In the domain of computation offloading from mobile devices, Space-Air-Ground Integrated Networks ) emerge as a potent paradigm, leveraging unmanned aerial vehicles (UAVs) and low earth orbit (LEO) satellites as computational resource providers. One prevalent strategy to reducing energy consumption within SAGIN systems is to jointly optimize UAV deployment and computation offloading. However, given the variable nature of mobile device positioning and computational requisites, the deployment and offloading must be continually recalibrated. This joint optimization faces a significant challenge of combinatorial explosion. Addressing the difficulties of real-time computation tasks necessitates efficacious methodologies for devising joint optimization schemes. In this paper, we introduce a convex optimization-based algorithm to minimzie the weighted total energy consumption within the SAGIN framework by jointly optimizing UAV deployment and computation offloading. The algorithm decomposes the original problem into a set of sub-problems: UAV deployment, ground device (GD) access, and computation offloading, employing the Block Coordinate Descent (BCD) method. By alternately addressing these sub-problems, the algorithm derives a near-optimal solution efficiently. Simulation results demonstrate that our approach can generate a joint optimization solution in a few seconds and diminish the weighted total energy consumption than other classic methods by 1.5% similar to 10.97%.

Automatic aerial vehicle (UAV) aided Mobile Edge Computing (MEC) has emerged as a promising technique to offer computing support for high-mobility and high-demand mobile devices (MDs). However, due to the dynamic scale of UAVs and MDs as well as their changeable resource availability and demands, it is very challenging to quickly make suitable UAV deployment plans for satisfying the real-time requirements. Existing solutions commonly adopt the centralized decision-making manner based on the global information, which leads to poor scalability, excessive search time, and repeated training costs. To address these important challenges, we propose a novel Federated deep Reinforcement learning based UAV Deployment optimization method (FRUD) for multi-UAV-aided MEC systems, aiming to minimize the average task response time via optimizing the real-time deployment locations of large-scale UAVs. In FRUD, each UAV independently conducts the deployment decision-making based on the local information of runtime environments rather than using the global information. Next, through feedback control and multi-UAV cooperation, an effective UAV deployment plan can be gradually formed. Simulation and real-world testbed evaluation results show that the proposed FRUD well handles the UAV deployment problem in large-scale and dynamic multi-UAV-aided MEC systems and outperforms the state-of-art methods.

Sharding is a promising technique for improving the throughput and scalability of blockchain systems. However, the imbalanced distribution of transactions across shards poses a major challenge: overloaded shards lead to congestion, while underutilized shards result in wasted resources, both of which hinder system performance. To obtain high throughput, it is crucial to adjust system parameters and allocate resources efficiently. In this paper, we formulate the problem, and propose a joint convex optimization framework that optimizes system parameters and resource allocation to enhance throughput in sharded blockchain systems. Unlike previous studies that focus solely on only one of these aspects, our approach integrates both. By leveraging the block coordinate descent method from convex optimization theory, our proposed approach iteratively solves two interdependent subproblems, system parameter tuning and resource allocation, leading to a near-optimal solution with guaranteed convergence. Extensive experiments demonstrate that our joint optimization algorithm achieves a near-optimal solution within 5 seconds and improves throughput by 3.32% to 14.56% compared to benchmark methods. These results validate the effectiveness of our approach in enhancing both the throughput and scalability of sharded blockchain systems.

Serverless edge computing (SEC) is an emerging paradigm for delivering low-latency and auto-scaling services on resource-limited edges. However, when applied to DNN inference applications, SEC suffers from significant cold-start overhead, because large model parameters need to be fetched from remote model registries before execution. Prior studies reduce modelfetching latency by prefetching models in SSDs. However, these methods are impractical for SEC, as resource-constrained edge servers cannot accommodate all required DNN models. To solve it, we propose AccSI, a novel edge-assisted pipelined model fetching framework to accelerate serverless inference. AccSI aggregates distributed edge storage to prefetch DNN model layers and employs a pipelined execution strategy that overlaps model fetching, model loading, and inference, thereby minimizing the overall application completion time (ACT). To fully unlock AccSI's potential, we jointly optimize application placement and model fetching decisions with a provable approximation ratio. In addition, considering the dynamic nature of SEC environments and application requests, we adaptively adjust application placement, model fetching, and prefetching decisions at runtime. Finally, extensive experiments demonstrate that AccSI achieves a speedup of up to 3.47× in average ACT.

