Decentralized Federated Distillation (DFD) has emerged as a significant research direction since it not only supports heterogeneous model training but also naturally avoids privacy risks and communication bottlenecks stemming from the central server. DFD shows great potential in various training scenarios, especially in cross-silo federated environments. Existing DFD algorithms have limitations, including reliance on public datasets for distillation, insufficient privacy protection, and high communication overhead. This paper introduces Ring-Distill, a novel DFD algorithm designed specifically for cross-silo federated environments, effectively addressing the aforementioned limitations. Ring-Distill allows clients to complete federated training within a few global epochs (i.e., communication rounds) without sharing public datasets, e.g.,$N$global epochs for a system involving$N$clients. To protect privacy and further reduce communication overhead, Ring-Distill contains a privacy-oriented automatic model pruning mechanism (PAMP) which can automatically create the privacy-preserving compressed model called proxy model for each client. These proxy models are employed for client-to-client distillation during each global epoch. Furthermore, Ring-Distill contains a historical model-based distillation (HMD) mechanism that allows local models to transfer knowledge from multiple historical proxy model replicas stored locally. Due to historical proxy model replicas, the HMD mechanism not only improves the distillation performance but also effectively mitigates client dropout issues. Theoretical analysis and comprehensive experiments show that Ring-Distill has significant advantages in terms of accuracy, privacy, and communication cost.

Knowledge Distillation (KD) is considered as an efficient way to replace the parameter averaging in federated learning, aiming to handle the clients with heterogeneous model architectures. Relying on the prepared distillation datasets across clients and the server, KD may encounter impractical difficulties in real-world implementations. Existing works explore the data-free KD in federated learning, which generates the distillation datasets on-site. However, the distillation datasets with global data distribution generated by these state-of-the-art schemes cannot be adapted to local non-IID data. In this article, we propose a new Personalized Data-Free Knowledge Distillation, namely PDKD, for federated learning under heterogeneous models and data. PDKD solves the problem of model drift caused by the inconsistent distribution of distillation datasets and the local data by generating personalized distillation datasets for each client while protecting client data privacy. In addition, we design a distillation dataset update scheme that maximizes the difference between teacher and client outputs on distillation datasets to accomplish deeper knowledge transfer. Furthermore, in order to accomplish the co-evolution of the teacher model and the clients’ model, PDKD incorporates a mutual distillation scheme. Numerous experiments show that PDKD significantly outperforms several state-of-the-art algorithms, with an 18% improvement in prediction accuracy and has a much lower communication cost than the compared algorithms.

We focus on the problem of fusing two or more heterogeneous large language models (LLMs) to leverage their complementary strengths.One of the challenges of model fusion is high computational load, specifically in fine-tuning or aligning vocabularies.To address this, we propose Cool-Fusion, a simple yet effective approach that fuses the knowledge of source LLMs, which does not require training.Unlike ensemble methods, Cool-Fusion is applicable to any set of source LLMs that have different vocabularies.To overcome the vocabulary discrepancies among LLMs, we ensemble LLMs on text level, allowing them to rerank the generated texts by each other with different granularities.Extensive experiments have been conducted across a variety of benchmark datasets.On GSM8K, Cool-Fusion increases accuracy from three strong source LLMs by a significant margin of 17.4%.

Sharding is a widely adopted solution for blockchain scalability, yet its performance is often constrained by the high overhead of cross-shard transaction (CTX) processing. Existing systems typically rely on costly coordination protocols for atomic execution of CTXs, leading to high latency and reduced throughput. While recent advances in transaction allocation have minimized the proportion of CTXs, we argue that a critical aspect of sharding performance has been overlooked: the optimization of intra-shard transactions (ITXs), which now dominate the workload. In this paper, we introduce B-Shard, a novel sharding architecture that reorients the optimization objective toward maximizing ITX throughput, leading to significantly improved overall performance. The B-Shard design consists of a two-layer architecture: a global Beacon Shard for cross-shard coordination and multiple Worker Shards for transaction processing. By drawing inspiration from off-chain solutions, Worker Shards can rapidly process and finalize ITXs locally through batched commits to the Beacon Shard. The full potential of this architecture is revealed by a synergistic mechanism: our Balancing Transaction Scheduling algorithm (BATCH) buffers CTXs to create extended “pure ITX processing windows,” which are then processed with maximum efficiency by a lightweight intra-shard certification protocol. This synergy substantially reduces system overhead and maximizes throughput. Our evaluation against traditional twophase commit (2PC) and modern protocols demonstrates that B-Shard achieves a significant increase in total throughput. By accepting a minor latency increase for some CTXs, we achieve an overall reduction in average confirmation latency, validating the effectiveness of our strategy.

Collecting data from simulated scenarios for training robotic skills provides a safer and more controllable alternative to real-world environments. However, it demands considerable effort, including the manual construction of simulation environments, the careful design of tasks, and the challenge of obtaining effective trajectories. These limitations hinder the efficiency of data collection from simulated scenarios. In this paper, we leverage the prior knowledge of Large Language Models (LLMs) and Large Multimodal Models (LMMs) to generate simulated scenarios and embodied tasks. We introduce a novel framework, ASCENT (Autonomous Skill learning toward Complex Embodied tasks with fouNdaTion models), designed to efficiently accomplish these tasks and generate trajectory data. ASCENT features a fully autonomous skill learning mechanism based on AI agent. During task training, the AI agent identifies suitable atomic skills from an atomic skill library to either directly complete the task or serve as an initial policy for further training. Newly acquired atomic skills are subsequently added to the library. To address training failures and enhance efficiency, the AI agent uses an LLM to automatically optimize the skill training process based on feedback received from simulations. Experimental results indicate that the number of training steps required for learning new tasks can be reduced by up to 65.9 %.

Payment channels enforce deposit locking at each hop to maintain atomicity guarantees in multi-hop off-chain transactions. Locked deposits remain unavailable for other transactions during this period. In cases of transaction failure or malicious delays, funds may be locked and occupied for a long time, severely degrading capital liquidity. To mitigate this issue, we propose a transaction binding mechanism based on path-channel overlap to reduce locked fund requirements. Our solution defines binding conditions for transaction pairs, implements a three-phase consensus protocol based on tripartite negotiation for establishing transaction bindings, and develops an atomic cross-path payment protocol with a mutual precondition mechanism. To evaluate the approach's effectiveness, we conduct comprehensive experiments using real-world transaction datasets from the Ripple network. Experimental results demonstrate a significant reduction in locked collateral, an notable improvement in transaction success ratio, while maintaining negligible increase in latency and guaranteed transaction security.

Modern cloud service providers increasingly rely on extensive natural language database queries for daily operational workflows, including system maintenance, fault recovery, and customer ticket resolution, all of which are critical for ensuring optimal data availability, service reliability, and customer experience. However, the specialized nature of cloud platform data inherently complicates this task: current practices either depend heavily on Site Reliability Engineer (SRE) experts for manual SQL query or require intensive data collection and manual annotation to train domain-specific Natural Language to SQL (NL2SQL) models for automation-both approaches suffering from high labour costs and limited scalability. To address these challenges, this paper introduces ASASQL, Active Learning-based Data Synthesis and Annotation Strategy for SQL Generation, a novel framework designed to automatically generate training data and select data for annotation to train domain-specific nl2sql models. ASASQL significantly reduces the annotation requirement for training accurate, cost-efficient, and scalable automatic data querying models. Experiments on realworld production datasets from a major cloud service provider demonstrate that ASASQL can reduce the amount of data annotation needed to achieve specific accuracy by $38.7 \%$.

Money laundering using cryptocurrency poses significant threats to the blockchain ecosystem. Due to the decentralized and anonymous nature of cryptocurrencies, detecting such laundering activities is difficult. Although substantial research has been conducted, almost all existing methods detect cryptocurrency laundering from an individual perspective, ignoring the fact that money laundering is typically a group behavior. Group information should be very helpful in laundering behavior analysis, but such laundering groups are hard to be recognized due to anonymity and diversity of purposes of cryptocurrency transactions. To address this challenge, we design a multi-persona grouping algorithm that can effectively group accounts into persona subgraphs. Then, we extract two subgraph features: cycle basis number and cycle overlapping ratio, and build an unsupervised model to evaluate laundering scores of each subgraph. Extensive experiments on both synthetic and real-world datasets demonstrate that, compared with existing methods, our proposed method can improve detection accuracy by 17.4 percentage points on average. To the best of our knowledge, this is the first work on group-based detection of cryptocurrency laundering.

GPU clusters have been widely used to co-locate various deep learning (DL) workloads in a multi-tenant way. Although such resource sharing can significantly reduce training cost, resource contention and interference among co-located workloads make task scheduling very complex and challenging. To simplify the scheduling problem, existing algorithms usually divide the procedure of scheduling into two sub-tasks, i.e., task placement and resource allocation, and allocate resources according to pre-defined and fixed resource demands. However, such a paradigm significantly constrains the selection of potential scheduling solutions. In this article, we present MAIFS, a novel multi-agent reinforcement learning based scheduling algorithm that handles task placement and resource allocation integratedly, and allows fungible resource allocation based on resource sensitivity of DL workloads. The core of MAIFS lies in two mechanisms. The multi-agent attention mechanism is designed to learn and share inter-related resource state features observed from different agents, which enables agents to explore fungible resource allocation solutions. The dynamic coordination graph mechanism is designed for coordinating interactive task placement decisions of agents during integrated scheduling, so as to mitigate potential task conflicts. Simulated experiments using two large scale production DL workload traces and physical deployment experiments based on a Kubernetes based GPU cluster show that MAIFS can outperform state-of-the-art scheduling algorithms by up to 44% in terms of makespan and 46% in terms of job completion time (JCT).

