provides a scalable and flexible structure to simplify network traffic management. It has been shown that rule placement plays an important role in the performance of SDN. However, since most existing work considers static network topologies of wired networks, it cannot be directly applied for mobile networks. In this paper, we propose MoRule, an efficient rule management scheme to optimize the rule placement for mobile users. To deal with the challenges of user mobility and rule capacity constraints of switches, we design a heuristic algorithm with low-complexity to minimize the rule space occupation while guaranteeing that the mobile traffic processed by local switches is no less than a threshold. By conducting extensive simulations, we demonstrate that our proposed algorithm significantly outperforms random solutions under various network settings.

Multicast is widely adopted by applications in cloud data centers. Due to the shortage of IP multicast groups, many efforts have been made to share multicast groups among multiple multicast sessions. However, existing group sharing schemes would lead to a serious privacy problem that some data will be exposed to receivers not belonging to the corresponding multicast session. In this paper, we propose a novel system called EPM for privacy-preserving multicast in cloud data centers. By enhancing a popular virtual machine manager Xen, it guarantees that the multicast packets received by each physical machine can be correctly forwarded only to virtual machines as destinations. Furthermore, we study a group assignment problem based on EPM to reduce redundant multicast traffic in cloud data centers, which will be solved by an efficient heuristic algorithm. We evaluate the effectiveness of EPM by implementing it on real systems, and conduct extensive experiments to show that the proposed algorithm can significantly reduce multicast traffic.

Software-defined network (SDN) is the next generation of networking architecture that is dynamic, manageable, cost-effective, and adaptable, making it ideal for the high-bandwidth, dynamic nature of today's applications. In SDN, network management is facilitated through software rather than low-level device configurations. However, the centralized control plane introduced by SDN imposes a great challenge for the network security. In this paper, we present a secure SDN structure, in which each device is managed by multiple controllers rather than a single one as in a traditional manner. It can resist Byzantine attacks on controllers and the communication links between controllers and SDN switches. Furthermore, we design a cost-efficient controller assignment algorithm to minimize the number of required controllers for a given set of switches. Extensive simulations have been conducted to show that our proposed algorithm significantly outperforms random algorithms.

After a decade of extensive research on application-specific wireless sensor networks (WSNs), the recent development of information and communication technologies make it practical to realize software-defined sensor networks (SDSNs), which are able to adapt to various application requirements and to fully explore the resources of WSNs. In SDSNs, wireless sensor nodes can be dynamically reprogrammed for different sensing tasks via the over-the-air-programming technique. For a given sensing task, it is usually required to guarantee certain quality-of-sensing, e.g., coverage ratio. Intuitively, the more sensors are deployed with a program, the higher quality-of-sensing of the corresponding task can be achieved. However, this is at the expense of high reprogramming energy consumption. In this paper, we investigate how to design an energy-efficient reprogramming strategy with guaranteed quality-of-sensing for a sensing task. To this end, two issues will be tackled: 1) the subset of sensors that shall be reprogrammed, i.e., reprogramming sensor selection and 2) the program distribution routing. They are jointly considered and formulated as an integer linear programming (ILP) problem, based on which an algorithm with low computation complexity is then proposed. The high efficiency of our algorithm is validated by extensive simulation studies.

Software-Defined Network (SDN) is a promising network paradigm that separates the control plane and data plane in the network. It has shown great advantages in simplifying network management such that new functions can be easily supported without physical access to the network switches. However, Ternary Content Addressable Memory (TCAM), as a critical hardware storing rules for high-speed packet processing in SDN-enabled devices, can be supplied to each device with very limited quantity because it is expensive and energy-consuming. To efficiently use TCAM resources, we propose a rule multiplexing scheme, in which the same set of rules deployed on each node apply to the whole flow of a session going through but towards different paths. Based on this scheme, we study the rule placement problem with the objective of minimizing rule space occupation for multiple unicast sessions under QoS constraints.We formulate the optimization problem jointly considering routing engineering and rule placement under both existing and our rule multiplexing schemes. Finally, extensive simulations are conducted to show that our proposals significantly outperform existing solutions.

