Text-attributed graphs (TAGs), which associate rich textual descriptions with each node, are widely employed to represent complex relationships among real-world textual entities. Currently, representation learning for TAGs leverages large language models (LLMs) to transform node-matched textual descriptions into node features or labels, followed by the message passing in graph neural networks (GNNs) that further improves the expressiveness of graph representation learning. Nevertheless, a simple experiment we conducted demonstrates that not all LLMs are readily compatible with GNNs. A salient finding indicates that architectural heterogeneity among LLMs manifests as substantial performance gap across diverse TAGs representation learning. Moreover, the node semantics encoded by LLMs are often misaligned with the message passing in GNNs, causing performance collapse. Motivated by this observation, we propose a novel self-supervised graph learning framework called Stage-Aware Graph Contrastive Learning (SAGCL). In particular, we propose the node-oriented mixture of experts (NodeMoE) to assign suitable candidate experts for each node. It flexibly balances the strengths of different language experts by low-rank decomposition and reparameterization strategies. Subsequently, to align the inductive biases of graph structures with the semantic perception capabilities of LLMs, the message passing in GNNs is decoupled into the feature transformation stage and the feature propagation stage. Given the two stage views, stage-aware graph contrastive learning is proposed to match the node semantics encoded by the LLM with the locally aware topological patterns within the GNN via self-supervised contrastive learning. Experiments on eight datasets and three downstream tasks demonstrate the effectiveness of SAGCL.

In recent years, Mobile Edge Computing (MEC) has been widely used for latency-sensitive tasks, but task scheduling in dynamic edge environments still faces two key challenges. First, edge devices are prone to failures, and existing fault-tolerance mechanisms lack task-aware modeling, making it hard to ensure timeliness and reliability under failures. Second, due to limited perception, high communication costs, and complex task structures, current scheduling strategies still struggle with adaptability and stability in dynamic systems. In this paper, we propose a Fault-Tolerant Discrete Soft Actor-Critic scheduling algorithm (FT-DSAC). Initially, we design a Primary-Backup-based Fault-Tolerant (PBFT) scheduling mechanism, which constrains task offloading locations and start times to effectively mitigate the impact of failures on task execution. Furthermore, we incorporate the Centralized Training and Distributed Execution (CTDE) architecture, which enables implicit collaborative scheduling decisions among edge servers to optimize system performance and reduce communication overhead. Finally, We conduct extensive experiments using both simulated data generated by DAGGEN and real-world workflow data. Experimental results show that the proposed algorithm significantly improves task execution success rates by 6%-19% and reduces latency by 9%-27% compared to mainstream benchmarks.

With the rapid advancement of large language models (LLMs) in both academia and industry, their growing size and complexity have introduced significant challenges in terms of computational cost and deployment efficiency. To address these issues, a wide range of inference optimization techniques-including but not limited to model compression-have been proposed to accelerate LLM inference while preserving model performance. This survey provides a comprehensive overview of LLM inference acceleration strategies, analyzing them from multiple perspectives, including foundational principles, algorithmic techniques, real-world applications, and open research challenges. We begin by introducing core concepts underlying inference optimization and propose a new taxonomy that categorizes existing approaches, including quantization, pruning, distillation, efficient architectures, compilation, and hardware-aware methods. Following the lifecycle of LLM development and deployment, we examine how these techniques interact with model training, fine-tuning, and serving. Furthermore, we highlight key applications of efficient LLMs and discuss emerging trends and unresolved issues in the field. By synthesizing recent advances, this survey aims to provide actionable insights and practical guidance for researchers and practitioners working with scalable and efficient LLM systems.

Nowadays, Mobile Crowd Sensing (MCS) has become a popular paradigm for large-scale data collection using ubiquitous mobile sensing devices. However, most existing works do not consider that requester's payments are unknown prior, and assume that workers are honest, which may not be true in practice. To address these problems, we propose a novel Truthful and Dual-direction Combinatorial Multi-Armed Bandit (TD-CMAB) scheme, which maximizes the total profit of the dual-direction platform for both the worker side and the requester side. Specifically, for the worker side, to overcome the problem that the platform is not clear whether sensed data are true, we propose a worker recruitment strategy that identifies and recruits honest workers at low cost through the Upper Confidence Bound (UCB) algorithm based on truth data discovery. For the requester side, where requesters’ payments are unknown prior, we model requester selection as a CMAB problem and solve it by the proposed adaptive UCB algorithm. Furthermore, we theoretically prove the worst regret bound of the TD-CMAB. Finally, we evaluate the effectiveness of the TD-CMAB scheme through extensive experiments using the Beijing taxi dataset.

The escalating prevalence of Advanced Persistent Threat (APT) malware demands more effective methods to accurately attribute malware to specific APT groups. Traditional manual attribution processes are labor-intensive and error-prone, while existing automated methods are hampered by small dataset sizes, inadequate representation learning, and poor noise reduction during preprocessing. To address these challenges, we introduce the AMG25 dataset, which expands the pool of malware samples labeled with APT group affiliations. Concurrently, we propose the MGAP3model (Malware Group Attribution based on PerceiverIO and Polytype Pre-training), which enhances attribution performance by incorporating hierarchical pre-training for disassembled codes and leveraging multi-view statistical features, all within a unified PerceiverIO architecture. This model adeptly captures complex program structures and interactions cross multiple code granularities, through a series of innovative polytype pre-training tasks. Additionally, we have developed a novel noise filtering technique that focuses on user-defined function codes, substantially reducing overfitting and boosting performance. Furthermore, a streamlined version of the model, MGAP3-Lite, has been developed to accelerate training while preserving robust performance. Extensive experiments have validated the effectiveness of our models and underscored the importance of the proposed pre-training technique.