Federated Distillation (FD), a popular variant of Federated Learning (FL), has attracted researchers' attention due to its ability to support heterogeneous model training. Generally, FD allows clients to upload logits associated with public datasets for knowledge transfer, yet logits may pose privacy risks. In this study, we provide the first demonstration of the impact of privacy risks caused by logits. Specifically, we design a data reconstruction attack against logits named L-Attack which can reveal sensitive information about the target client without access to the target model. Via the zeroth-order optimization technique, L-Attack involves training a server-side generator that unveils certain features of private data owned by the target client. To defend against L-Attack, we propose a label aggregation-based FD algorithm called LabelAvg which allows clients to upload predicted hard labels for knowledge transfer instead of logits. Due to the insufficient information in labels for distillation, LabelAvg provides a voting-based label smoothing mechanism that enables the server to construct smooth labels from received labels. The generated smooth labels which stand for the consensus among all clients, indicate the approximate probability distribution. Thus, these smoothed labels bear a striking similarity to logits and can be used for distillation. Analysis and experimental results prove LabelAvg is superior to baselines in terms of accuracy, privacy, and communication data volume.

Communication cost is a main challenge in Federated Learning (FL). Gradient sparsification is one of the effective ways to reduce communication data volumes by allowing clients to send only a small portion of gradient elements to the server. Existing gradient sparsification methods, such as Top-K, analyze and decide the contribution of gradient elements for uploading by comparing their values. However, gradient elements with large values may not necessarily contain more information. To improve the accuracy while effectively reducing communication costs, in this paper, we propose a novel gradient sparsification-based FL algorithm called FedLOC, which sparsifies gradients not only based on the value of the gradient elements but also on the layer output of the local model. In particular, we design a two-phase compression operation to reduce the communication data volumes. First, FedLOC sparsifies gradients based on their values to construct initial compressed gradients. Subsequently, since the performance of the local model is related to the output of each layer, FedLOC analyzes the contribution of initial compressed gradient elements and constructs masks for the second sparsification operation based on each layer’s extend output derived from the original layer output. Convergence analysis and experimental results show that FedLOC can achieve superior performance in terms of accuracy and convergence when effectively reducing data communication volumes.

Workload prediction plays a crucial role in resource management of large scale cloud datacenters. Although quite a number of methods/algorithms have been proposed, long-term changes have not been explicitly identified and considered. Due to shifty user demands, workload re-locations, or other reasons, the “resource usage pattern” of a workload, which is usually quite stable in a short-term view, may change dynamically in a long-term range. Such long-term dynamic changes may cause significant accuracy degradation for prediction algorithms. How to handle such long-term dynamic changes is an open and challenging issue. In this article, we propose Evolution Graph for Workload Prediction (EvoGWP), a novel method that can predict long-term dynamic changes using a delicately designed graph-based evolution learning algorithm. EvoGWP automatically extracts shapelets to explicitly identify resource usage patterns of workloads in a fine-grained level, and predicts workload changes by considering factors in both temporal and spatial dimensions. We design a two-level importance based shapelet extraction mechanism to mine new usage pattern changes in temporal dimension, and design a novel evolution graph model to fuse the interference among resource usage patterns of different workloads in spatial dimension. By combining temporal extraction of shapelets from each single workload and spatial interference of shapelets among different workloads, we then design a spatio-temporal GNN-based encoder-decoder model to predict the long-term dynamic changes of workloads. Experiments using real trace data from Alibaba, Tencent and Google show that EvoGWP improves the prediction accuracy by up to 58.6% over the state-of-the-art prediction methods. Moreover, EvoGWP can outperform the state-of-the-art prediction methods in terms of model convergence. To the best of our knowledge, this is the first work that explicitly identifies fine-grained workload resource usage patterns to accurately predict long-term dynamic changes of workloads.

Scalability is one of the main problems limiting blockchain applications. Most recent research on blockchain scalability has focused on improving the consensus layer, but the most advanced consensus protocols can only reach a few thousand transactions per second(tps), which is far below the capacity of typical distributed databases. With the use of blockchain sharding technology, the time overhead corresponding to the execution layer and the consensus layer in the blockchain is gradually reduced. Additionally, improving the efficiency of transaction execution could incentivize nodes to actively verify transactions and enhance blockchain security. In this paper, we propose a Sharding Based Parallel Execution of Block Transactions (SPEx-Tran) framework in blockchain, which transforms the sequential execution of intra-block transactions into parallel execution to improve the performance of blockchain. In addition, based on the characteristics of blockchain sharding transactions, we further propose a Blockchain Transaction Scheduling (BCTS) algorithm to improve the performance of blockchain by optimizing the parallel execution process of transactions. In the BCTS algorithm, the mechanism of local multi-processors sharing the Conflict Transactions cache Table (CTT) simplifies the verification process of conflicting transactions, thus improving the blockchain performance. Experimental results show that the throughput of blockchain using the SPExTran framework and BCTS algorithm is significantly higher than that of typical sharding technologies.

Function as a Service (FaaS) has become a popular computing paradigm. FaaS shifts the burden of resource management to cloud providers and enables users to focus only on the logic of the code. At the same time, FaaS has thrived due to its scalability and billing strategies. However, cold start has been always a major challenge for FaaS which causes additional delay. Existing cold start mitigation methods ignore dependencies among functions resulting in bloated function combinations and consequently unnecessary resource consumption. In this paper, we propose FaaSCom, a FaaS scheduler which mitigates cold start problem from the perspective of preallocating containers. We mine causality among functions and divide the function dependency graph into communities. By scheduling containers according to the function communities, we can reduce the number of containers provisioned for functions to decrease resource consumption while not increasing cold start occurrences. Besides, we design an extended histogram policy to mitigate cold starts for long-time interval functions. We evaluate our method on an industrial serverless dataset. Compared with baseline methods, FaaSCom reduces the cold start rate by 8.4% while saving 50% memory consumption.

Cryptocurrency phishing scams is a significant treat to Ethereum, one of the most popular blockchain platforms. Most of existing Ethereum phishing detection methods are based on traditional machine learning or graph representation learning, which mostly rely on only statistical and structural features in local scope. In this paper, we propose Multi-transaction-view Graph Attention Network (MTvGAT), a novel phishing scam detection model that can make use of transaction patterns of different scopes. To obtain global-view information, we apply graph clustering and construct the global-view graph with multiple clusters, including all the nodes of the original transaction network. To obtain local view information, we apply neighborhood sampling, and construct local-view graphs with target nodes and their neighborhood nodes. Then, node features, edge features, and attention coefficients are aggregated to merge multi-view information into representation of nodes. We further combine global-view and local-view representations to finally identify phishing addresses from target nodes. Extensive experiments demonstrate that the proposed method can outperform existing ones with significant improvement.

With the development of intelligent transportation system, it is urgent to transmit and analyze traffic data based on Internet of Vehicles. Federated learning has the characteristics of privacy protection, distributed learning and making full use of the computing power of local devices, which is very suitable for IoV's streaming big data analysis. Classical federated learning has only one global model, thus it can't be applied directly to IoV, where different region may have different data distribution. It is preferable that model of different region can be trained separately. Besides, the high mobility of vehicle nodes allows them to enter other areas while collecting data. This paper tries to solve the above issues. Firstly, a new vehicle-edge-cloud three-layer federated learning architecture for Io V, the Mobile Vehicle Multiple Access Federated Online Learning (MVMAFOL) algorithm, is proposed, where each vehicle terminal can upload the local model to multiple edge servers to maximize the use of its local model and effectively improve the generalization ability of the global model. Besides, a Model Similarity Vehicle Multiple Access (MSVMA) algorithm is proposed to improve the performance of MVMAFOL in the regional model aggregation phase, and an Edge Cloud Parameter Amendment (ECPA) algorithm is proposed for the global model aggregation phase. Simulation results show that this scheme can effectively improve the accuracy of the trained model.

Client selection (a.k.a., client sampling) is one of the hot topics in Federated Learning (FL). In each communication round, selecting some clients to participate in aggregation can effectively reduce the communication overhead caused by exchanging model parameters. However, due to statistical heterogeneity in FL, selecting clients randomly may affect the performance of aggregated global models. existing approaches regarding client selection firstly cluster clients and then sample(select) some representative clients from each cluster. However, these clustering-based approaches may be either time-intensive or high complexity. To address these issues, In this paper, we introduce Bi-level Sampling, a clustering-based approach for client selection. After multinomial distribution sampling, Bi-level Sampling clusters clients based on weighted per-label mean class scores and then selects participating clients for federated learning in each round. Bi-level Sampling can lead to better client representativity and the reduced variance of the client’s stochastic aggregation weights in FL. Our approach can be integrated into typical FL frameworks. Experimental results show that, compared with state-of-the-art approaches, our approach demonstrates significantly more stable and accurate convergence behavior-getting higher test accuracy and less training time, especially in highly Non-IID settings.

Data heterogeneity across clients is one of the key challenges in Federated Learning (FL), which may slow down the global model convergence and even weaken global model performance. Most existing approaches tackle the heterogeneity by constraining local model updates through reference to global information provided by the server. This can alleviate the performance degradation on the aggregated global model. Different from existing methods, we focus the information exchange between clients, which could also enhance the effectiveness of local training and lead to generate a high-performance global model. Concretely, we propose a novel FL framework named FedCME by client matching and classifier exchanging. In FedCME, clients with large differences in data distribution will be matched in pairs, and then the corresponding pair of clients will exchange their classifiers at the stage of local training in an intermediate moment. Since the local data determines the local model training direction, our method can correct update direction of classifiers and effectively alleviate local update divergence. Besides, we propose feature alignment to enhance the training of the feature extractor. Experimental results demonstrate that FedCME performs better than FedAvg, FedProx, MOON and FedRS on popular federated learning benchmarks including FMNIST and CIFAR10, in the case where data are heterogeneous.

Co-location, which deploys long running applications and batch-processing applications in the same computing cluster, has become a promising way to improve resource utility for large cloud datacenters. However, co-location brings huge challenges to task scheduling because different types of workloads may affect each other. Existing works on task scheduling rarely focus on the scenario of co-location. This article presents Co-ScheRRL, a scheduling algorithm delicately designed for co-located workloads. Co-ScheRRL consists of two major mechanisms: i) a self-attention encoding mechanism which encodes and represents states of the computing cluster as a set of embedding feature vectors; ii) a deep reinforcement learning (DRL) relational reasoning mechanism which calculates and compares different scheduling actions under different co-located workloads pattern via DRL feedback reward signals based on these feature vectors. Our two mechanisms can tackle complicatedly and dynamically varying behaviors of co-located workloads. With the help of these two mechanisms, Co-ScheRRL is able to construct high-quality scheduling policies. Trace-driven simulation demonstrates that Co-ScheRRL outperforms existing scheduling algorithms in terms of makespan by more than 38.4% and throughput by more than 166.7%.