Cooperative communication (CC) can offer high channel capacity and reliability in an efficient and low-cost way by forming a virtual antenna array among single-antenna nodes that cooperatively share their antennas. It has been well recognized that the selection of relay nodes plays a critical role in the performance of multiple source-destination pairs. Unfortunately, all prior work has made an unrealistic assumption that spectrum resources are unlimited and each source-destination pair can communicate over a dedicated channel with no mutual interference. In this paper, we study the problem of maximizing the minimum transmission rate among multiple source-destination pairs using CC in a cognitive radio network (CRN). We jointly consider the relay assignment and channel allocation under a finite set of available channels, where the interference must be considered. In order to improve the spectrum efficiency, we exploit the network coding opportunities existing in CC that can further increase the capacity. Such max-min rate problems for cognitive and cooperative communications are proved to be NP-hard and the corresponding MINLP (Mixed-Integer Nonlinear Programming) formulations are developed. Moreover, we apply the reformulation and linearization techniques to the original optimization problems with nonlinear and nonconvex objective functions such that our proposed algorithms can produce high competitive solutions in a timely manner. Extensive simulations are conducted to show that the proposed algorithms can achieve high spectrum efficiency in terms of providing a much improved max-min transmission rate under various network settings.

Software-defined network (SDN) is the next generation of networking architecture that is dynamic, manageable, cost-effective, and adaptable, making it ideal for the high-bandwidth, dynamic nature of today’s applications. In SDN, network management is facilitated through software rather than low-level device configurations. However, the centralized control plane introduced by SDN imposes a great challenge for the network security. In this paper, we present a secure SDN structure, in which each device is managed by multiple controllers, not just a single as in a traditional manner, with the dynamic and isolated instance provided by the cloud. It can resist Byzantine attacks on controllers and the communication links between controllers and SDN switches. Furthermore, we study a controller minimization problem with security requirement and propose a cost-efficient controller assignment algorithm with a constant approximation ratio. From the experiment result, the secure SDN structure has little impact on the network latency, provide better security than general distributed controller, and the proposed algorithm performs higher efficiency than random assignment.

Cooperative communication (CC) has been proposed for achieving spatial diversity without requiring multiple antennas on the same node. Many efforts in exploiting the benefits of CC focus on improving the performance in terms of outage probability or channel capacity. However, the energy efficiency of CC, which is critical for the applications with energy constraints, has been little studied. In this paper, we study the lifetime maximization problem for multiple source-destination pairs using CC in multi-channel wireless networks by an optimal dynamic allocation of resources in terms of power, channel, cooperative relay, and transmission time fraction. We prove it NP-hard and formulate it as a mixed-integer nonlinear programming (MINLP) problem, which is then transformed into a mixed-integer linear programming (MILP) problem using linearization and reformulation techniques. By exploiting several problem-specific characteristics, a time-efficient branch-and-bound algorithm is proposed to solve the MILP problem. Extensive simulations are conducted to show that the proposed algorithm can significantly improve the performance of energy efficiency over existing solutions.

Multicast is an important mechanism in modern wireless networks and has attracted significant efforts to improve its performance with different metrics including throughput, delay, energy efficiency, etc. Traditionally, an ideal loss-free channel model is widely used to facilitate routing protocol design. However, the quality of wireless links is affected or even jeopardized resulting in transmission failures by many factors like collisions, fading or the noise of environment. In this paper, we propose a reliable multicast protocol, called CodePipe, with energy-efficiency, high throughput and fairness in lossy wireless networks. Building upon opportunistic routing and random linear network coding, CodePipe can not only eliminate coordination between nodes, but also improve the multicast throughput significantly by exploiting both intra-batch and inter-batch coding opportunities. In particular, four key techniques, namely, LP-based opportunistic routing structure, opportunistic feeding, fast batch moving and inter-batch coding, are proposed to offer significant improvement in throughput, energy-efficiency and fairness. Moreover, we design an efficient online extension of CodePipe such that it can work in a dynamic network where nodes join and leave the network as time progresses. We evaluate CodePipe on ns2 simulator by comparing with other two state-of-art multicast protocols, MORE and Pacifier. Simulation results show that CodePipe significantly outperforms both of them.

Most existing research studies in cooperative communication are based on a half-duplex assumption. Motivated by recent successes in hardware implementation of wireless full-duplex transmission, we propose a full-duplex cooperative communication (FDCC) approach to maximize the minimum transmission rate among a set of users to a common destination with the help of a dedicated relay. Under the consideration of hardware cost, only the relay node requires full-duplex wireless equipment in our design. We derive the achievable transmission rate for the proposed FDCC scheme under both amplify-and-forward (AF) and decode-and-forward (DF) modes. Further, as the transmission scheduling of users plays a critical role in determining the achievable transmission rate in FDCC, we formulate the max-min rate scheduling problem as a nonconvex mixed integer nonlinear programming (MINLP) problem. By applying linearization and convex approximation techniques, we propose an optimal algorithm based on a branch-and-bound framework to solve the problem efficiently. Extensive simulation results show that FDCC can significantly improve the transmission rate as compared with direct transmission and half-duplex cooperative communication (HDCC).