Noise perturbation introduced by differential privacy (DP) could degrade the quality of essential services like dynamic pricing and ride-matching in ride-on-demand (RoD) services. In this paper, we focus on RoD services under an honest-but-curious server, and propose a Pricing-Aware Differentially Private framework (PADP-RoD) to protect users’ location privacy while providing them with high-quality location-based services. Specifically, given that a price multiplier is subject to abrupt changes in response to shifts in supply and demand, especially near hotspots, we propose an adaptive supply and demand aware grid to capture the changes. Powered by the grid, we put forward two utility metrics for quantifying the quality loss of dynamic pricing and ride-matching services caused by perturbation, respectively. With those metrics, PADP-RoD is formulated as a minimization problem, aiming to minimize the quality loss of services given DP constraint. In this way, we can achieve an optimal balance between privacy and service quality. Due to the problem being a multi-objective optimization, we decompose it into a dynamic-pricing utility sub-problem and a ride-matching utility sub-problem, and solve them separately. To solve the dynamic pricing utility sub-problem, we propose a heuristic algorithm named the dynamic pricing mapping algorithm. Since the semi-infinite and non-differentiable nature of the ride-matching utility sub-problem, we transform this sub-problem into an unconstrained problem by the exact penalty function method, and solve it employing the particle swarm optimization algorithm. Our theoretical analysis demonstrates that PADP-RoD satisfies both$\varepsilon _{d}$-DP and$\varepsilon _{d}$-identifiability, and extensive experiments on a real-world dataset show that it can provide high-quality dynamic pricing and ride-matching services.

Multi-task assignment is widely used in mobile crowdsensing (MCS) to efficiently utilize limited resources such as shared user pool, user capability constraints and so on. In MCS, users need to submit data streams to perform sensing tasks, which involve a large amount of private information. However, the privacy leakage when users perform tasks across different types and submit multimodal data streams in multi-task assignment has not been fully addressed in current works. Privacy requirements vary for users with different activity levels in multi-task assignment. Specifically, users with higher activity levels tend to handle more task types and submit more data types, which poses more serious consequences of privacy leakage. Meanwhile, the privacy requirements of users are dynamic due to the user’s changing activity. In this work, we propose a user-driven local differential privacy framework for multi-task assignment called UD-LDP. First, we design a flexible privacy model called$w$-adjacent-event privacy to provide accurate privacy protection for users with different activity levels. Then, we introduce information entropy to quantify privacy requirements of user’s activity in real-time. After that, we propose a privacy-aware budget allocation method to dynamically allocate personalized privacy budgets for each user. At last, we design a variance-optimized selection method that chooses rational privacy budgets and users for release to improve data utility. The effectiveness of our framework is supported by experiments conducted on both real-world and synthetic datasets.

Mobile Edge Computing (MEC) facilitates computing and storage at edge nodes near user devices, reducing latency and optimizing bandwidth. Federated Learning (FL) complements MEC by enabling privacy-preserving collaborative model training across edge nodes without sharing raw data. However, in MEC environments, FL faces challenges such as communication inefficiency and data heterogeneity (Non-IID), which degrade model performance and hinder convergence. To address these issues, we propose FedLFP, a communication-efficient personalized federated learning approach using label-free prototypes for Non-IID data in MEC. FedLFP employs three key strategies: (1) a Label-Free Prototype strategy to reduce communication costs and mitigate privacy risks, (2) a centroid prototype and combined clustering weight strategy to improve global prototype quality by considering data quantity and confidence levels, and (3) a multifaceted weighted contrastive learning strategy to enhance local representation learning and global alignment. We evaluated FedLFP on Android malware recognition using the KronoDroid dataset and standard image classification tasks, with eight configurations representing practical Non-IID settings. Experimental results show that FedLFP consistently outperforms thirteen state-of-the-art FL methods in accuracy, communication and computational efficiency. Additionally, we provide theoretical guarantees for the convergence of FedLFP under Non-IID conditions.

Differential privacy (DP) is widely used for protecting privacy in crowdsensing by adding noises. However, malicious attackers can exploit noise to launch covert data poisoning attacks. In this paper, we propose a game-based defense approach to resist such data poisoning attacks in DP-based crowdsensing systems. In this approach, attackers are believed to be powerful as they can refine their attack strategy based on the observations of deployed defenders’ defense strategy. Specifically,the defendersformulate the defense as a functional minimization problem (which cannot be directly solved by numerical optimization algorithms because its decision variable is a set of functions), resisting data poisoning attacks by deleting data shared by identified malicious workers through the log-likelihood ratio test. To obtain a current defense strategy, the decision variable of the problem is relaxed into the coefficients of basis-based linear combinations through the variable-basis approximation, and then solved using the simulated annealing genetic algorithm. Correspondingly,the attackersformulate their attack strategy as a bi-level maximization problem (which is an NP-hard problem), biasing crowdsensing results as much as possible while remaining undetected. Since the attackers can know the defense strategy, they may bypass the defenders by constraining the expected log-likelihood ratio test. Additionally, the attackers can evade truth discovery methods deployed in crowdsensing using DP noise. To determine a current attack strategy, the bi-level problem is decomposed into upper-level and lower-level sub-problems, wherein the upper-level sub-problem is solved by the variational methods, and then these sub-problems are alternately optimized. Finally, we propose a local minimax points calculating algorithm to obtain an equilibrium point in the defenders-attackers game, thereby finding an optimal defense strategy to resist the powerful data poisoning attack. Extensive experiments on real-world and synthetic datasets show that the proposed game-based defense approach can effectively defend powerful and covert attackers.

Power consumption prediction is one of the keys to optimize the energy consumption of servers. Existing traditional regression-based methods are too simple and poorly generalized, while popular deep learning methods require too much data. Therefore, they are difficult to be widely generalized. In this study, we propose a framework of linear interpolation multi-head sparse temporal pattern attention (Li-MSA) based on few-shot learning for power consumption prediction of servers with small-scale datasets in environments such as cloud data centers or edge computing. First, the interpolation reconstruction module extends and smooths the data. Then, the embedding learning module is used to narrow the scope of the hypothesis space. Finally, the multi-head sparse temporal pattern attention module emphasizes features and predicts power consumption. The results of the experiments show that Li-MSA outperforms the best results among the other methods for two datasets with different time steps in the RMSE metric by 15.34%, 17.35%, 18.18%, 6.28%, 4.05%, 7.73%.