Unstable transmission conditions degrade the offloading performance of Mobile Edge Computing (MEC) servers in vehicular communication environments. Cache-enabled unmanned aerial vehicles (UAVs) with computation and caching capabilities offer a potential solution by acting as mobile edge servers in the air. Due to superior channel conditions, UAVs provide more reliable communication than traditional ground servers in vehicular networks. Additionally, caching popular task results in UAV storage improves service efficiency. However, challenges remain in designing offloading schemes that balance energy consumption, service revenue, and secure caching. This paper formulates a joint optimization problem based on the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm to determine UAV trajectory, offloading schemes, and UAV-side resource allocation. The objective is to minimize energy consumption and maximize service revenue in secure cache-enabled UAV-assisted vehicular networks under delay and resource constraints. To tackle the high nonlinearity and intricate action space, we propose a novel Twin Actor Network combined with a hierarchical reward mechanism, together forming the Enhanced Reward Twin Actor Twin Delayed Deep Deterministic Policy Gradient (ERTATD3) framework. Simulation results show that the proposed algorithm outperforms existing methods in terms of convergence, reward, caching security, delay, energy efficiency, and revenue optimization.

In Wireless Metropolitan Area Networks (WMANs), multi-edge collaboration has emerged as a promising solution to improve user experience and service quality by offloading tasks to multiple edges. Due to the varying service rate among edges, unreasonable task offloading may lead to resource wastage or excessive load, making efficient load balancing solutions essential. However, existing solutions commonly neglect the network topology, failing to extract deeper potential features, leading to suboptimal performance. In addition, most solutions only consider fixed or simple linear granularity to adjust policies, making it difficult to adapt to dynamic edge changes. To address these critical challenges, we propose an adaptive graph reinforcement learning method that combines Graph Convolutional Network (GCN) and Deep Q-Network (DQN), called LBGRL, for achieving load balancing of multi-edge collaboration in WMANs. To extract the potential features of network topology, LBGRL considers the system as a graph and designs a graph embedding learning process using GCN to learn the representation of nodes and edges. To further adapt to changing edge loads and response times, LBGRL designs a dynamic fine-tuning mechanism that adaptively adjusts task offloading granularity. Simulation experiments demonstrate that LBGRL outperforms the state-of-the-art methods by 1.08%-12.09% and approximates the theoretical optimal result.

With the rapid advancement of deep learning (DL), different types of DL models exhibit significant variations in their resource utilization, including CPU, GPU, storage IO, and network IO. However, scheduling DL training jobs with diverse resource bottlenecks in resource-constrained clusters to minimize makespan remains challenging. Most existing studies primarily focus on GPU allocation, which limits their ability to effectively address the multi-resource demands of various model types. To address this challenge, we propose a multi-resource sharing model that leverages the staged iteration characteristics of DL jobs, thereby enabling parallel execution through temporal resource interleaving. Next, we propose a DL training job scheduling strategy based on the Dynamic Multi-Swarm Particle Swarm Optimization Algorithm with Genetic Algorithm Operators (DMPSO-GA). This approach dynamically partitions the swarm into subswarms based on particle fitness and enhances the algorithm's exploration capability by designing differentiated update strategies for particles at different evolutionary stages within the subswarms.