In this paper, we consider the reliable broadcast and multicast lifetime maximization problems in energy-constrained wireless ad hoc networks, such as wireless sensor networks for environment monitoring and wireless ad hoc networks consisting of laptops or PDAs with limited battery capacities. In packet loss-free networks, the optimal solution of lifetime maximization problem can be easily obtained by tree-based algorithms. In unreliable networks, we formulate them as min-max tree problems and prove them NP-complete by a reduction from a well-known minimum degree spanning tree problem. A link quality-aware heuristic algorithm called Maximum Lifetime Reliable Broadcast Tree (MLRBT) is proposed to build a broadcast tree that maximizes the network lifetime. The reliable multicast lifetime maximization problem can be solved as well by pruning the broadcast tree produced by the MLRBT algorithm. The time complexity analysis of both algorithms is also provided. Simulation results show that the proposed algorithms can significantly increase the network lifetime compared with the traditional algorithms under various distributions of error probability on lossy wireless links.

Cooperative communication (CC) has been proposed to increase the wireless channel capacity and reliability by employing multiple single-antenna nodes to form a virtual antenna array. Many efforts focus on exploiting the benefits of CC among multiple source-destination pairs with an unrealistic assumption that each of them communicates over a dedicated channel without interference. In this paper, we investigate the transmission scheduling problem for multiple source-destination pairs under the assistance of a set of dedicated relay nodes on a single channel. By applying the protocol interference model, we propose a concept of cooperative link to characterize the interference regions of CC. Due to the NP-completeness of optimal scheduling, LP (linear programming) based heuristic algorithms are proposed to maximize the minimum transmission rate under a given relay assignment. Then, without specifying a relay node for each source-destination pair, we study the max-min rate problem by jointly considering transmission scheduling and relay assignment. Heuristic algorithms are proposed to solve this more challenging problem. Finally, extensive simulations are conducted to show that the proposed algorithm outperforms direct transmission substantially.

Cooperative communication (CC) has been proposed to achieve spatial diversity without requiring multiple antennas on a single device. Many efforts in exploiting the benefits of CC focus on improving the performance in terms of outage probability and channel capacity. However, the energy efficiency of CC, which is critical for the applications with energy constraints, has been little studied. In this paper, we study the problem of energy-efficient CC in cognitive radio networks by taking power control, relay assignment and channel allocation into consideration. We obtain the optimal power control for a single source-destination pair. After that, a heuristic algorithm is proposed to solve the the relay assignment and channel allocation problem for multiple source-destination pairs. Extensive simulations are conducted to show that the proposed algorithm can significantly improve the performance of energy efficiency over the direct transmission approach.

We study a fundamental problem in determining the multicast capacity in energy-constrained wireless networks with lossy transmission links. The multicast capacity in our paper is defined as the maximum number of packets that can be disseminated from the source and successfully received by all multicast destinations. To explore the expected multicast capacity, we propose a framework for the joint optimization of both dynamic power control and error control. In our framework, the lossy wireless transmission links are characterized by the Rayleigh fading model, which reveals the realistic relationship among link quality, transmission power, and path attenuation. Under this model, we exploit the reliability gain of random linear network coding, also referred to as intrabatch coding in this paper, by disseminating data in batches. To maximize multicast capacity, another type of network coding opportunities across batches, referred to as interbatch coding, is also explored. Our analytical framework based on intrabatch and interbatch network coding eventually leads to a linear programming formulation that is proved to obtain the optimal multicast capacity. To approach the theoretical results in practice, we propose an algorithm called DMCC that exploits the intrabatch and interbatch coding via dynamically constructing bottleneck trees. Extensive simulations are conducted to show that its performance is very close to the optimal solution.