Fine-tuning Pre-trained Language Models (PLMs) is a popular Natural Language Processing (NLP) paradigm for addressing Named Entity Recognition (NER) tasks. However, neural network models often demonstrate poor generalization capabilities due to significant disparities between the knowledge learned by PLMs and the distribution of the target dataset, as well as data scarcity issues. In addition, token omission in predictions due to insufficient learning remains a challenge in NER. In this paper, we propose a kNN retrieval enhancement algorithm (WkNER) that incorporates word segmentation information to enhance the model’s generalization ability and alleviate the problem of missing entity tokens in prediction. The introduction of word segmentation information is used to preliminarily determine the boundaries of entities and alleviate the common prediction errors of missing tokens within entities made by the fine-tuned model. Secondly, we find that non-entities in the retrieval table contain a large amount of redundant information, and explore the effects of introducing non-entity information of different scales on the model. Experimental results show that our proposed method significantly improves the performance of baseline models, and achieves better or compared recognition accuracy than previous state-of-the-art models in multiple public Chinese and English datasets. Especially in low-resource scenarios, our method achieves higher accuracy on 20% of the dataset than the original method on the full dataset.

The time-varying characteristics of user mobility, node and services connections, and edge resources in Mobile Edge Computing (MEC) scenarios pose significant challenges for designing efficient service migration strategies to enhance system performance. This paper introduces and quantifies affinity and anti-affinity metrics to evaluate the discrepancies between the resource requirements of containers and the available resources of nodes, as well as the competition level of computing resources during the migration of containers among different nodes within the Kubernetes cluster. To adapt to the complexity of the edge environment, such metrics are integrated into the reward update mechanism of the Deep Deterministic Policy Gradient (DDPG) reinforcement algorithm. Besides, the Multi-Objective Evolutionary Algorithm (MOEA) is employed to dynamically adjust the weights of various reward objectives, forming a self-adaptive online container migration strategy named MO-DDPG. Finally, we construct a real-world heterogeneous Kubernetes edge node cluster in experiments and use a public dataset to simulate multi-modal connections between mobile user trajectories and service demands. Compared to the greedy and heuristic strategies that consider only single metrics, our experiment results show that the proposed strategy improves energy efficiency and reduces latency by 21.30% and 29.49%, respectively. Moreover, the MO-DDPG improves resource utilization of the node cluster compared to the default Kubernetes scheduler.

With the development of loT and edge computing, data analytics on edge has become popular, and text analytics directly on compression (TADOC) has been proven to be a promising technology for edge data analytics. At the same time, Field Programmable Gate Array (FPGA) also has broad application prospects in data analytics systems. Unfortunately, there is no work to date showing how to support TADOC using FPGAs. We propose FPGA-based text analytics directly on compression with HLS, namely F - TADOC, which is the first framework using HLS to provide FPGA-based text analytics directly on compressed data. It effectively supports efficient text analytics on FPGA without decompressing input data. F-TADOC addresses three major challenges. First, TADOC involves a large number of dependencies with unbalanced workload of rules, which causes extremely low pipeline efficiency on FPG As. To solve it, we use layer-wise approach to traverse the DAG composed of rules and allocate different pipeline processing strategies for rules of different sizes. Second, the data volume required can be large that beyond the on-chip memory capacity of FPGAs. We develop a memory pool supporting hash structure and on-chip caches on FPGA to deal with this challenge. Third, when traversing the DAG, there are massive indirect addressing with a large number of random accesses. This leads to redundant time overhead caused by the latency in accessing the High Bandwidth Memory (HBM) during the pipeline. We optimize the F - TADOC algorithm by using dataflow to expand the nested loop, thus eliminate indirect addressing. With four widely used datasets, experiments show that F - TADOC achieves 4.63 x and 1.49 x performance speedup over TADOC and G- TADOC.

Edge Intelligence aims to push deep learning (DL) services to network edge to reduce response time and protect privacy. In implementations, proximity deployment of DL models and timely updates can improve the quality of experience (QoE) for users, but increase the operation cost as well as pose a challenge for task assignment. To address the challenge, a joint online optimization problem for DL model deployment (including placement and update) and freshness-sensitive task assignment is formulated to improve QoE and application service provider (ASP) profit. In the problem, we introduce the age of information (AOI) to quantify the freshness of the DL model and represent user QoE as an AOI based utility function. To solve the problem, an online model placement, update, and task assignment (MPUTA) algorithm is proposed. It first converts the time-slot coupled problem into a single time-slot problem using the regularization technique, and decomposes the single time-slot problem into model deployment and task assignment subproblems. Then, using the randomized round technique to deal with the model deployment subproblem and the graph matching technique to solve the task assignment subproblem. In simulation experiments, MPUTA is shown to outperform other benchmark algorithms in terms of both user QoE and ASP profit.

In order to achieve high availability and low storage costs in distributed storage systems, erasure code is widely used instead of replication. Compared to replication, erasure code can reduce storage costs, but also brings higher repair costs. There are currently many repair algorithms to reduce the block reconstruction time of single block failure. However, applying the existing methods to multi-block failures may lead to unbalanced network traffic, unnecessary network transfers, and network congestion at data collection node during the repair process, which can not make full use of the bandwidth between nodes.To solve this problem, we propose a novel repair algorithm called Partial Merge Repair (PMR) for multi-block failures, which is a scheduling algorithm that considers network load between nodes and combines multiple failed blocks to recover together. It first divides all surviving nodes into different groups, and then the data collection nodes within the group collect the data needed to repair multiple blocks through cross merging. Finally, the data collection node sends the collected blocks to the repair node to complete the repair. Our study presents a formal definition and proof of network transfer time in the modeled repair process of PMR, highlighting its superior efficiency compared to existing methods in homogeneous environments.We implement a prototype of PMR to evaluate its performance. The experimental results indicate that compared to existing repair technologies, PMR improves repair throughput by 28%-256% for various scenes.

Existing methods for decomposing monolithic applications into microservices in cloud environments primarily rely on the call relationships within itself. However, these methods are difficult to apply directly in resource-constrained and distributed edge network scenarios without considering the heterogeneity of the device. Therefore, this paper proposes a clustering method that ensembles graph structures and device features based on attention mechanism, which utilizes attention encoders to learn node embeddings and employs a spectral clustering algorithm to obtain decomposition results, optimizing the affinity and matching degree between microservices and devices. Experimental results demonstrate that the proposed method exhibits excellent performance in terms of functional independence, modularity, and adaptability of microservices.