Mobile Edge Computing (MEC) distributes resources such as computing, storage, and bandwidth to the side close to users, which can provide low-latency services to in-vehicle users, thus promising a more efficient and safer driving environment. However, due to the dynamic scale of vehicle and the variability of resource requirements, it is a significant challenge to quickly obtain effective task offloading in large-scale vehicle scenarios. The existing studies generally adopt the centralized decision-making method, with long decision-making time and high computational overhead, which cannot effectively achieve good offloading decisions in large-scale scenarios. To address these problems, we propose a Multi-agent Collaborative Method for vehicular task offloading using Federated Deep Reinforcement Learning called MCM-FDRL. First, each vehicle as an agent, independently makes offloading decisions based on local information. Next, the offloading decision model of each vehicle is obtained through federated reinforcement learning training. At runtime, an effective vehicle offloading plan can be gradually developed through multi-agent collaboration. Using two real-world datasets, experiments show that the MCM-FDRL has good adaptability and scalability. Moreover, compared to the state-of-the-art methods, the task's average response time of the MCM-FDRL is reduced by 9.75%-64.90%, respectively.

Vehicular Edge Computing (VEC) is a feasible solution for autonomous driving as it can offload latency-sensitive and computation-intensive tasks from vehicle terminals to roadside units (RSUs) for real-time processing. Due to the high computational resource the workflow applications (i.e., autonomous driving) required, the small vehicle-to-RSUs communication range, and the scarce available resources for each vehicle, it might be difficult to accomplish complex workflow applications through the single-hop offloading paradigm in the Internet of Vehicles (IoV). In this paper, we propose a reinforcement learning (RL) based multi-hop computation offloading scheme for workflow applications to reduce their execution latency in the VEC networks, which considers the data dependency in workflow applications as well as the trust relationship and communication interference among RSUs. Firstly, a preprocessing method is employed to merge the cut-edges in a workflow to reduce the offloading scale of tasks and compress the encoding dimension of offloading solutions. Then, a Deep Q-network algorithm using Phase-optimal State update (DQPS) is proposed to update the offloading policy distribution, in which the deviation of the phase-optimal state from the current state is estimated as the reward of RL to promote the algorithm's convergence to adapt the dynamic VEC networks. Simulation results show that DQPS has the best performance compared to other benchmark schemes. Moreover, the latency of the identical applications can be reduced by 2.5%-25.3% through our offloading scheme with a multi-hop model compared to that with the single-hop paradigm in IoV.

Computation offloading utilizes powerful cloud and edge resources to process workflow applications offloaded from Mobile Devices (MDs), effectively alleviating the resource constraints of MDs. In end-edge-cloud environments, workflow applications typically exhibit complex task dependencies. Meanwhile, parallel tasks from multi-MDs result in an expansive solution space for offloading decisions. Therefore, determining optimal offloading plans for highly dynamic and complex end-edge-cloud environments presents significant challenges. The existing studies on offloading tasks for multi-MD workflows often adopt centralized decision-making methods, which suffer from prolonged decision time, high computational overhead, and inability to identify suitable offloading plans in large-scale scenarios. To address these challenges, we propose a Multi-agent Collaborative method for Workflow Task offloading in end-edge-cloud environments with the Actor-Critic algorithm called MCWT-AC. First, each MD is modeled as an agent and independently makes offloading decisions based on local information. Next, each MD's workflow task offloading decision model is obtained through the Actor-Critic algorithm. At runtime, an effective workflow task offloading plan can be gradually developed through multi-agent collaboration. Extensive simulation results demonstrate that the MCWT-AC exhibits superior adaptability and scalability. Moreover, the MCWT-AC outperforms the state-of-art methods and can quickly achieve optimal/near-optimal performance.

Spatiotemporal Graph (STG) forecasting is an essential task within the realm of spatiotemporal data mining and urban computing. Over the past few years, Spatiotemporal Graph Neural Networks (STGNNs) have gained significant attention as promising solutions for STG forecasting. However, existing methods often overlook two issues: the dynamic spatial dependencies of urban networks and the heterogeneity of urban spatiotemporal data. In this paper, we propose a novel framework for STG learning called Dynamic Meta-Graph Convolutional Recurrent Network (DMetaGCRN), which effectively tackles both challenges. Specifically, we first build a meta-graph generator to dynamically generate graph structures, which integrates various dynamic features, including input sensor signals and their historical trends, periodic information (timestamp embeddings), and meta-node embeddings. Among them, a memory network is used to guide the learning of meta-node embeddings. The meta-graph generation process enables the model to simulate the dynamic spatial dependencies of urban networks and capture data heterogeneity. Then, we design a Dynamic Meta-Graph Convolutional Recurrent Unit (DMetaGCRU) to simultaneously model spatial and temporal dependencies. Finally, we formulate the proposed DMetaGCRN in an encoder-decoder architecture built upon DMetaGCRU and meta-graph generator components. Extensive experiments on four real-world urban spatiotemporal datasets validate that the proposed DMetaGCRN framework outperforms state-of-the-art approaches.