Cooperative communication provides an efficient and low-cost way to achieve spatial diversity without deploying multiple antennas on each node in wireless networks. In a channel-constrained environment, such as cognitive radio networks, the channel allocation as well as relay assignment have been identified as two critical factors in determining the performance of multiple source-destination pairs. However, the advantage of channel diversity has little been exploited in such networks under a simplified model where the transmissions within a cooperative communication pair are on a common channel. In this paper, we consider a fine-grained channel allocation scheme that the source and the relay can work on different channels to complete a signal transmission. We study its performance gain in maximizing the minimum throughput among multiple source-destination pairs in channel-constrained wireless networks with a number of dedicated relay nodes. This problem is proved to be NP-hard and an online algorithm is proposed for a dynamic wireless network where the accessible channels of each node may vary from time to time. Extensive simulations are conducted to show that the proposed fine-grained channel allocation scheme can effectively improve the performance under various network settings.

Cooperative communication (CC) can offer high channel capacity and reliability in an efficient and low-cost way by forming a virtual antenna array among single-antenna nodes that cooperatively share their antennas. It has been well recognized that the selection of relay nodes plays a critical role in the performance of multiple source-destination pairs. Unfortunately, all prior work has made an unrealistic assumption that each source-destination pair communicates over a dedicated channel with no mutual interference. In this paper, we study the problem of capacity maximization using cooperative communication in a cognitive radio network by jointly considering the relay assignment and channel allocation under a finite set of available channels, where the interference must be considered. It is proved to be NP-hard and a heuristic algorithm is proposed. Moreover, we exploit the network coding opportunities existing in CC that can further increase the capacity. Extensive simulations are conducted to show that the proposed algorithms can achieve high total capacity under various network settings.

Multicast is an important mechanism in modern wireless networks and has attracted significant efforts to improve its performance with different metrics including throughput, delay, energy efficiency, etc. Traditionally, an ideal loss-free channel model is widely used to facilitate routing protocol design. However, the quality of wireless links would be affected or even jeopardized by many factors like collisions, fading or the noise of environment. In this paper, we propose a reliable multicast protocol, called CodePipe, with advanced performance in terms of energy-efficiency, throughput and fairness in lossy wireless networks. Built upon opportunistic routing and random linear network coding, CodePipe not only simplifies transmission coordination between nodes, but also improves the multicast throughput significantly by exploiting both intra-batch and inter-batch coding opportunities. In particular, four key techniques, namely, LP-based opportunistic routing structure, opportunistic feeding, fast batch moving and inter-batch coding, are proposed to offer substantial improvement in throughput, energy-efficiency and fairness. We evaluate CodePipe on ns2 simulator by comparing with other two state-of-art multicast protocols, MORE and Pacifier. Simulation results show that CodePipe significantly outperforms both of them.

In traditional stop-and-wait strategy for reliable communications, such as ARQ, retransmission for the packet loss problem would incur a great number of packet transmissions in lossy wireless ad-hoc networks. We study the reliable multicast lifetime maximization problem by alternatively exploring the random linear network coding in this paper. We formulate such problem as a min-max problem and propose a heuristic algorithm, called maximum lifetime tree (MLT), to build a multicast tree that maximizes the network lifetime. Simulation results show that the proposed algorithms can significantly increase the network lifetime when compared with the traditional algorithms under various distributions of error probability on lossy wireless links.

The reliable broadcast and multicast lifetime maximization problems in energy-constrained wireless ad-hoc networks are considered in this paper. In packet loss-free networks, the optimal solution of lifetime maximization problem can be easily obtained by tree based algorithms. In unreliable networks, we formulate them as min-max tree problems. A link quality-aware heuristic algorithm called MLRBT (Maximum Lifetime Reliable Broadcast Tree) is proposed to build a broadcast tree that maximizes the network lifetime. The reliable multicast lifetime maximization problem can be solved as well by pruning the broadcast tree produced by the MLRBT algorithm. Simulation results show that the proposed algorithms can significantly increase the network lifetime compared with the traditional algorithms under various distribution of unreliable communication links.

Thread-Level Speculation (TLS) has shown great promise as an automatic parallelization technique to achieve high level performance by partitioning a sequential program into threads, which are expected to be optimistically executed in parallel. In this paper, we propose a load-balancing approach to save energy using dynamic voltage scaling. By scaling the voltage of processors running short threads, energy consumption on these processors can be reduced while keeping a similar speedup of the overall system. Two voltage selection strategies have been investigated. With the assistance of some profiling tools, we propose a static voltage selection algorithm that can minimize energy consumption without degrading the parallelism provided by the pure TLS. The other dynamic algorithm selects voltage for each thread with prediction during the execution. Our experimental results show that its energy consumption is reduced to 78.8% and execution time is stretched to 1.07 times, on average, of the pure TLS in a 16-core CMP processor.