Recent years witness the increasing popularity of ride-on-demand (RoD) services such as Uber and Didi. Compared with traditional taxi, RoD service is more “data-driven” and adopts dynamic pricing to manipulate the supply and demand in real time. Dynamic price could be viewed as an accurate and quantitative indicator of the supply and demand, and could provide clues to drivers, passengers, and the service providers, possibly reshaping the ways in which some problems are solved. In this paper, we focus on the seeking route recommendation problem that aims at increasing driver revenue by recommending highly profitable seeking routes to drivers of vacant cars with the help of dynamic prices. We first justify our motivation by showing the importance of route recommendation and answering why it is necessary to consider dynamic prices, based on the analysis of real service data. We then design a dynamic price prediction model to generate the dynamic prices at any given time and location based on multi-source urban data. After that, a reinforcement learning model is adopted to perform seeking route recommendation based on predicted dynamic prices. We conduct extensive experiments in different spatio-temporal combinations and make comparisons with multiple baselines. Results first show that our dynamic price prediction model achieves an accuracy ranging from 83.82% to 90.67% under different settings. It also proves that considering the real-time predicted dynamic prices significantly increases driver revenue by, for example, 12% and 47.5% during weekday evening rush hours, than merely using the average prices or completely ignoring dynamic prices.

Mobile Crowd Sensing (MCS) has become a prevalent large-scale and low-cost data collection paradigm by employing workers, and the location and bid privacy of both task and workers should not be leaked to the third party to prevent the adversary from attacking. Existing privacy preserving worker recruitment schemes have taken the location and quality into consideration, but ignore the bid privacy. To tackle this issue, a two-stage Location and Bid Privacy Preserving based Quality-aware Worker Recruitment (LBPP-QWR) scheme is proposed in this paper. In the first stage, to select those workers who satisfy the specified location and bid range of the task in the encrypted state, we propose a hybrid encryption scheme of matrix encryption and asymmetric encryption technique in the MCS platform. For the second stage, after obtaining the preliminary worker set via the platform, we propose a Knapsack Worker Selection (KWS) algorithm to recruit those high-quality and low bid workers under the budget constraint in the Data Requester (DR). Considering that there are quality-unknown workers, we further propose an improved.-KWS algorithm based on.-greedy algorithm by combining the exploration and exploitation mechanism to learn the quality of worker. Extensive experiments conducted on real-world datasets demonstrate that our proposed scheme can improve the average total quality by 17.96%-83.34%, and the cost efficiency by 27.99%-67.90% for the DR compared with other benchmark methods.

Mobile crowdsourcing (MCS) endeavors to attain reliable truth by recruiting large numbers of users with handheld mobile devices to collect the data. However, during the early stages of platform development, MCS encounters the cold start problem, failing to complete the task. Existing research addresses this issue by leveraging social networks for user recruitment. Nevertheless, there is a predominant focus on the user quantity, and the quality of task completion is ignored. Additionally, fairness considerations among users are lacking. Therefore, this article proposes recruitment based on social users’ trust (RSUT) to solve the cold start problem while maintaining high task completion quality. Specifically, we propose the activation model based on the user awareness to simulate the influence of social users and task attributes on activation from the perspective of unregistered users, which is more realistic. Additionally, we measure the user’s contribution and then design a reward system based on the user’s contribution to ensure fairness. Finally, social network-based trust evaluation is proposed to identify malicious users and update rewards in real time according to task requirements to ensure high-quality completion of tasks within budget constraints. Extensive experimental results demonstrate the superior performance of RSUT compared to the state-of-the-art methods in task completion quality, user recruitment, and task completion rate.

Deploying deep neural networks (DNNs) with satisfactory performance in resource-constrained environments is challenging. This is especially true of microcontrollers due to their tight space and computational capabilities. However, there is a growing demand for DNNs on microcontrollers, as executing large DNNs on microcontrollers is critical to reducing energy consumption, increasing performance efficiency, and eliminating privacy concerns. This paper presents a novel and systematic data redundancy elimination method to implement efficient DNNs on microcontrollers through innovations in computation and space optimization. By making the optimization itself a trainable component in the target neural networks, this method maximizes performance benefits while keeping the DNN accuracy stable. Experiments are performed on two microcontroller boards with three popular DNNs, namely CifarNet, ZfNet and SqueezeNet. Experiments show that this solution eliminates more than 96% of computations in DNNs and makes them fit well on microcontrollers, yielding 3.4-5$\times$speedup with little loss of accuracy.

Edge-cloud networks face security threats during data collection, data routing, and service construction, resulting in data tampering, stealing, and communication interruption. Trust mechanism can predict data quality and cooperation probability of nodes before purchasing data or establishing cooperation, so as to select trusted participants for data perception and interaction. However, there are some problems with existing trust methods, such as limited evaluation scope, incomplete trust evidence, and inaccurate evaluation results. To address these issues, a Trust mechanism-based Multi-Tier Computing system (TMTC) is proposed in this paper. Specifically, we propose a two-tier trust evaluation model. At the data collection layer, it conducts trust evaluation on data reporters based on data submission and communication interactions. At the network layer, it evaluates trust of routers through path backtracking verification, multi-service analysis and coincident path analysis. Then, based on evaluation results, a differentiated trust detection is initiated for normal and abnormal nodes. And high-frequency detection tasks are initiated for malicious nodes to improve accuracy, sparse detection tasks are initiated for normal nodes to reduce costs. Finally, extensive experiments conducted on the synthetic and real-world datasets demonstrate that, TMTC can resist data tampering and good-bad mouth attacks effectively. And whether in a dense or uniform scene, it outperforms two benchmark methods by increasing malicious node detection rate by 13.37%-21.87% and reducing cost by 18.8%-50.32%.