Payment channel, which allows two parties to perform micropayments without involving the blockchain, has become a promising proposal to improve the scalability of decentralized ledgers. Payment channels have been extended to payment networks to utilize existing channels as intermediary links to route coins to others. However, payments through multiple channels bear nontrivial overheads, including time cost, charges at intermediate nodes, etc. In this paper, we propose channel hub, a novel off-chain system to shorten the routing path for the payment network. Channel hub allows transferring coins directly from one payment channel to another within the same hub, so that it can be viewed as a shortcut device for the underlying payment network. Compared with payment hub in literature, which connects participant nodes, our channel hub is more efficient and lightweight, because it connects channels already constructed by the payment network. Based on channel hub, we design a new protocol named Boros to perform secure off-chain cross-channel transfers. We present the security definition of Boros and formally prove its security using the UC-framework. To demonstrate the feasibility of Boros, we develop a proof-of-concept prototype running on Ethereum. Performance evaluation shows that Boros can effectively shorten the off-chain payment path.

This study is motivated by the maximum connected coverage problem (MCCP), which is to deploy a connected UAV network with given$K$UAVs in the top of a disaster area such that the number of users served by the UAVs is maximized. The deployed UAV network must be connected, since the received data by a UAV from its served users need to be sent to the Internet through relays of other UAVs. Motivated by this application, in this paper we study a more generalized problem – the$h$-hop independently submodular maximization problem, where the MCCP problem is one of its special cases with$h=4$. We propose a$\frac {1-1/e}{2h+3}$-approximation algorithm for the$h$-hop independently submodular maximization problem, where$e$is the base of the natural logarithm. Then, one direct result is a$\frac {1-1/e}{11}$-approximate solution to the MCCP problem with$h=4$, which significantly improves its currently best$\frac {1-1/e}{32}$-approximate solution. We finally evaluate the performance of the proposed algorithm for the MCCP problem in the application of deploying UAV networks, and experimental results show that the number of users served by deployed UAVs delivered by the proposed algorithm is up to 12.5% larger than those by existing algorithms.

Resource allocation has been always a major concern of cloud providers. Workload prediction can effectively improve resource utility by providing information about resources available in the future. Machine learning-based workload prediction has been widely studied and deployed in large-scale clouds. However, workload prediction in Inter-Cloud environments has not been considered. Since more and more cloud services are orchestrated by containers (or virtual machines) distributed in multiple clouds, intelligent models for workload prediction should be trained based on trace data across clouds. Due to concerns raised by data privacy, the trace data of clouds may not be shared with each other, and general distributed training is not applicable. In this paper, we propose a general and flexible framework, namely EFL-WP, for training workload prediction models in the Inter-Cloud environment. EFL-WP adopts the federated learning approach and allows cloud providers to collaborate to train prediction models without sharing traces. EFL-WP considers the difference among workloads in different cloud (i.e., the Non-IID characteristic of traces) by using two novel techniques: participant selection mechanism and multi-global models aggregation. The former can prevent some local models trained on non-IID traces from participating in global aggregation, which is beneficial to the global model. The latter allows the coordinator to aggregate several global models according to the difference among local models. To further improve accuracy, EFL-WP adopts an ensemble inference strategy. Experimental results show that the proposed framework can be superior to baselines on both Alibaba dataset and Tencent Games Traces.

This study is motivated by the maximum connected coverage problem (MCCP), which is to deploy a connected UAV network with given K UAVs in the top of a disaster area such that the number of users served by the UAVs is maximized. The deployed UAV network must be connected, since the received data by a UAV from its served users need to be sent to the Internet through relays of other UAVs. Motivated by this application, in this paper we study a more generalized problem – the h-hop independently submodular maximization problem, where the MCCP problem is one of its special cases with h = 4. We propose a $\frac{{1 - 1/e}}{{2h + 3}}$-approximation algorithm for the h-hop independently submodular maximization problem, where e is the base of the natural logarithm. Then, one direct result is a $\frac{{1 - 1/e}}{{11}}$-approximate solution to the MCCP problem with h = 4, which significantly improves its currently best $\frac{{1 - 1/e}}{{32}}$-approximate solution. We finally evaluate the performance of the proposed algorithm for the MCCP problem in the application of deploying UAV networks, and experimental results show that the number of users served by deployed UAVs delivered by the proposed algorithm is up to 12.5% larger than those by existing algorithms.

Welcome to the 18th IEEE International Conference on Mobility, Sensing and Networking (MSN 2022)! Though the conference is finally determined to be held in a virtual format due to the COVID-19 pandemic, an attractive program with keynote addresses, panel discussions, technical sessions, and workshops has been successfully organized.

The co-location of online services and offline tasks has become very popular in data centers, which can largely improve resource utilization. Scheduling co-located offline tasks is challenging due to the interference with online services. Existing co-location scheduling algorithms try to find the best combination of different workloads to avoid performance interference and maximize the utilization of data centers, but few of them take the time varying resource constraints into account. We propose a heuristic algorithm named interval matching scheduling algorithm based on the idea that the time series of available resources and task scheduling can be regarded as interval endpoints. The proposed scheduling algorithm makes decisions based on a scoring method that calculates the matching degrees of the tasks and the changing resource series. The experimental results show that the proposed algorithm has achieved better performance under different parameter settings comprehensively.

Federated Learning (FL) has attracted a lot of attention from both academia and industry. In FL, user data no more need to be transmitted to the data center and each client device trains the deep model using its personal data so as to protect user privacy from being revealed. There exists two kinds of FL architecture, cloud based and edge based. Researchers have proposed the client-edge-cloud hierarchical FL system combining their advantage together to take the full advantage and avoid the defects. To improve the utilization ratio of local model parameters and data, so that enhance every single edge model accuracy, and ultimately achieve a more outstanding global model, we propose an algorithm Client Model Multiple Access (CMMA). CMMA allows clients associate with a set of edge servers, and uploads its training results to all servers in the set. That said, the local model of one client is reused by multiple edge servers. Such reuse can improve model performance particularly when the number of clients is small. Empirical experiments demonstrate the superiority of our scheme in different datasets, CNN models and data distributions. The results have not just shown the general suppress CMMA against hierarchical FL, but also validate the advantage of CMMA guaranteeing global model accuracy in unstable network or short of clients dilemma.

Mobile edge computing (MEC) is a key technology that enables the deployment of applications (or services) at the proximity of mobile users. However, the performance of mobile edge computing is sensitive to the quality and availability of underlying connection links. It is still unclear to what extent intermittent connectivity affects the performance of mobile edge computing. In this paper, we make the first attempt to quantify the influence of intermittent connectivity on mobile edge computing from a theoretical perspective. Specifically, we propose an analytical framework based on discrete-time Markov chain and derive a closed-form expression of the task processing time under different network conditions. Our model can be further extended to account for the case with group task arrivals. We also conduct extensive simulations to examine the accuracy of our proposed analytical models with both synthetic and real-world user mobility traces. The results show that our model can well capture the influence of intermittent connectivity on MEC. Our model sheds important insights into the impact of intermittent connectivity on task processing in MEC, which we believe should be taken into account when designing future MEC systems.

Byzantine Fault Tolerant (BFT) state machine replication protocols are used to achieve agreement among replicated servers with arbitrary faults. Most existing BFT protocols perform well in fault-free cases, but usually suffer from serious performance degradation when faults occur. In this paper, we present DBFT, a BFT protocol that realizes graceful performance degradation in faulty cases. The major novelty of DBFT lies in the double-response mechanism, which lets replica nodes deterministically respond to clients twice: one is after the speculative execution phase and the other is after the commitment phase. The double-response mechanism ensures good performance in spite of inconsistency in speculative execution. Also, to further alleviates undetectable performance attacks by a smartly malicious primary, we change primary upon every outstanding request. Moreover, DBFT does not involve clients in critical consensus operations so as to reduce the load of clients. We prove the correctness properties, i.e., safety and liveness, of DBFT. We conduct extensive experiments and the results show that, DBFT outperforms similar BFT protocols obviously in normal cases.

Recent advancements in edge computing and edge caching provide a feasible solution to support a plethora of new applications such as on-demand videos, AR/VR, road surveillance. However, to apply edge caching in vehicular scenarios is still difficult due to the unkonwn request pattern of vehicular users and intermittent service links between vehicles and edge servers (e.g., Road Side Units, RSUs). In this paper, we aim to investigate the vehicular edge caching problem in practical vehicular scenarios by considering higher hit ratio, while avoiding interruption of caching services. Specifically, to obtain a higher hit ratio, we firstly propose an on-demand adaptive cache algorithm. The algorithm can adjust the eviction time of cached contents by tracking the dynamics of requests and content popularity. We then develop an analysis framework to model the interruption performance of caching services from RSUs. Through diffraction approximation theory, the service process can be modeled as a joint process of the movement and stopping of vehicles to deduce the interruption ratio. To apply the on-demand adaptive cache algorithm in practical scenarios, the final caching decisions should be corrected by incorporating the interruption performance. Therefore, a$\alpha $-fair utility-oriented vehicular edge caching scheme is developed, which can achieve the tradeoff of hit ratio and interruption ratio. Performance evaluation shows the advantages of our proposed vehicular caching scheme in hit ratio, accuracy of analysis model, utility, respectively.

Split Learning (SL) is a distributed machine learning setting that allows several nodes to train neural networks based on model parallelism. Since SL avoids sharing raw data among training nodes, it can protect data privacy by nature. However, recent studies show that, raw data may be reconstructed from activations in training, which may cause data privacy leakage. Besides raw data, label sharing in SL may also cause privacy problems. In order to address these issues, we propose a novel mechanism called multiple activations and labels mix (MALM). By taking advantage of the diversity of sample categories, MALM generates mixed activations that preserve a low distance correlation with the raw data so as to reduce the risk of reconstruction attacks. To protect label information, MALM creates obfuscated labels associated with the raw data so as to prevent adversaries from inferring ground-truth labels. Since clients with few sample categories may not effectively generate mixed activations and obfuscated labels, we propose a bipartite graph based assistant client match technique for MALM, which lets clients with a large number of categories provide mixed activations and obfuscated labels for clients with few categories. Those clients with few categories can mix the obtained mixed activations and obfuscated labels with their own activations and labels. Experimental results show that, compared with baselines, MALM can reduce the risk of raw data and label information leakage with lower cost, while achieving comparable even better model performance.