The emerging load prediction techniques support up-front and rational resource provisioning in edge systems to enhance system efficiency and Quality-of-Service (QoS). Classic prediction methods may handle loads with apparent trends, but they cannot achieve accurate prediction for highly-variable edge loads. With the advantage of sequential data analysis, recurrent neural networks (RNNs) are often used for load prediction but reveal limited generalization ability and low training efficiency. Moreover, it is hard to obtain a well-performed prediction model by discrete single-edge training with insufficient historical data. To address these important challenges, we propose a novel Multi-edge Cooperative universal framework for load Prediction with Personalized Federated deep learning (MC-2PF), enabling multi-edge cooperative training of load prediction models. Specifically, to solve the client-drift issue in federated learning (FL) caused by distinct data distribution, we customize personalized models for each edge by independent control parameters and theoretically analyze the model convergence improvement. Meanwhile, we prove the generalization bound of the MC-2PF and its universality to RNN-based prediction models through a practical example. Using the real-world testbed and load datasets, extensive experiments verify the effectiveness and practicality of the MC-2PF for different RNN-based prediction models. Compared to state-of-the-art frameworks, the MC-2PF achieves higher prediction accuracy, faster convergence, and stronger adaptiveness.

Mobile Edge Computing (MEC) can augment the capability of Internet of Things (IoT) mobile devices (MDs) through offloading the computation-intensive tasks to their adjacent servers. Synergistic computation offloading among MEC servers is one possible solution to reduce the completion time of system during peak hours. However, due to the large number of servers and the long distance between base stations (BSs), synchronizing the information of all servers takes a long time, which is not applicable to the fluctuant environments. Meanwhile, each server from different BSs is typically selfish and rational, and can only obtain the imperfect information from its adjacent servers, which is a challenge for computation offloading among servers from a global perspective. This article proposes a game-based computation offloading scheme with imperfect information in multi-edge environments. First, a non-cooperative game with imperfect information is designed to analyze the complex interactions during synergistic computation offloading among MEC servers. Second, a Synergistic Balancing Offloading Algorithm (SBOA) through distributed decision-making manner to obtain the optimal offloading decision is proposed, which guarantees that the game converges to a Nash Equilibrium (NE) point. Extensive simulation results reveal the fast convergence of SBOA. As the percentage of high-load servers rises and the number of heavy tasks increases, SBOA performs better than other benchmark algorithms in terms of timeliness, effectiveness, and system completion time.

GUI testing ensures the software quality and user experience in the ever-changing mobile application development. Using test scripts is one of the main GUI testing manner, but it might be obsolete when the GUI changes with the app’s evolution. Current studies often rely on textual or visual similarity to perform test repair, but may be less effective when the interacted event sequence changes dramatically. In the interaction design, practitioners often provide multiple entry points to access the same function to gain higher openness and flexibility, which indicates that there may be multiple routes for reference in test repair. To evaluate the feasibility, we first conducted an exploratory study on 37 tests from 18 apps. The result showed that over 81% tests could be represented with alternative event paths, and using the extended paths could help enhance the test replay rate. Based on this finding, we propose a test-extension-based test repair algorithm named ExtRep. The method first uses test-extension to find alternative paths with similar test objectives based on feature coverage, and then finds repaired result with the help of sequence transduction probability proposed in NLP area. Experiments conducted on 40 popular applications demonstrate that ExtRep can achieve a success rate of 73.68% in repairing 97 tests, which significantly outperforms current approaches Water, Meter, and Guider. Moreover, the test-extension approach displays immense potential for optimizing test repairs. A tool that implements the ExtRep is available for practical use and future research.