In this paper, we explore data poisoning attacks and their defenses in local differential privacy (LDP)-based crowdsensing systems. First, we construct data poisoning attacks launched by corrupted workers to subvert crowdsensing results by tampering information reported. Specifically, the attacks are formulated as a bi-level optimization problem where attackers strive to conceal their malicious behavior by delicately exploiting noise perturbation introduced by LDP protocols. In this way, the attacks can not be detected, even with the weight-based truth discovery methods. Due to the NP-hard nature of the bi-level problem, we decompose it into upper-level and lower-level sub-problems and employ the augmented Lagrangian method to iteratively solve them, ultimately identifying optimal attack strategies. Second, we propose corresponding countermeasures to defend against the attacks. The countermeasures are formulated as a minimization problem, with the objective of minimizing disruptions caused by attacks through the identification and removal of corrupted workers from crowdsensing systems. To solve the problem, we utilize a differential evolution algorithm instead of gradient-based methods since the objective function of the problem is not differentiable. Extensive experiments on real-world datasets are conducted to evaluate the performance of the proposed attacks and defenses. The evaluation results demonstrate that LDP perturbation indeed facilitates the success of data poisoning attacks, and the proposed defenses can accurately distinguish malicious behaviors disguised.

Sparse Mobile CrowdSensing (SMCS) effectively lowers sensing costs while maintaining data quality, offering an alternative approach to data collection. Unfortunately, the fact that data contain sensitive information raises serious privacy concerns. Local Differential Privacy (LDP) has emerged as the de facto standard for ensuring data privacy. However, the LDP based on the perturbation concept causes a substantial reduction in the data utility of the SMCS system. To address this problem, we propose a novel scheme named enhancing Sparse mobile crowdsensing With manifold Optimization and differential Privacy (SWOP). Specifically, we first revisit the Gaussian mechanism based on the fact that data utility intervals are ubiquitous in sensing tasks, and introduce a novel perturbation mechanism, namely Truncated Gaussian Mechanism (TGM). Subsequently, we perturb user-collected data by locally injecting noise sampled from TGM and deduce a sufficient condition for the scale parameter to ensure ϵ-LDP. Furthermore, we model the data inference with privacy-preserving properties as an unconstrained optimization problem on a Riemannian manifold and solve it using the nonlinear conjugate gradient method. Extensive experiments on large-scale real-world and synthetic datasets are conducted to evaluate the proposed scheme. The results demonstrate that SWOP can greatly enhance the utility of data inference while ensuring workers’ data privacy compared to baseline models.

Unmanned aerial vehicle (UAV) edge computing systems provide easy-to-deploy and low-cost services at those areas with inadequate infrastructure by deploying UAVs as moving edge servers for large-scale users. However, user devices are generally distributed unevenly in a large area, which makes it difficult for existing efforts to cope with this realistic scenario for optimal deployment of UAVs. Therefore, this paper considers a multiple UAV (Multi-UAV) Collaborative edge Computing (UCC) system by utilizing collaboration among them to split computation tasks at UAVs to balance the load and improve resource utilization. In order to maximize the energy-efficiency of the UCC system under the satisfaction of the delay constraint, we study the joint problem of UAV deployment, task collaborative offloading, computation and communication resource allocation in UCC system. We propose a bi-level optimization framework to solve the formulated non-convex mixed-integer optimization problem. In the upper level, the UAV deployment is optimized based on an improved differential evolution (DE) algorithm, and in the lower level the offloading decision and resource allocation are optimized based on a Reinforcement Learning (RL) algorithm with Twin Delayed Deep Deterministic policy gradient. Experimental results demonstrate the effectiveness and superiority of multi-UAV collaborative computing, with the proposed framework achieving a 32.4% reduction in energy consumption and an average 30% increase in task completion rate compared to DDPG, ToDeTaS, and other benchmark schemes.

As automobiles continue to develop in the direction of intelligence and networking, the requirements for in-vehicle network bandwidth and deterministic time delay continue to increase. However, existing in-vehicle network standard protocols such as Controller Area Network (CAN) cannot meet the increasing bandwidth needs of in-vehicle networks, and Time Sensitive Networking (TSN) has emerged a research hot-spot for next-generation in-vehicle network standards. The next-generation in-vehicle network is developing towards a domain network architecture with TSN as the backbone and other conventional buses as branches. In this architecture, the TSN-CAN gateway is an important component that handles the data communication between the TSN domain and CAN. Due to the large difference in transmission rates between TSN and CAN, the TSN-CAN heterogeneous gateway suffers from congestion resulting in unguaranteed real-time transmission of messages across the gateway. To address this issue, a high response ratio priority scheduling algorithm (HRRP) for TSN-CAN gateways based on worst-case response time analysis theory is proposed in this paper. The algorithm assigns forwarding priority to CAN messages based on the value of their response ratio, and the experimental demonstrate show that the method can significantly improve the schedulability and reduce latency, improving the real-time performance of the system.

The potentially malicious devices in D2D networks may spread forged videos to compromise system reliability. The prevailing passive trust model solutions have limitations, such as insufficient and inaccurate trust evidence, as well as weaker resistance to collusion attacks. This paper presents theTrustEvaluation JointActiveDetection (TEAD) method, which employs content correctness as trust evidence and allows devices to verify the authenticity of received videos via the base station. TEAD incorporates an active detection method to proactively determine the trustworthiness of devices. This promotes interaction among devices, resulting in an increase in trust evidence and an improvement in the accuracy of evaluation. Moreover, TEAD introduces a trust calculation method with a penalty mechanism to strengthen the system's resilience against malicious attacks. Empirical results show that TEAD outperforms state-of-the-art methods by achieving a more accurate trust evaluation and faster trust convergence with low extra energy consumption.