The technology of colocation, which can effectively improve resource utilization by colocating online services and offline tasks in the same data center, is being adopted by more and more large scale cloud data centers. Predication of resource demands of online services is at the core of colocation since the offline task scheduler relies on the information of future idle resources. Due to the complex running environment, such prediction is far from trivial. XGBoost should be the most popular model for resource demand prediction in recent years. However, our testing and observation show that XGBoost cannot predict values out of the range of training data. We call such problem “range drift”. By addressing this problem, we propose a data center scale online service resource demand forecasting method, named Enhanced-XGB, which consists of three major parts: XGBoost, Coefficient CNN, and Offset CNN. The weight of each tree in XGBoost is learnt dynamically based on the features extracted by the Coefficient CNN, and the offset value is data-dependent functions learnt through the Offset CNN. Finally, the prediction result is obtained by the weighted sum of trees plus the offset value. With the help of CNNs integrated with XGBoost, our model can handle resource demand series with range drift, which makes it more powerful and generic than the original XGBoost. The Enhanced-XGB model has been deployed in a production system at Tencent Games. The experiment results show that our model is superior to XGBoost and LSTM based models in terms of prediction accuracy.

Federated learning is a new and increasingly popular distributed machine learning paradigm, which can employ multiple clients such as mobile phones to collaboratively train a deep neural network model under the coordination of the central server. The raw data is always kept locally on the clients, and only the model parameters are communicated between clients and the server, so that data privacy can be largely preserved. To cope with the effect of not independent and identically distributed (Non-IID) data, personalized federated learning that can provide personalized and customized models for clients participating in federated training has been proposed and widely studied. In this paper, we propose a personalized federated learning algorithm that can provide a personalized local model for each client while storing the latest global model on the central server through only one federated training. The deep neural network model is divided into two parts of global layers and personalized layers, and the whole training process is partitioned into earlier and later stages. Based on the division of layer-wise parameters, the cumulative learning strategy of updating parameters on different layers with variable frequency is adopted to better personalize local models in the later stage based on the global features learned in the earlier stage. Compared with existing personalized federated learning algorithms, our proposed algorithm can achieve a good balance between the personalized model and the global model in federated learning, and performs better in communication efficiency and model accuracy.

Energy harvesting rates of sensors in renewable (e.g., solar energy) wireless sensor networks are not only lower than their energy consumption rates but also temporally varying. Existing studies exploited spatial data correlations among sensors to reduce their energy consumptions, where the data correlations mean that the sensing data of nearby sensors have high similarities. They assumed that the sensing data of nearby sensors are very likely to highly correlated. They adopted a coarse-grained spatial-correlation model, in which sensors are partitioned into different clusters such that the sensors in the same cluster have high data similarities with each other. Then, only the sensor with the maximum residual energy in each cluster sends its sensing data, while the other sensors do not. We, however, notice that the data similarities among nearby sensors in real sensor networks may vary significantly, i.e., ranging from very similar to not similar at all. Since the existing algorithms require that the sensors in the same cluster have high data similarities with each other, the sensors in a network may be partitioned into many clusters and each cluster consists of only a few sensors, where two nearby sensors belong to two different clusters if the sensing data of the two sensors are not highly correlated. Therefore, in the existing studies, many sensors have to send all their data as there are many clusters. Unlike the existing studies, in this article, we first propose a fine-grained spatial correlation model, in which sensors are partitioned into only a few clusters and each cluster consists of many sensors. Then, each cluster master sensor sends all its data to the sink, while the majority of other sensors in the cluster transmit only their nonredundant data, thereby significantly saving sensor energy consumptions. We formulate a novel sensor clustering problem under the proposed model, which is to partition sensors into different clusters and choose a representative sensor for each cluster such that the amount of suppressed redundant data transmissions is maximized. We propose a randomized (0.5-ε)-approximation algorithm for the clustering problem, where E is a given constant with 0 <; ε ≤ 0.5. To further reduce sensor energy consumption, we consider temporal data correlations, where the sensing data by a sensor in a short period are likely to be highly correlated. We investigate a data utility maximization problem that allocates sensor data rates and routing so that the accumulative utility of both spatially and temporally correlated data received by the sink is maximized. We devise a near-optimal algorithm for the problem. We finally evaluate the performance of the proposed algorithms through experiments. the experimental results show that the proposed algorithms are very promising.

Virtual machine (VM) management is at the core of virtualized cloud data centers. Among others, how to reduce the startup delay of VMs is a key issue for improving user experience and resource utility. In this paper, we study this issue by jointly considering VM placement and VM image caching. We formulate the joint placement problem and design several joint algorithms, including both online and offline algorithms, to speed up VM startup. In our design, we adopt the cooperative caching approach, where image cache copies are shared among physical machines (PMs) so as to reduce image retrieval time. The key point of our algorithms lies in how to appropriately place VM image cache among PMs so as to speed up VM startup as much as possible. The proposed algorithms are evaluated by extensive simulations via SimGrid. The results show that our algorithms can achieve shorter startup delay in most cases, compared with existing ones.

To mitigate the scalability problem of decentralized cryptocurrencies such as Bitcoin and Ethereum, the payment channel, which allows two parties to perform secure coin transfers without involving the blockchain, has been proposed. The payment channel increases the transaction throughput of two parties to a level that is only limited by their network bandwidth. Recent proposals focus on extending the two-party payment channel to the N-party payment hub. Unfortunately, none of them can achieve efficiency, flexibility in the absence of a trusted third-party. In this paper, we propose Garou, a secure N-party payment hub that allows multiple parties to perform secure off-chain coin transfers. Except in the case of disputes, participants within the payment hub can make concurrent and direct coin transfers with each other without the involvement of the blockchain or any third-party intermediaries. This allows Garou to achieve both high-performance and flexibility. If all participants keep online, Garou guarantees all honest users’ balance security against strong adversarial capabilities. To demonstrate the feasibility of the Garou protocol, we develop a proof of concept prototype for the Ethereum network. Experimental evaluations show that, with 300 participants and without disputes, that the maximum transaction throughput of Garou is$20{ \times }$higher than that of state-of-the-art payment hubs.

Mobile edge computation partitioning is an effective technique to improve the applications performance on the mobile devices by selectively offloading some computations from the devices to the nearby edge cloud. Most previous works focus on the computation partitioning for a single user. Recent works begin to study the computation partitioning in a multiple user environment in which a number of users compete for the constrained computation resources on the edge cloud. However, these works neglect the fact that the users normally also share the network resources to access the edge cloud, and thus the allocation of bandwidth to the users significantly affects the overall performance of the users. In this paper, we studynetwork aware mobile edge computation partitioning in multi-user environments, i.e., to decide for each user which parts of the application should be offloaded onto the edge cloud, and which others should be executed locally, and meanwhile to allocate the access bandwidth among the users, such that the average application performance of the users is maximized. This problem is new in that we consider the competition among users for the network bandwidth as well as the computation resources in a multi-user environment. With a set of novel models and formulations, we transform the problem into the classic Multi-class Multi-dimensional Knapsack Problem, and develop an effective algorithm, namely Performance Function Matrix based Heuristic (PFM-H), to solve it. We further consider the user mobility and design effective online algorithms that could be easily deployed in practical systems. Comprehensive trace driven simulations show that our proposed algorithm outperforms the benchmark algorithms significantly in the average application performance.

With the fast development of Internet of Vehicles (IoV), various types of computation-intensive vehicular applications pose significant challenges to resource-constrained vehicles. The emerging Vehicular Edge Computing (VEC) and Edge Intelligence (EI) can alleviate this situation by offloading the computation tasks of vehicles to the roadside edge servers. However, with many vehicles contending for the communication and computation resources at the same time, how to quickly and efficiently make an optimal computation offloading decision for individual vehicles represents a fundamental research issue. In this paper, we propose a self-learning based distributed computation offloading scheme for IoV. Note that without any centralized controller, a fully distributed algorithm is necessary. The proposed scheme is devised based on a game-theoretic model. Specifically, through establishing an offloading framework with communication and computation for IoV, the computation offloading problem is first formulated as a distributed offloading decision-making game, in which each vehicle as a player makes its best response decision to minimize its joint cost (including latency and offloading cost). The existence of Nash Equilibrium can be proved. We then propose a self-learning based distributed computation offloading (DISCO) algorithm to reach the Nash Equilibrium, where a mutually satisfactory solution among vehicles is obtained and no vehicle is willing to change its decision. Using extensive simulations, we verify that DISCO can outperform the counterparts and achieve at least an order-of-magnitude improvement on time overhead and 88% performance gain on message overhead, only at up to 12% performance loss on joint cost over the centralized scheme.

Distributed deep learning has been proposed to train large scale neural network models with huge amounts of datasets using multiple workers. Since workers need to frequently communicate with each other to exchange gradients for updating parameters, communication overhead is always a major challenge in distributed deep learning. To cope with this challenge, gradient compression has been used to reduce the amount of data to be exchanged. However, existing compression methods, including both gradient quantization and gradient sparsification, either hurt model performance significantly or suffer from inefficient compression. In this paper, we propose a novel approach, called Standard Deviation based Adaptive Gradient Compression (SDAGC), which can simultaneously achieve model training with low communication overhead and high model performance in synchronous training. SDAGC uses the standard deviation of gradients in each layer of the neural network to dynamically calculate a suitable threshold according to the training process. Moreover, several associated methods, including residual gradient accumulation, local gradient clipping, adaptive learning rate revision and momentum compensation, are integrated to guarantee the convergence of the model. We verify the performance of SDAGC on various machine learning tasks: image classification, language modeling and speech recognition. The experiment results show that, compared with other existing works, SDAGC can achieve a gradient compression ratio from 433× to 2021× with similar or even better accuracy.