The development of the Space-Air-Ground Integrated Network (SAGIN) represents a significant advancement in satellite Internet technology, primarily because it facilitates the provision of pervasive networking solutions for the Internet of Things (IoT). Central to minimizing energy consumption in this context is the strategic deployment of Unmanned Aerial Vehicles (UAVs) and the efficient offloading of computational tasks. This is due to the fact that the quality of communication between UAVs and ground-based mobile devices is a crucial determinant of both communication latency and energy usage. This study presents an in-depth exploration of a Mobile Edge Computing (MEC) framework within a SAGIN context, in which multiple UAVs serve as edge computing servers, thereby accelerating computational services to terrestrial devices. In parallel, Low Earth Orbit (LEO) satellites are leveraged to provide cloud computing services. We introduce a Dual-layer nested joint optimization technique fortified by an improved differential evolution. The outer layer of the algorithm employs a phased differential evolution technique to optimize the positioning of UAVs. Simultaneously, the inner layer adopts a greedy approach to optimize the offloading ratio for computational tasks. The effectiveness and superior performance of the proposed method are substantiated through extensive numerical simulations. Comparative analysis with existing benchmark techniques demonstrates a significant reduction in the energy consumption of the system, underscoring the potential of the proposed framework.

Through deploying computing resources at the network edge, Mobile Edge Computing (MEC) alleviates the contradiction between the high requirements of intelligent mobile applications and the limited capacities of mobile End Devices (EDs) in smart communities. However, existing solutions of computation offloading and resource allocation commonly rely on prior knowledge or centralized decision-making, which cannot adapt to dynamic MEC environments with changeable system states and personalized user demands, resulting in degraded Quality-of-Service (QoS) and excessive system overheads. To address this important challenge, we propose a novel Personalized Federated deep Reinforcement learning based computation Offloading and resource Allocation method (PFR-OA). This innovative PFR-OA considers the personalized demands in smart communities when generating proper policies of computation offloading and resource allocation. To relieve the negative impact of local updates on global model convergence, we design a new proximal term to improve the manner of only optimizing local Q-value loss functions in classic reinforcement learning. Moreover, we develop a new partial-greedy based participant selection mechanism to reduce the complexity of federated aggregation while endowing sufficient exploration. Using real-world system settings and testbed, extensive experiments demonstrate the effectiveness of the PFR-OA. Compared to benchmark methods, the PFR-OA achieves better trade-offs between delay and energy consumption and higher task execution success rates under different scenarios.

Due to the complicated data dependencies between the tasks in a workflow application and the heterogeneous resources in edge-cloud environments, it is difficult to select an optimal tasks-servers solution for scheduling workflow applications in the complex environments. Current research on workflow applications scheduling is mainly concentrated on certain conditions, ignoring the fact that the scheduling environments usually fluctuate. In this article, we deal with reducing the execution cost of multiple workflow applications within the corresponding deadline constraints and improving the network robustness in fuzzy edge-cloud environments. Triangular Fuzzy Numbers (TFNs) are employed to describe the computing capacity of servers and the bandwidth between them in uncertain environments. Specially, a novel Scheduling Strategy based on Particle Swarm Optimization algorithm employing the Quadratic Penalty Function (SSPSO_QPF) is proposed for scheduling multiple workflow applications. Compared with other classic scheduling strategies, simulation results demonstrate that the proposed strategy can generate feasible scheduling schemes even with the strict deadline constraints, and significantly reduce the fuzzy execution cost of multiple workflow applications.