Federated class-incremental learning (FCIL) allows multiple clients in a distributed environment to learn models collaboratively from evolving data streams, where new classes arrive continually at each client. Some existing works in FCIL combine traditional federated learning methods with class-incremental methods. However, the global model affected by data heterogeneity can aggravate local forgetting through the direct combination of traditional methods. To tackle this issue, we propose FCIDF, a novel Federated Class-Incremental learning approach based onDynamic feature extractor Fusion. FCIDF learns personalized and incremental models for each client by introducing personalized fusion rates to integrate global knowledge into local features. Leveragingmeta-learningduring each incremental round, FCIDF ensures involvement of both old and new task knowledge in personalized training. Besides, we further propose a new Storing strategy based on Accumulated Global Feature Means (AGFMS), which helps the model review unbiased old knowledge and compensates for local forgetting. Experiment results show that FCIDF outperforms the baseline methods in both accuracy and forgetting on most settings, and AGFMS improves the performance of FCIDF on most evaluated scales.

Mobile crowdsourcing aims to recruit enough workers holding mobile devices to collect data. Nevertheless, the platform will have cold start problems when the number of workers is limited. Existing studies have proposed solving this problem by propagating tasks to social networks for social recruitment. However, they neglect to verify workers’ propagation, leading to malicious workers reducing the platform's utility. Furthermore, during propagation verification, it is imperative to protect the privacy of social relationships among workers, as it can significantly influence the propagation. Therefore, this paper proposes Zero-knowledge Propagation Verification based on Social Relationship Encryption (ZPV-SRE) to improve the platform's utility. Specifically, we transform the propagation verification problem into a problem of computing the solution of the function. Then, the Zero-knowledge proof is used to prove the propagation, in which the worker's social relationship is protected through homomorphic encryption. Considering that ZPV-SRE will incur a significant time cost, we propose Trust-guided Zero-knowledge Propagation Verification based on Social Relationship Encryption (TZPV-SRE), which updates the worker's trust based on the verification results and selects suspicious workers for verification. The experimental results show ZPV-SRE improves the platform's utility as high as 104.05% over the state-of-the-art methods, while TZPV-SRE reduces time costs and ensures improvement.

Sparse Mobile CrowdSensing (MCS) aims to reduce sensing cost while ensuring high task quality by intelligently selecting small regions for sensing and accurately inferring the remaining areas. Data inference and cell selection are crucial components in Sparse MCS. However, cell division, which is a prerequisite for cell selection, has received insufficient attention. The existing uniform division method disregards the correlation of the sensing area. In addition, the impact of sparse noise on both data inference and cell selection has been ignored, potentially undermining the effectiveness of Sparse MCS. To address these issues, we propose a novel scheme termed Robust data Inference and Cost-Effective cell Selection for Sparse MCS (Rices). Specifically, we first design an adaptive region division strategy that captures the correlation of sensing regions. Subsequently, we tackle the robust data inference problem in the presence of sparse noise by formulating it as a dual-objective optimization. Furthermore, we optimize the cell selection strategy to dynamically adjust the set of sampled cells under the constraints of data inference quality. Extensive experiments on large-scale real-world datesets are conducted to evaluate the proposed scheme. The results demonstrate that Rices can accurately recover missing data with 20% sparse noise and significantly reduce sensing costs compared to baseline models.

Incentive mechanisms are essential to incentive workers carrying mobile handheld devices to participate in mobile crowd sensing and finally achieve good truth discovery performance. However, malicious workers may report false or malicious data to defraud rewards, resulting in service quality degradation. Moreover, the existing incentive mechanism is challenging to identify malicious workers when recruiting workers in reality, which results in low accuracy of truth discovery and waste of cost. In this paper, we propose an Incentive-based Truth Discovery (ITD) scheme to incentive credible workers to submit high-quality data, thereby enhancing the accuracy of truth discovery. In the ITD, an unmanned aerial vehicle (UAV)-assisted split-aggregation truth discovery mechanism is proposed firstly to infer the truth. The addition of the UAV can improve the accuracy of truth discovery and assist in evaluating workers’ trust. Then, we evaluate the quality of participants’ data and propose a data quality-based trust meter to update each worker’s trust to guide future recruitment efforts. Finally, a Quality-aware Trustworthy Incentive (QTI) mechanism is proposed to select credible workers for data collection and provide them with reasonable payments. The experimental results show that ITD improves the accuracy of truth discovery by 98.47%, over the state-of-the-art, at a sensing cost reduced by as high as 43.89%.

Many artificial intelligence applications based on convolutional neural networks are directly deployed on mobile devices to avoid network unavailability and user privacy leakage. However, the significant increase in model parameter volumes makes it difficult to achieve high-performance convolutional neural network inference on these mobile devices with limited computing power. Weight pruning is one of the main approaches to compress models by reducing model parameters and computational operations, which also introduces irregular sparsity of neural networks, leading to inefficient computation and memory access during inference. This work proposes an end-to-end framework, namely MCPruner, for efficient inference of pruned convolutional neural networks on mobile devices by aligning the sparse patterns with hardware execution features in computation, memory access, and parallelism. It first co-designs pruning methods and code generation optimizations for the alignment of non-zero weight count and vector width, to improve computational efficiency while ensuring accuracy. During the code generation, it applies a sparse pattern-aware format to reduce inefficient memory accesses. Besides, convolution computations are reordered for alignment, and then mapped to parallel threads on accelerated units to achieve high parallelism. Experimental results using several commonly used models and datasets on the ARM-based Hikey970 demonstrate that our work outperforms state-of-the-art methods in inference efficiency, with no accuracy degradation.

The widespread adoption of mobile smart devices has ushered in the era of Mobile Crowd Sensing (MCS), serving as an efficient method for large-scale data collection. Inherently location-sensitive, the service construction of MCS faces a crucial challenge of Location Privacy Preservation (LPP). Prior studies for LPP often necessitate a Trusted Third Party (TTP), which is not always feasible. Moreover, these privacy-preserving techniques may inadvertently obscure dishonest or malicious behaviors, leading to compromised Quality of Service (QoS). Motivated by this, we propose a Quality-enabled Bilateral Location Privacy-Preserving (Q-BLPP) service construction scheme, ensuring Bilateral LPP without TTP, while maintaining QoS. To achieve bilateral LPP, we introduce a novel BI-LBE algorithm using Bloom Indexing (BI) and Location-Based Encryption (LBE). Additionally, for high-quality recruitment, we present a Combinatorial Multi-Armed Bandit (CMAB) approach to balance exploration and exploitation. Furthermore, to ensure privacy during recruitment, worker profiles are anonymized using differential privacy. To our knowledge, our approach is the first to integrate QoS and LPP in MCS, with theoretical proofs of truthfulness and individual rationality. Simulations demonstrate that our Q-BLPP scheme strikes a favorable balance between computational efficiency, privacy security, and service quality, outperforming existing schemes.