Payment channel is an effective and popular technique to improve the scalability and throughput of blockchains by transferring transactions from on-chain to off-chain. Multiple payment channels can constitute a payment network and realize transaction execution via multi-hop paths. How to find a feasible and efficient transaction path, i.e., transaction routing, is a key issue in payment channel networks, and different solutions have been proposed. However, the problem of funds skewness, which may cause routing failures, has been largely ignored in existing routing algorithms. In this work, we design FSTR, a routing algorithm that attempts to route transactions using a funds skewness based path selection scheme so as to reduce funds skewness and increase transaction success probability. To evaluate the performance of FSTR, we conduct experiments using the real-world dataset of Ripple. The experiment results show that FSTR outperforms existing routing algorithms, in terms of success ratio, delay, and overhead.

Job scheduling in cluster is often considered as a difficult online decision-making problem, and its solution depends largely on the understanding of the workload and environment. People usually first propose a simple heuristic scheduling algorithm, and then perform repeated and tedious manual tests and adjustments based on the characteristics of the workload to gradually improve the algorithm. In this work, focusing on multi-cluster environments, load balancing and efficient scheduling, we present RLSK, a deep reinforcement learning based job scheduler for scheduling independent batch jobs among multiple federated cloud computing clusters adaptively. By directly specifying high-level scheduling targets, RLSK interacts with the system environment and automatically learns scheduling strategies from experience without any prior knowledge assumed over the underlying multi-cluster environment and human instructions, which avoids people's tedious testing and tuning work. We implement our scheduler based on Kubernetes, and conduct simulations to evaluate the performance of our design. The results show that, RLSK can outperform traditional scheduling algorithms.

Distributed parallel training using computing clusters is desirable for large scale deep neural networks. One of the key challenges in distributed training is the communication cost for exchanging information, such as stochastic gradients, among training nodes. Recently, gradient sparsification techniques have been proposed to reduce the amount of data exchanged and thus alleviate the network overhead. However, most existing gradient sparsification approaches consider only synchronous parallelism and cannot be applied in asynchronous distributed training.

Data shuffling can improve the statistical performance of distributed machine learning. However, the obstruction of applying data shuffling is the high communication cost. Existing works use coding technology to reduce communication cost. These works assume a master-worker based storage architecture. However, due to the demand for unlimited storage on the master, the master-worker storage architecture is not always practical in common data centers. In this paper, we propose a new coding method for data shuffling in the decentralized storage architecture, which is built on a fat-tree based data center network. The method determines which data samples should be encoded together and from which the encoded package should be sent to minimize the communication cost. We develop a real-world test-bed to evaluate our method. The results show that our method can reduce the transmission time by 6.4% over the state-of-art coding method, and by 27.8% over Unicasting.

Automatic driving services have large volume, location-aware, and time-changing contents, which are suitable to be cached by the edge. However, the traffic on the edge will be extremely high especially in the area with high vehicle density, if the vehicles directly access the contents from the edge as they demand. A hybrid data dissemination model with both vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) disseminations has been proposed to reduce the traffic on the edge, in which the edge (infrastructure) selectively injects data to the vehicles and leverages the vehicle network to disseminate the data. In this article, we study the hybrid data dissemination problem, i.e., to optimally determine when and which vehicle the data are injected into, and whether the vehicle acquires the demanded data directly from the edge or from nearby neighbors, with the aim of minimizing the traffic cost on the edge and meeting the deadlines of acquiring the data. Existing approach prioritizes the selection of V2I disseminations at first and then explores the V2V disseminations which have no conflict with the V2I disseminations. This approach cannot fully take advantage of V2V to reduce the traffic cost on the edge. We propose a new data dissemination algorithm, named the offline algorithm for hybrid data dissemination (OFDD), which seeks the most beneficial V2V broadcasts with priority, and then choose feasible V2I disseminations. Based on OFDD, we develop both the snapshot and prediction-based online algorithms. We follow with extensive simulations to validate the proposed algorithms. The results show that our algorithms significantly outperform the state-of-the-art approaches in terms of data acquisition rate and traffic cost.

Fog computing is envisioned as a promising approach for supporting emerging computation-intensive applications on capacity and battery constrained mobile Internet of Things (IoT) devices. Technically speaking, a massive crowd of devices in close proximity can be harvested and collaborate for computation and communication resource sharing. Hence fog computing enables significant potentials in low-latency and energy-efficient mobile task execution. However, without an efficient incentive mechanism to stimulate resource sharing among devices, the benefits of fog computing cannot be fully realized. Leveraging coalitional game theory, this work presents an efficient incentive mechanism to incentivize mutually-beneficial resource cooperation among the devices for collaborative task execution. In particular, to efficiently achieve mutually beneficial task execution, the proposed mechanism groups the devices into multiple micro computing clusters (MCCs). Within each MCC, devices can exchange mutually beneficial actions by helping to compute or transmit tasks, making all of their performances no worse than local execution or execution in the fog server. The solution to the MCC formation is devised by both centralized and decentralized schemes and further proven to admit nice properties such as top coalition, core solution, individual rationality and computational efficiency. Extensive numerical studies demonstrate the superior performance of our MCC formation mechanisms.

State Machine Replication (SMR) is a fundamental fault tolerant technique for distributed systems. SMR traditionally requires sequential execution of commands at each replica node, so as to guarantee strong consistency among replicas. To achieve high performance at large scale cloud datacenters, SMR has been parallelized by employing multiple threads at each replica. In this paper, we propose SMR-X, a novel parallel SMR scheme, which realizes flexible mapping of commands for parallel executing at each replica. The mapping between clients' requests and work threads is dynamically adjusted according to the load level of work threads. Therefore, workloads of different threads can be well balanced and high system throughput can be achieved. The major challenge in our work lies in the inconsistency problem caused by dynamic changes in request-thread mapping. To cope with this, we design delicate mechanisms to synchronize mapping function, so that strong consistency among replicas can be guaranteed. The correctness of the proposed scheme is rigorously proved and its performance is evaluated via simulations. Simulation results show that SMR-X can achieve better load balance and lower access latency than existing parallel SMR schemes.

Byzantine Fault-Tolerant (BFT) state machine replication protocol is an important building block for highly available distributed computing. This paper presents COMBFT, a BFT protocol that achieves both efficiency and robustness simultaneously. The major novelty of COMBFT lies in Conflicting-Order-Match (COM), a new request ordering mechanism that uses a new way to select the available sequence number for requests, and detects the possible malicious primary early. COM assigns sequence number based on request interference, and requires both primary and backup nodes to conduct request ordering, which can greatly reduce the impact of malicious primary and clients. When the backup suspects the primary may be malicious, it triggers an efficient commit protocol with two phases (i.e., suspect phase and commit phase) to further confirm whether the primary is malicious, and commit the request. The performance of COMBFT is evaluated via simulations and the results illustrate the outstanding performance of COMBFT in terms of throughput, latency and fault scalability.

Aspect-based sentiment analysis is a fine-grained task of sentiment classification for multiple aspects in a sentence. While the overall polarity of a sentence has been successfully predicted with neural network architectures, aspect-specific sentiment analysis still faces challenges. In this paper, we propose a novel neural network named IRP-BERT. It is based on BERT which is the newest breakthrough in natural language processing. IRP-BERT aims at improving the accuracy of predicting the right polarity of aspects in a sentence, especially the neutral ones, because some sentimental-related words such as "love" in a noun phrase which is neutrally sentimental could have a wrong impact on the polarity classification to aspects. More specifically, we design a module named Irrelevance-Removing Process (IRP) to counteract the impact from irrelevant sentimental-related features. The final prediction depends on word representations from the outputs of both IRP and BERT. Experiments on a twitter dataset show that our model outperforms both existing state-of-the-art methods and fine-tuned BERT baselines.

Edge computing is envisioned to be the de-facto paradigm of hosting emerging low latency Internet-of-Things (IoT) data streaming services.For IoT data streaming in edge computing, cost management is of strategic significance, due to the low cost-efficiency of edge servers. While existing literature adopts a reactive approach to dynamically provisioning edge servers to reduce cost, the delay of server activation and instantiation has been mostly ignored. In this paper, we target a proactive approach to dynamic edge server provisioning for real-time IoT data streaming across edge nodes, which adjusts server provisioning ahead of time, based on prediction of the upcoming workload. To effectively predict upcoming workload, a learning-based method online gradient descent is applied. We further combine the online learning method with an online optimization algorithm for server provisioning in a joint online optimization framework, through (1) minimizing of the regret incurred by inaccurate workload prediction, and (2) minimizing the cost incurred by near-optimal online decisions. The resulting predictive online algorithm can well leverage the power of prediction and achieve a good performance guarantee, as verified by both rigorous theoretical analysis and extensive trace-driven evaluations.

The surging interest in blockchain has revitalized the search for efficient Byzantine fault-tolerant (BFT) protocols, which are used for blockchains to achieve consensus among replicated data blocks. Most existing BFT protocols perform well in fault-free cases, but they usually suffer from serious performance degradation when faults occur. In this paper, we present DBFT, a BFT protocol that realizes graceful performance degradation in normal cases. The major novelty of DBFT lies in the double-response mechanism, which lets replica nodes deterministically respond to clients twice: one is after the speculative execution phase and the other is after the commitment phase. The double-response mechanism can handle inconsistency in speculative execution, so as to alleviate performance degradation caused by faults. Moreover, DBFT does not involve clients in critical consensus operations so as to reduce the load of clients. The correctness properties, i.e., safety and liveness, of DBFT is rigorously proved. The performance of DBFT is evaluated via experiments and the results show that, DBFT outperforms similar BFT protocols obviously in normal cases.

State machine replication (SMR) is a fundamental fault tolerant technique for distributed systems to guarantee consistency among replicas via sequential execution of commands. With the development of cloud computing, parallel SMR has been recently proposed for large scale cloud datacenters. In this paper, we propose PDFE, a novel parallel SMR scheme, which realizes flexible dispatch of parallel ordered commands for parallel executing. In PDFE, the mapping/binding between ordering threads and work threads becomes dynamic, and commands can be dispatched according to the work load level of different threads. Such flexibility can help achieve two levels of load balancing: load balancing between ordering threads and work threads, and load balancing among work threads. The major challenge in our work lies in the inconsistency problem caused by dynamic changes in command dis-patch, and it is addressed by a specially designed mechanism. Compared with existing parallel SMR schemes, PDFE can achieve better load balancing and higher system efficiency. Such advantages are validated by experimental performance evaluation.