As an effective technique to relieve the problem of resource constraints on mobile devices (MDs), the computation offloading utilizes powerful cloud and edge resources to process the computation-intensive tasks of mobile applications uploaded from MDs. In cloud-edge computing, the resources (e.g., cloud and edge servers) that can be accessed by mobile applications may change dynamically. Meanwhile, the parallel tasks in mobile applications may lead to the huge solution space of offloading decisions. Therefore, it is challenging to determine proper offloading plans in response to such high dynamics and complexity in cloud-edge environments. The existing studies often preset the priority of parallel tasks to simplify the solution space of offloading decisions, and thus the proper offloading plans cannot be found in many cases. To address this challenge, we propose a novel real-time and Dependency-aware task Offloading method with Deep Q-networks (DODQ) in cloud-edge computing. In DODQ, mobile applications are first modeled as Directed Acyclic Graphs (DAGs). Next, the Deep Q-Networks (DQN) is customized to train the decision-making model of task offloading, aiming to quickly complete the decision-making process and generate new offloading plans when the environments change, which considers the parallelism of tasks without presetting the task priority when scheduling tasks. Simulation results show that the DODQ can well adapt to different environments and efficiently make offloading decisions. Moreover, the DODQ outperforms the state-of-art methods and quickly reaches the optimal/near-optimal performance.

Workflow decision making is critical to performing many practical applications of scientific principles and data. Scheduling in edge-cloud environments can address the high complexity of workflow applications, while decreasing the data transmission delay between the cloud and end devices. However, due to the heterogeneous resources in edge-cloud environments and the complicated data dependencies between the tasks in a workflow, significant challenges for workflow scheduling remain, including the selection of an optimal tasks-servers solution from the possible numerous combinations. Existing studies are mainly done subject to rigorous conditions without fluctuations, ignoring the fact that workflow scheduling is typically present in uncertain environments. In this study, we focus on reducing the execution cost of multiple workflow applications mainly caused by data transmission and task computation, while satisfying the required deadline constraints. Triangular fuzzy numbers are employed to represent the computing performance of servers and transmission bandwidth in fuzzy edge-cloud environments. A cost-driven scheduling strategy for multiple Poisson-arrived workflow applications using partial critical paths is proposed. It firstly merges cut edges through preprocess to reduce the workflow scale, then uniformly schedules all tasks on each partial critical path to avoid data transmission between dependent tasks and reduce the data transmission cost. The experimental results show that our strategy can obtain the optimal feasible scheduling scheme and have better robustness and real-time performance with different deadline constraints, compared with other benchmark strategies. Note to Practitioners —Vehicle identification is one of the workflow decision making systems in transportation environments, whose core technology is Deep Neural Networks (DNN). Traffic cameras with limited process capacity periodically record the images of on-road vehicles, and usually fail to complete the applications within their deadlines. Workflow decision making is one of the key issues to performance DNNs in vehicle identification applications. The uncertain environments have a great impact on the system latency for such problems, which can easily lead to the misjudgement of the optimal scheduling. In addition, it is difficult to select an optimal layers-servers solution from the numerous combinations. Therefore, we can employ the scheduling strategy (i.e., SWPCP) to make intelligent and faster workflow decisions for vehicle identification applications, which can reduce the execution cost mainly caused by layer computation and data transmission between layers within their deadlines, even in uncertain edge-cloud environments. Complex DNN layers (tasks) in vehicle identification applications can be scheduled to the cloud for execution, while simple ones are processed on the edge. The cloud and edge platforms collaborate with each other and execute the DNN layers with low system cost and latency.

Mobile edge computing (MEC) offers a promising technology that deploys computing resources closer to mobile devices for improving performance. Most of the existing studies support on-demand remote execution of the computing tasks in applications through program transformation, but they commonly assume that mobile devices merely resort to a single server for computation offloading, which cannot make full use of the scattered and changeable computing resources. Thus, for object-oriented applications, we propose a novel approach, called FUNOff, to support the dynamic offloading of applications in MEC at the function granularity. First, we extract a call tree via code analysis and locate the function invocations that are suitable for offloading. Next, we refactor the code of related object functions according to a specific program structure. Finally, we make offloading decisions referring to the context at runtime and send function invocations to multiple remote servers for execution. We evaluate the proposed FUNOff on two real-world applications. The results show that, compared with other approaches, FUNOff better supports the computation offloading of object-oriented applications in MEC, which reduces the response time by 10.7%-58.2%.