In recent years, a number of lightweight single-image super-resolution (SISR) network methods heave been proposed. However, most existing approaches do not make full use of the information before and after the convolution and the high-frequency information of the image. In this paper, we propose a lightweight deep feature aggregation network (DFAnet), which fuses the outputs of all the deep feature aggregation blocks (DFAB) through the designed nonlinear global feature fusion (NGFF) module. The DFAB includes deep feature aggregation structure (DFAS) and non-local sparse attention mechanism (NLSA), where DFAS consists of several aggregation convolutions and information rearrangement operations. Then the output of DFAS is assessed by non-local sparse attention module to form our basic block DFAB. Furthermore, we design a nonlinear global feature fusion (NGFF) module to learn the nonlinear relationship between the output of each DFAB, which encourages every DFAB to pay attention to different patterns of the image. The qualitative and quantitative experimental results on several benchmark datasets show the proposed method achieves the state-of-the-art results in term of reconstruction accuracy, computational complexity and memory consumption.

In this paper, a novel task offloading architecture called Flex-MEC is proposed, which achieves efficient task allocation and scheduling (TAS) between MEC servers. By adding metadata before task data, we redesign the offloading process in Flex-MEC, the TAS planning can be conducted without finishing the task data receiving. Once planning is done the task data can be directly forwarded to the allocated server and executed. This reduces latency compared to the traditional way of transmitting, planning, forwarding and executing sequentially. For TAS planning, a multi-server multi-task allocation and scheduling (MMAS) problem is formulated to maximize the MEC system profit. The MMAS problem is proven as an NP-complete problem, thus is challenging to solve. Then, a distributed scheme and a centralized scheme are proposed to solve the MMAS problem with low complexity. In the distributed scheme, the MMAS problem is converted into a non-cooperative game and the existence of Nash Equilibrium (NE) is proven and a low complexity response update algorithm is proposed to converge to NE. And the centralized scheme is based on a greedy idea and runs on a MEC controller in a centralized way. Verified by experiments, these two schemes can achieve better performance than compared schemes.

Multi-tier computing (MC) is a promising architecture that integrates cloud computing, fog computing, and edge computing to provide users with a consistent experience of computing services by fusing computing devices within the network through virtualization technology. Although MC combines powerful computation and communication resources, the massive demand from Service Function Chain (SFC) deployments continues to make it challenging regarding resource constraints, latency satisfaction, and revenue-cost tradeoffs. To this end, in this article, we study an SFC deployment problem in MC and formulate a problem for maximizing the revenue of online SFC deployment under latency, computation resources, and communication resources constraints. To solve this online problem better, we construct a computation and communication resource cost model and transform the original online problem into a deployment cost minimization problem and a request admission problem by an alternating optimization approach. To solve the two subproblems, we propose an online approximation algorithm with a provable competitive ratio for the particular scenario with no latency requirements. Then, based on the cost model, we propose an online heuristic algorithm that adopts a binary search method for the original problem with latency requirements. Simulation experiments show that our two proposed online algorithms have advantages in total revenue, running time, and load balancing compared with other comparison algorithms.

With the development of multi-access edge computing (MEC), the cloudlet at the edge of the network can provide nearby high-performance computing services, thus reducing the computational consumption of user equipments (UEs). To provide more real-time computing services to UEs, service providers face the challenge of optimizing the assignment of requests and the allocation of cloudlets’ computing resources to achieve low latency while dealing with the large number of offloaded requests from UEs. Therefore, in this paper, we study the problem of minimizing the total latency to complete the requests in the MEC network by jointly optimizing request assignment and computing resource allocation. The problem is formulated as a mixed integer nonlinear programming (MINLP) problem which is NP-hard. To solve the problem, we decompose the problem into two subproblems which respectively optimize the request assignment and the computing resource allocation. We first deal with the computing resource allocation problem by utilizing the Lagrangian multiplier method, and the resulting solution is applied for the request assignment problem. Then a novel primal-dual based approximation algorithm is devised to address the request assignment problem. Finally, to verify the efficiency of the proposed algorithm, we provide an upper bound on the approximation ratio. The experiment results show that the proposed algorithm outperforms baseline algorithms in terms of total latency, loading balancing, and computational speed.

Cloud computing and mobile edge computing techniques supply efficient ways to solve the contradiction between the increasing computing and storage demands of portable terminals and the limited capacity. In this paper, we conduct a three-tier hierarchical service system with multiple UEs, multiple MECs, and a single cloud center. It's worth noting that multiple UEs with personalized options generate a large number of different tasks in real time. To deal with this offloading problem, a response ratio offloading strategy (RROS) centered on user preference and real-time nature is designed to make MECs or CC serve as many UEs as possible. Therefore, a MEC-choosing preference list of each UE is created based on its past experiences at first. Then, each MEC iteratively sorts UEs with its ranking in the UEs' preference list. In order to avoid that the first task arriving at MEC occupies too many resources of MEC and cannot achieve global optimization, we also adopt loop iterative sequencing for multiple tasks arriving within a stipulated time. Lastly, by comparing the optimal response ratio on different MECs and CC, multiple MECs and the CC collaborative offload computing tasks of multiple UEs. Experimental results show that the algorithm significantly outperforms conventional techniques.

Recently, the siamese convolutional neural network plays an important role in the field of visual tracking, which can obtain high tracking accuracy and good real-time performance. However, the requirement of offline training a specific neural network results in the hardware source and time consumption. In order to improve the tracking efficiency and save computation resources, we adopt pre-trained densely connected neural network to extract robust target features. Since the pre-trained model is mainly used for classification task, it is not appropriate to directly adopt these deep features for visual tracking. We design a regression network to measure the importance of each channel to the target, and then propose a weighting fusion strategy to select the suitable features for visual tracking. Besides, we provide deep analysis about the proposed channel weighting method to demonstrate the superiority of this method through visualization of feature heatmaps. Extensive experiments on four classical benckmarks show that compared with state-of-the-art methods, our algorithm achieves the best results on several standard indicators and comparable results on other indicators.