Named Data Networking (NDN) is a promising technology for content centric networks, and it is suitable for vehicular networks since no IP architecture is required. Quite a number of solutions have been proposed for vehicular NDN (V- NDN), but high communication cost due to frequent topology changes caused by high mobility of vehicles is still a challenge to be addressed. In this paper, we study how to disseminate traffic information to vehicles via V-NDN. Different from existing works, we consider navigation route based data interests, i.e., a vehicle is concerned about the traffic information along road segments planned to take. According to such a data interest scenario, we propose a tree based data interest structure and associated maintenance operations to merge identical data interests due to overlapping navigation routes among different vehicles. With the tree based data interest management, the number of interest packets can be significantly reduced. Then, we propose trigger based mechanisms for data interest packet re-sending and forwarding, which can avoid unnecessary interest packets re-sending. With our design, traffic information can be disseminated to interested nodes with high success ratio and low communication cost simultaneously. Simulations via SUMO and ndnSIM confirm such advantages of our work.

Fog computing is envisioned as a promising approach for supporting emerging mission-critical applications on capacity and battery constrained mobile devices. By harvesting and collaborating a massive crowd of devices in close proximity for computation and communication resource sharing, it enables significant potentials in low-latency and energy-efficient mobile task execution. It is readily acknowledged, however, that without an efficient incentive mechanism that stimulates resources sharing among devices, the benefits of fog computing cannot be fully realized. Leveraging coalitional game theory, this work presents an efficient incentive mechanism to incentivize mutually-beneficial resource cooperation among the devices for collaborative task execution. Specially, to prevent the over-exploiting and free-riding behaviors that harm resource-rich device's willingness to collaborate, the proposed mechanism groups the devices into multiple micro computing clusters (MC-C). Within each MCC, devices can exchange mutually beneficial actions by helping to compute or transmit tasks, making all of them better off. The solution of the MCC formation is devised by a network-assisted mechanism, which is further proven to admit nice properties such as top coalition and core solution.

Mobile edge cloud has been increasingly concerned by researchers due to its closer distance to mobile users than the traditional cloud on Internet. Offloading computations from mobile devices to the nearby edge cloud is an effective technique to accelerate the applications and/or save energy on the mobile devices. However, the mobile edge cloud usually has limited computation resources and constrained access bandwidth shared by multiple users in its proximity. Thus, allocation of resources and bandwidth among the users is significant to the overall application performance. In this paper, we study network aware multi-user computation partitioning problem in mobile edge clouds, i.e., to decide for each user which parts of the application should be offload onto the edge cloud, and which others should be executed locally, and meanwhile to allocate the access bandwidth among the users, such that the average application performance of the users is maximized. This problem is novel in that we consider the competition among users for both computing resources and bandwidth, and jointly optimizes the partitioning decisions with the allocation of resources and bandwidths among users, while most existing works either focus on the single user computation partitioning or study the multiple user computation partitioning without regard of the constrained network bandwidth. We first formulate the problem, and then transform it into the classic Multi-class Multi-dimensional Knapsack Problem and develop an effective algorithm, namely Performance Function Matrix based Heuristic (PFM-H), to solve it. Comprehensive simulations show that our proposed algorithm outperforms the benchmark algorithms significantly in the average application performance.

A smart computing system is usually composed of smart mobile nodes, such as mobile phones and/or sensor nodes. In this paper, we study the consensus problem in a smart computing system, which has many applications in smart computing, such as smart grid. Due to mobility and failures, smart computing nodes may join and leave a system from time to time. Such dynamicity in node set of a smart computing system brings new challenges in consensus protocol design. A process newly joining the system may miss messages sent out before the joining, which will cause existing consensus algorithms blocked forever and cannot terminate. Process leave may cause the value of previous decision lost, which will make existing consensus algorithms decide on multiple values and violate the agreement property. To address such challenges, we firstly define necessary dynamicity model to constrain the change of process set, and then design a dynamic consensus algorithm which can handle process join and leave effectively via two novel mechanisms. The first mechanism lets old processes send redeeming messages to new processes, while the second one lets processes jump to higher rounds under some specific scenarios. The correctness of our proposed algorithm is rigorously proved, in terms of termination, agreement and validity.

Vehicular network has been recently used to achieve high efficient and flexible traffic scheduling at intersection roads for smart transportation systems. Different from existing works, where traffic signal is used to schedule waiting vehicles at each lane, we propose to divide vehicles in the same lane into small groups and schedule vehicle groups via wireless communication rather than traffic lights. Such direct scheduling of vehicles can reduce waiting time and improve fairness, especially when the traffic volume in different lanes is imbalanced. The key challenge in such a design lies in determining appropriate size of groups with respect to real-time traffic conditions. To cope with this issue, we propose a neuro-fuzzy network-based grouping mechanism, where the network is trained using reinforcement learning technique. Also, vehicle groups are scheduled via a neuro-fuzzy network. Simulations using ns3 are conducted to evaluate the performance of our algorithm and compare it with similar works. The results show that our algorithm can reduce waiting time and at the same time improve fairness in various cases, and the advantage against traffic light algorithms can be up to 40%.

In an IMANET, mobile users access both text and media web contents on the Internet through gateway nodes, with web access patterns, i.e., the Zipf-like distribution or the Stretched Exponential distribution. To reduce data access delay from the Internet, we consider the cache placement problem in cooperative caching, which is that each mobile node selects a subset of web contents to cache cooperatively in its limited cache so that total access cost is minimized. It has been proved NP-hard. We propose a solution named Adaptive Allocation Cooperative Caching (AACC), which adaptively divides the cache space of each node into two components: altruistic and selfish, according to detected data access patterns. AACC aims to find the optimal cache allocation solution to allocate appropriate cache spaces for two components in order to minimize total access cost. Given the Zipf-like access distribution, we find a near-optimal allocation solution to the cache placement problem. Simulation results show that AACC achieves much better performance than the existing best cooperative caching strategy in IMANETs in terms of average query delay, caching overheads, and query success ratio. In particular, AACC reduces caching overheads by 40% in average.

This paper comes up with a SDN based On-Demand Routing Protocol, SVAO, which separates data forwarding layer and network control layer, as in SDN, to enhance the data transmission efficiency within VANETs. The Roadside Service Unit plays the role of Local Controller and is in charge of selecting vehicles to forward packet within a road segment. All the vehicles state in the road. Correspondingly, a two-level design is used. Global Level is distributed and adopts a ranked query scheme to collect vehicle information and determine the road segments along which a message should be forwarded. And the Local Level is in charge of selecting forwarding vehicles in each road segment determined by the global level. We compare SVAO with popular ad-hoc network routing protocols, including OLSR, DSR, DSDV, and DB via simulations. We consider the impact of vehicle density, speed on data transmission rate and average packet delay. The simulation results show that in case of large network scales or high vehicle speed, SVAO performs better than the others.

With the development of the technology, Vehicular Ad-hoc Network is developing rapidly and continuously. And many new algorithms in VANET were put forward. For example, many people apply the Named Data Networks or Software Defined Networking6 to the VANET. However, few researches are looking into the analysis of the topology of the VANET, among most of which are based on simple situations and calculation of the track of vehicles. And therefore, these analysis are inadequate for analyzing the topology under complex situations. This paper is going to put forward a new method for analyzing the change of the topology of VANET, which is capable to quantitatively analyze the change under different and complex situations. In the real experiment, when compared with traditional analysis methods, this new method can analyze more precisely the impact of some possible wireless communication problem on the change of the topology, for example, the impact of hidden node on the vehicular wireless communication and whether the number or the density of vehicles makes the topology of VANET more stable. And this new method can provide a numerical result instead of purely qualitative analysis.

Large Web applications usually require replicating data across geo-distributed datacenters to achieve high locality, durability and availability. However, maintaining strong consistency in geo-replicated systems usually suffers from long latency due to costly coordination across datacenters. Among others, transaction chopping is an effective and efficient approach to cope with such a challenge. In this paper, we propose DTC (Dynamic Transaction Chopping), a novel technique that chops transactions and checks their conflicts in a dynamic and automatic way, during application execution. DTC mainly consists of two parts: a dynamic chopper that chops transaction dynamically according to data partition scheme, and a conflict detection algorithm for determining the safety of the dynamic chopping. Compared with existing transaction chopping technique for geo-replicated systems, DTC has several advantages, including transparency to programmers, flexibility in conflict analysis, high degree of piecewise execution, and adaptability to dynamic partition schemes. We implement our DTC technique and conduct experiments to examine the correctness of DTC and evaluate its performance. The experiment results show that our DTC technique can achieve much more piecewise execution than the existing chopping approach does, and reduce execution time obviously.

Traffic control at intersections is a key issue and hot research topic in intelligent transportation systems. Existing approaches, including traffic light scheduling and trajectory maneuver, are either inaccurate and inflexible or complicated and costly. More importantly, due to the dynamics of traffic, it is really difficult to obtain the optimal solution in a real-time way. Inspired by the emergence of vehicular ad hoc network, we propose a novel approach to traffic control at intersections. Via vehicle to vehicle or vehicle to infrastructure communications, vehicles can compete for the privilege of passing the intersection, i.e., traffic is controlled via coordination among vehicles. Such an approach is flexible and efficient. To realize the coordination among vehicles, we first model the problem as a new variant of the classic mutual exclusion problem, and then design algorithms to solve new problem. Both centralized and distributed algorithms are. We conduct extensive simulations to evaluate the performance of our proposed algorithms. The results show that, our approach is efficient and outperforms a reference algorithm based on optimal traffic light scheduling. Moreover, our approach does not rely on traffic light or intersection controller facilities, which makes it flexible and applicable to various kinds of intersections.

This paper presents a simple broadcast algorithm suited to dynamic systems where links can repeatedly appear and disappear. The algorithm is proved correct and a simple improvement is introduced, that reduces the number and the size of control messages. As it extends in a simple way a classical network traversal algorithm to the dynamic context, the proposed algorithm has also pedagogical flavor.