Large-scale Internet applications running on data centers are typically instantiated as a set of containers. Assigning a container to its affinity machine can reduce communication and transport costs while assigning it to the anti-affinity machine may affect the proper operation of the container. Existing container scheduling methods cannot accommodate these two types of requirements. In order to reduce the operation and maintenance cost of data centers, this paper focuses on the container instance allocation problem in heterogeneous server cluster, and proposes a global cost-aware scheduling algorithm (GCCS) to solve it. The purpose is to minimize the total power consumption of the cluster from a global perspective, while trying to meet the affinity/anti-affinity requirements of applications. We study the number of containers per server selected by the application, model it as an integer linear program (ILP), and then propose a heuristic search algorithm to repair the relaxation solution of the ILP into a suboptimal feasible solution. In particular, we use Bayesian optimizer to perform a number of automated development and exploration processes for the selection of the cost coefficient. The experiments are carried out with the best cost coefficient recommended by Bayesian optimizer. Finally, the results demonstrate that GCCS can significantly reduce the total power consumption of the cluster, while maintaining a high affinity satisfaction ratio.

With the aggravation of data explosion and backhaul loads on 5 G edge network, it is difficult for traditional centralized cloud to meet the low latency requirements for content access. The federated learning (FL)-basedproactive contentcaching (FPC) can alleviate the matter by placing content in local cache to achieve fast and repetitive data access while protecting the users’ privacy. However, due to the non-independent and identically distributed (Non-IID) data across the clients and limited edge resources, it is unrealistic for FL to aggregate all participated devices in parallel for model update and adopt the fixed iteration frequency in local training process. To address this issue, we propose a distributed resources-efficient FPC policy to improve the content caching efficiency and reduce the resources consumption. Through theoretical analysis, we first formulate the FPC problem into a stacked autoencoders (SAE) model loss minimization problem while satisfying resources constraint. We then propose an adaptive FPC (AFPC) algorithm combined deep reinforcement learning (DRL) consisting of two mechanisms of client selection and local iterations number decision. Next, we show that when training data are Non-IID, aggregating the model parameters of all participated devices may be not an optimal strategy to improve the FL-based content caching efficiency, and it is more meaningful to adopt adaptive local iteration frequency when resources are limited. Finally, experimental results in three real datasets demonstrate that AFPC can effectively improve cache efficiency up to 38.4$\%$and 6.84$\%$, and save resources up to 47.4$\%$and 35.6$\%$, respectively, compared with traditional multi-armed bandit (MAB)-based and FL-based algorithms.

With the development of the Internet of Things, the data that needs to be processed is increasing rapidly. Therefore, the collaboration of cloud and edge emerges as the times require. Edge nodes are mainly responsible for collecting data, and decide to process the data locally or offload to cloud data centers. Cloud data centers are suitable for data analysis, model training, and managing edge nodes. In this article, we focus on the task assignment problems in collaborative edge and cloud environments and study it in a distributed, non-cooperative environment. An M/M/1 queueing model is established to characterize the task transmission. Because of the multi-core processors, we set an M/M/C queueing model to characterize the task computation. We consider the problem from the perspective of game theory and formulate it into a non-cooperative game among multi-agents (multiple edge data centers) in which each agent is informed with incomplete information (allocation strategies) of others. For each agent, we define a function of the expected cost of tasks as the disutility function, and minimize it subject to the QoS constraint. We analyze the existence of Nash equilibrium and develop a Greedy Energy-aware Algorithm (GEA) to choose active servers using the Limit Searching Algorithm (LSA) to find the ceiling utilization. Then we propose the Best Response Algorithm (BRA) to optimize the utility function. The convergence of the BRA algorithm has been discussed. Finally, the results demonstrate that the BRA algorithm can get a solution close to Nash equilibrium and reach it quickly.

Convolutional neural networks (CNNs) have been widely used in various complicated problems, such as image classification, objection detection, semantic segmentation. To meet diversified CNN structures, the deep learning processing unit (DPU) is designed as a general accelerator on field programmable gate array (FPGA) to support various CNN layers, such as convolution, pooling, activation, etc. However, low DPU utilization and schedule efficiency appear when DPU used to multitask application completed by CNN models. In this paper, an effective design including multi-core with different size (MCDS) and DPU Plus is proposed to improve the efficiency of DPUs usage from the two dimensions of time and space. Through increasing the number of DPU cores on an FPGA and the utilization of single DPU core, the design of MCDS can effectively improve the overall throughput with restricted on-chip resources. Furthermore, the design of DPU Plus is proposed to improve the schedule efficiency of DPUs through simultaneously implementing DPU with other significant auxiliary modules of the application system on the same FPGA. Finally, a color space conversion module is implemented cooperate to the DPU cores to testify its performance, and the experimen shows that compared with running on the the CPU completely, it achieves16.2x acceleration, and increases the throughput of the entire system by 3.0x.

Deduplication is a data redundancy elimination technique, designed to save system storage resources by reducing redundant data in cloud storage systems. With the development of cloud computing technology, deduplication has been increasingly applied to cloud data centers. However, traditional technologies face great challenges in big data deduplication to properly weigh the two conflicting goals of deduplication throughput and high duplicate elimination ratio. This paper proposes a similarity clustering-based deduplication strategy (named SCDS), which aims to delete more duplicate data without significantly increasing system overhead. The main idea of SCDS is to narrow the query range of fingerprint index by data partitioning and similarity clustering algorithms. In the data preprocessing stage, SCDS uses data partitioning algorithm to classify similar data together. In the data deletion stage, the similarity clustering algorithm is used to divide the similar data fingerprint superblock into the same cluster. Repetitive fingerprints are detected in the same cluster to speed up the retrieval of duplicate fingerprints. Experiments show that the deduplication ratio of SCDS is better than some existing similarity deduplication algorithms, but the overhead is only slightly higher than some high throughput but low deduplication ratio methods.