Dynamic network is the abstraction of networks with frequent topology changes arising from node mobility or other reasons. With the model of dynamic network, fundamental distributed computing problems can be formally studied with rigorous correctness. In this work, we study the counting problem, which plays a significant role in many applications, such as all-to-all information dissemination. Existing counting algorithms for dynamic networks are mostly estimation based and randomized, and cannot achieve accurate results deterministically. To address this problem, we adopt the diffusing computation approach in this paper. Although this approach has been widely used to solve different distributed computing problems, due to dynamic topology changes, existing diffusing computation algorithms cannot work under dynamic network models. To address such challenge, we firstly define a new dynamicity model, named (Q, S)-distance, which is used to describe dynamic changes of information propagation time. Based on (Q, S)-distance, we design two different counting algorithms. The first algorithm focuses on extending typical diffusing computation approach by handling topology changes. Different from existing algorithms, which have two phases, our algorithms contain three phases and the new phase is used to determine and notify the termination of tree growing, which the major challenge is caused by topology dynamicity. The second algorithm further extends the first one by incorporating a cluster-based hierarchy to reduce communication cost. The correctness of both the algorithms is formally proved and their performances in time cost and communication cost are analyzed.

Cooperative caching is a desirable approach to achieve efficient data access in multi-hop wireless networks. Existing cooperative caching algorithms mostly focus on cache placement. Another key issue, cache consistency, has not been adequately addressed. In this paper, we propose CACC, a cooperative approach to maintain cache consistency for wireless mesh networks. CACC combines push and pull by making use of the hierarchical architecture of mesh networks. The key contribution of CACC lies in two techniques that introduce cooperation among network nodes in delivering invalidation reports (IRs) so as to reduce communication cost and tolerate message losses. The first technique, IR integration, buffers and merges IRs at gateway nodes and periodically broadcasts them. The second technique, cooperative IR re-sending, lets intermediate nodes resend missed IR messages upon request. The interval of IR broadcast is optimized to achieve the optimal tradeoff between push and pull. We conduct numerical analysis to find optimal values for different scenarios. We also perform simulation to confirm our analysis results and compare with existing approaches. The results show that CACC can save message cost significantly (50-70 percent).

Mutual exclusion (ME) is a fundamental problem for resource allocation in distributed systems, It is concerned with how the various processes access shared resources in a mutually exclusive way. Besides the classic ME problem, several variant problems have been proposed and studied. In this paper, drawing inspiration from the scenario of controlling autonomous vehicles at intersections, we have defined a new ME problem, called Local Group Mutual Exclusion (LGME), w here mutual exclusion is necessary only among the processes requesting overlap but not the same set of resources. In comparison with other variant problems of ME, LGME is more general but also more challenging. To solve the LGME problem, we propose a novel notion called strong coterie, which can handle the complex process relationship in LGME. Based on strong coterie, we have designed an ME algorithm, which can handle concurrent CS execution and message asynchrony. The correctness of our algorithm is rigorously proved.

Dynamic network is the abstraction of networks with frequent topology changes arising from node mobility or other reasons. With the model of dynamic network, distributed computation problems can be formally studied with rigorous correctness. Information dissemination is one of such problems and it has received much attention recently. However, existing works focus on the time cost of dissemination, i.e. how fast the information can be disseminated to all nodes, and communication cost has been largely ignored. Our work focuses on high communication efficiency in information dissemination with correctness property guaranteed. We achieve this by making use of cluster-based hierarchy. Clustering has been widely studied and used in wireless networks to reduce communication cost. However, to the best of our knowledge, it has never been considered in the study of dynamic networks. In this paper, we firstly propose a dynamic network model, named (T, L)-HiNet, to extend existing dynamic network model with clusters. (T, L)-HiNet includes several properties defining the dynamics of cluster hierarchy in a dynamic network. Base on (T, L)-HiNet, we design hierarchical information dissemination algorithms for different scenarios of dynamics. The correctness of our algorithms is proved and the performance is analyzed. Compared with the algorithm recently proposed by Kuhn, Lynch and Oshman [7], our design can significantly reduce communication and our objective is fully achieved.

Cooperative caching is an effective approach to share cache copies among network nodes and then reduce data access cost in ad hoc networks. Since cached data may be privacy sensitive, privacy preservation should be considered. In this paper, we study this problem by achieving anonymity in cooperative caching. Our anonymous caching protocol borrows idea from anonymous routing protocols. Data query and access are realized based on anonymous hop-by-hop broadcasting. Pseudonyms are used in communications so as to protect the real identity. Security analysis confirms that our protocol can achieve sender anonymity, receiver anonymity, and communication anonymity. We also conduct simulations to evaluate the effect of anonymity mechanism on system performance. Different variants of anonymous caching algorithm and non-anonymous algorithms are simulated for comparisons. The results show that, anonymity will cause additional message cost and time cost, but the cost is reasonable. To the best of our knowledge, this is the first work on anonymity in cooperative caching.

Wireless ad hoc network is a promising networking technology to provide users with Internet access anywhere anytime. To cope with resource constraints of wireless ad hoc networks, data caching is widely used to efficiently reduce data access cost. In this paper, we propose an efficient data caching algorithm which makes use of the overhearing property of wireless communication to improve caching performance. Due to the broadcast nature of wireless links, a packet can be overheard by a node within the transmission range of the transmitter, even if the node is not the intended target. Our proposed algorithm explores the overheard information, including data request and data reply, to optimize cache placement and cache discovery. To the best of our knowledge, this is the first work that considers the overhearing property of wireless communications in data caching. The simulation results show that, compared with one representative algorithm and a naive overhearing algorithm, our proposed algorithm can significantly reduce both message cost and access delay.

Cooperative caching is especially desirable for wireless ad hoc networks to achieve efficient data access. Although studies have been conducted on cooperative cache placement and cache consistency, the cache discovery problem has been largely ignored. In this paper, we propose a Voronoi Diagram (VD) based cooperative cache discovery approach to reduce data access cost by limiting the cache information update and query within a single Voronoi Region (VR). By addressing several challenging issues caused by the lack of geometrical coordinate in ad hoc networks, we design two topological VD algorithms for different network environments. To our knowledge, this is the first work for the construction of topological VD. Simulation results show that our proposed approach can get cache copies faster while incurring much lower message cost compared with existing ones.

Cooperative caching is especially desirable for multi-hop wireless networks to achieve efficient data access. Existing cooperative caching algorithms for wireless networks mostly focus on cache placement. Another key issue, cache consistency maintenance has not been adequately addressed. In this paper, we propose the first cooperative approach to maintain cache consistency for wireless mesh networks. It basically combines push and pull by making use of the hierarchical architecture of mesh networks. More precisely, we propose two techniques introducing cooperation among network nodes in delivering Invalidation Reports (IR) so as to reduce communication cost and tolerate message losses: IR integration buffers and integrates IRs at the gateway nodes and periodically broadcasts them, Cooperative IR re-sending lets the intermediate nodes resend missed IR messages upon request. The most challenging issue in our design is the determination of the optimal IR broadcast period in order to achieve the optimal tradeoff between push and pull. We conduct numerical analysis to get optimal values for different scenarios. Simulation results confirm our analysis well and comparisons with existing approaches show that our approach can save message cost significantly (50%-70%).

Service unavailability frequently occurs in dynamic environments, like mobile ad hoc networks (MANET), due to service provider failure, network partitioning, or service scope outage by service provider or user mobility. Service handoff is needed to provide users with alternate matching services in case the original service becomes unavailable. However, existing service discovery solutions for MANET do not address this problem. In this paper, we propose a novel service discovery solution which employs service handoff to facilitate seamless service access for mobile users. The major concerns of service handoff are -- reducing handoff frequency, message cost, and delay as well as balancing loads on service providers. Addressing the above issues are challenging in the dynamic MANET environment. We design a reliable and seamless service discovery solution which employs two different service handoff protocols for different situations. Simulation results show that our protocols can support seamless service access for mobile users at low message cost and time delay while achieving high load balance among service providers.

Wireless Mesh Network (WMN) is a promising technology for extending the Internet access to wider areas with high bandwidth. Cooperative caching is especially desirable for multi-hop wireless networks, including WMNs, to achieve efficient data access. Although quite a number of works have been done on data cooperative caching in wireless networks, WMN's characteristic in network architecture necessitates the specially designed caching system. In this paper, we design the first cache cooperative caching system for WMN environments. A hierarchical cache system architecture is proposed, where each mesh access point acts as the coordinator of the client nodes inside the cell. Cache placement is conducted in the cell level and the network level respectively. Accordingly, we design two different placement metrics for different levels of placement decisions. Simulation results show that our algorithm performs much better than similar ones, in terms of both access delay and message cost.

Particle swarm optimization (PSO) is predominately used to find solutions for continuous optimization problems. As the operators of PSO are originally designed in ann-dimensional continuous space, the advancement of using PSO to find solutions in a discrete space is at a slow pace. In this paper, a novel set-based PSO (S-PSO) method for the solutions of some combinatorial optimization problems (COPs) in discrete space is presented. The proposed S-PSO features the following characteristics. First, it is based on using a set-based representation scheme that enables S-PSO to characterize the discrete search space of COPs. Second, the candidate solution and velocity are defined as a crisp set, and a set with possibilities, respectively. All arithmetic operators in the velocity and position updating rules used in the original PSO are replaced by the operators and procedures defined on crisp sets, and sets with possibilities in S-PSO. The S-PSO method can thus follow a similar structure to the original PSO for searching in a discrete space. Based on the proposed S-PSO method, most of the existing PSO variants, such as the global version PSO, the local version PSO with different topologies, and the comprehensive learning PSO (CLPSO), can be extended to their corresponding discrete versions. These discrete PSO versions based on S-PSO are tested on two famous COPs: the traveling salesman problem and the multidimensional knapsack problem. Experimental results show that the discrete version of the CLPSO algorithm based on S-PSO is promising.

Data caching is one of the attractive techniques that can increase the efficiency of data access in wireless multi-hop ad hoc networks. However, it is still a challenging task to achieve an optimal trade-off between the total traffic cost and average access delay by properly selecting cache nodes, i.e. cache placement. In this paper, we address the problem of cache placement in wireless multi-hop ad hoc networks. We consider dynamic networks, in which there is a data source that stores one data item and other client nodes require accessing the data item. We define the cache placement problem on a dynamic network topology as Dynamic Cache Placement (DCP). Most of the existing cache placement algorithms use hop counts to measure the total cost of caching systems. Due to the impact of contentions in wireless networks, one hop delay is different from each other and it varies with the traffic load. Thus the previous algorithms cannot achieve their expected performance. We propose two heuristic cache placement algorithms, named Centralized Contention-Aware Caching Algorithm (CCCA) and Distributed Contention-Aware Caching Algorithm (DCCA), which detect the variation of contentions to evaluate the benefit of selecting a node as cache node. Simulation results show that the proposed algorithms achieve better performance than other alternative ones in terms of average query delay, caching overheads, and query success ratio.

Re-authentication delay is a major concern of fast handoff in wireless mesh networks. Most recent efforts focus on proactive approaches by transferring security context to potential target mesh routers before handoff occurs. However, the efficiency and effectiveness of such approaches are significantly affected by the density of potential targets and the accuracy of handoff prediction which is in fact a very difficult task. In this paper, we propose a reactive scheme to reduce re-authentication delay with low communication overhead in IEEE 802.11 wireless mesh network. Our proposed scheme consists of two steps: Immediate Authentication (lA) and Full Authentication (FA) to exploit the tradeoff between system performance and security. Numerical results show our proposed scheme can reduce re-authentication delay by as much as 51.9% within the sacrifice of the degradation of security 6.04%.

This paper proposes a modular approach to the design of hierarchical consensus protocols for the mobile ad hoc network with a static and known set of hosts. A two-layer hierarchy is imposed on the network by grouping mobile hosts into clusters, each with a clusterhead. The messages from and to the hosts in the same cluster are merged/unmerged by the clusterhead so as to reduce the message cost and improve the scalability. The proposed modular approach separates the concerns of clustering hosts from achieving consensus. A clustering function, called eventual clusterer (denoted as diamC), is designed for constructing and maintaining the two-layer hierarchy. Similar to unreliable failure detectors, diamC greatly facilitates the design of hierarchical protocols by providing the fault-tolerant clustering function transparently. We propose an implementation of diamC based on the failure detector diamS. Using diamC, we design a new hierarchical consensus protocol. As shown by the performance evaluation results, the proposed consensus protocol can save both message cost and time cost. Our proposed modular design is therefore effective and can lead to efficient solutions to achieving consensus in mobile ad hoc networks.

In this paper, we propose a distributed algorithm to elect the top K leaders among the nodes in a wireless ad hoc network. Leader election is a fundamental distributed coordination problem arising from many applications, e.g. token regeneration, directory service. However, there is no deterministic algorithm proposed for electing k leaders. In our algorithm, election is based on the weight values of the nodes, which can represent any performance related attribute such as the node's battery power, computational capabilities etc. To achieve message efficiency, coordinator nodes are first selected locally and then the coordinator nodes collect the weight information of other nodes using a diffusing computation approach. The coordinator nodes collaborate with each other to further reduce the message cost. Node failures are also considered in our design. The simulation results show that, compared with a naive solution, our proposed algorithm can elect top K leaders with much less message cost.

Gossiping has been widely used for disseminating data in large scale networks. Existing works have mainly focused on the design of gossip-based protocols but few have been reported on developing models for analyzing the fault tolerance property of these protocols. In this paper, we propose a general gossiping algorithm and develop a mathematical model based on generalized random graphs for evaluating the reliability of gossiping, i.e., to what extent gossip-based protocols can tolerate node failures, yet guarantee the specified message delivery. We analytically derive the maximum ratio of failed nodes that can be tolerated without reducing the required degree of reliability. We also investigate the impact of the parameters, namely the fanout distribution and the nonfailed member ratio, on the protocol reliability. Simulations have been carried out to validate the effectiveness of our analytic model in terms of the reliability of gossiping and the success of gossiping. The results obtained can be used to guide the design of fault tolerant gossip-based protocols.

Abstract As a widely used fault‐tolerance technique, checkpointing can be divided into three categories: independent checkpointing, coordinated checkpointing, and communication‐induced checkpointing (CIC). Independent checkpointing and coordinated checkpointing have been widely used in the design of fault‐tolerant mobile agent systems, but CIC has not attracted much attention. CIC is flexible, efficient, and scalable, which is desirable for mobile agent (MA) systems. Furthermore, CIC can be well integrated with independent checkpointing for reliable MA migrations. In this paper, we propose two CIC‐based checkpointing algorithms for MA systems. The Basic‐CIC algorithm is developed by applying the classic CIC technique to MA systems, while the deferred message processing‐based CIC (DM‐CIC) improves the time efficiency by deciding when to make the forced checkpointing. Simulation results show that DM‐CIC is stable and can help the system recover fast from failures. Copyright © 2007 John Wiley & Sons, Ltd.

Eventual leadership has been identified as a basic building block to solve synchronization or coordination problems in distributed computing systems. However, it is a challenging task to implement the eventual leadership facility, especially in dynamic distributed systems, where the global system structure is unknown to the processes and can vary over time. This paper studies the implementation of a leadership facility in infrastructured mobile networks, where an unbounded set of mobile hosts arbitrarily move in the area covered by fixed mobile support stations. Mobile hosts can crash and suffer from disconnections. We develop an eventual leadership protocol based on a time-free approach. The mobile support stations exchange queries and responses on behalf of mobile hosts. With assumptions on the message exchange flow, a correct mobile host is eventually elected as the unique leader. Since no time property is assumed on the communication channels, the proposed protocol is especially effective and efficient in mobile environments, where time-based properties are difficult to satisfy due to the dynamics of the network.

This paper studies the design of hierarchical consensus protocols for mobile ad hoc networks. A two-layer hierarchy is imposed on the mobile hosts by grouping them into clusters, each with a clusterhead. The messages from and to the hosts in the same cluster are merged/unmerged by the clusterhead so as to reduce the message cost and improve the scalability. We adopt a modular method in the design, separating clustering from achieving consensus using the clusters. The clustering function, named eventual clusterer (denoted as diamC), is designed to construct a cluster-based hierarchy over the mobile hosts in the network. Since diamC provides the fault tolerant clustering function transparently, it can be used as a new oracle (i.e. an abstract tool to provide some kind of information about the state of the system) for the design of hierarchical consensus protocols. Based on diamC, we design a new consensus protocol, which can significantly reduce the message cost of achieving consensus. We also propose an implementation of the diamC oracle based on the failure detector diamS.

Designing protocols for solving the consensus problem faces new challenges in mobile computing environments. Among others, how we can achieve message efficiency for saving resource consumption has been the focus of research. In this paper, we present the HC protocol, a message efficient consensus protocol for MANETs. We consider the widely used system model where the hosts fail by crashes and the system is equipped with Chandra-Toueg's unreliable failure detectors. Unlike existing consensus protocols, the HC protocol uses a two-layer hierarchy based on clusters to achieve message efficiency. The messages from and to the hosts in the same cluster are merged so as to reduce the message cost. However, adding such a hierarchy is not trivial. Due to host movements and failures, the hierarchy changes from time to time and this may cause message loss. In designing HC, we also propose methods to handle such message losses. Extensive simulations have been carried out to evaluate and compare the performance of the HC protocol and similar protocols in a MANET environment. Simulation results show that, in most cases, our protocol can significantly reduce both the message cost and time cost. With increases in the system scale or the percentage of faulty hosts, the advantage of our protocol becomes more obvious.

This paper investigates the use of additional synchronization messages in round-based message-passing synchronous systems. It first presents a synchronous computation model allowing a process to send such messages. The difference with respect to the traditional round-based synchronous model lies in the sending phase, where a process can first send a data message to each other process, and then, without a break, a synchronization message (their sendings can be pipelined). This model is suited to the class of local area networks where communication channels are reliable. (It is not for networks where unreliable communication requires message retransmission.) To illustrate the model, the paper presents a uniform consensus algorithm suited to this model. This algorithm, based on the rotating coordinator paradigm, allows the processes to decide in at most f + 1 rounds where f is the actual number of processes that crash in the corresponding run. (This improves the f + 2 lower bound of the traditional synchronous model.) In addition to its efficiency, the algorithm enjoys another first class property, namely, design simplicity. The paper focuses also on lower bound results, and shows that any uniform consensus algorithm designed for the proposed model, requires at least f + 1 rounds in the worst case. The proposed algorithm is consequently optimal. In that sense the paper has to be seen as an investigation of both the power and the limit of adding synchronization messages to synchronous systems built on top of local networks with reliable communication

Mobile ad hoc networks (MANETs) raise new challenges in designing protocols for solving the consensus problem. Among the others, how to design message efficient protocols so as to save resource consumption, has been the focus of research. In this paper, we present the design of such an efficient consensus protocol. We consider the system model for MANETs with host crashes, but equipped with Chandra-Toueg's unreliable failure detectors of class /spl diams/P. At most f hosts can crash where f < n/2 (n is the total number of the hosts). The protocol adopts the coordinator rotation paradigm to achieve consensus. Unlike existing consensus protocols, the proposed protocol is based on a two-layer hierarchy with hosts associated with proxies. At least f + 1 hosts act as proxies and each host is associated with one proxy host. The messages from and/or to the local hosts of the same proxy are merged so as to reduce the message cost. Moreover, the hierarchical approach can improve the scalability of the consensus protocol. Performance analysis shows that the proposed protocol can significantly save cost compared existing protocols.

Mobile ad hoc networks (MANETs) introduce new challenges to designing algorithms for solving the distributed mutual exclusion (MUTEX) problem. In this paper we propose the first permission-based MUTEX algorithm for MANETs. In order to reduce the messages cost, the algorithm uses the so called "look-ahead" technique, which enforces MUTEX only among the hosts currently competing for the critical section (CS). We propose mechanisms to handle the "doze" mode and "disconnection" of mobile hosts. The constraint of FIFO channel, which is not feasible in MANETs, is also relaxed. Using timeout, a fault tolerance mechanism is introduced to tolerate transient link and host failures. The simulation results show that the proposed algorithm works well under various mobility levels, load levels and system scales, especially when the mobility is high or load level is low. It is also shown that the algorithm has good scalability, especially when most of the requests come from a few active hosts.