复杂移动边缘计算场景中基于动态信任评估的任务卸载方案

Task Offloading Scheme Based on Dynamic Trust Evaluation in Complex Moblie Edge Com- puting

摘要

任务卸载是移动边缘计算中的关键研究方向。尽管现有研究在优化计算任务时延和能耗方面取得了显著进展，但多数研究未充分考虑边缘计算场景的复杂性。边缘环境中资源节点不可靠性，设备异构性及任务多样性等因素，影响了任务卸载决策和系统性能。因此，本文提出了基于动态信任评估的任务卸载方案。该方案结合了两种算法：首先，仿照人类社会信任的演化机制，在边缘网络中构建了设备间的信任关系，为任务卸载提供可靠的资源节点信息。通过该机制，可有效避免因设备故障，恶意攻击或资源不足等问题引发的卸载失败。其次，采用改进的 Q-learning 算法求解任务卸载问题，以实现系统成本最小化的优化目标。本文提出的方案相较于启发式方案，系统成本降低了 16.3% ，任务成功率提升了 32.1% 。同时，实验进一步验证了信任值机制在筛选可靠资源节点方面的有效性。

Abstract

Task offloading represents a pivotal research direction within mobile edge computing.While existing studies have achieved remarkable progress in optimizing computation latency and energy consumption，most fail to adequately address the inherent complexi- ties of edge computing environments.Factors such as the unreliability of resource nodes，device heterogeneity，and task diversity signif- icantly affect offloading decisions and overall system performance.To address these challenges，this paper proposes a task offloading scheme grounded in dynamic trust evaluation.The proposed approach integrates two algorithms ：Firstly，drawing inspiration from the e- volution of trust mechanisms in human society，a trust relationship model between devices is established within the edge network.This mechanism ensures the provision of reliable resource node information for task offloading，effectively mitigating failures caused by de- vice malsunctions，malicious allacks，or resource insufficiency.Subsequently，an improved Q-learning algorithm is employed to solve the task offloading problem，aiming to minimize system costs.Compared to heuristic scheme，the proposed scheme reduces system costs by 16.3% and improves task success rates by 32.1%.Furthermore，experimental results validate the effectiveness of the trust- value mechanism in identifying reliable resource nodes.

关键词

移动边缘计算 / 任务卸载 / 强化学习 / 信任值

Key words

mobile edge computing / task offloading / reinforcement learning / trust value

参考文献

[1] Schneider M, Haag F, Khalil A K, et al. Evaluation of communica- tion technologies for distributed industrial control systems :concept and evaluation of 5G and WiFi 6[J]. Procedia CIRP, 2022, 107: 588-593.

[2] Cisco. Annual internet report (2018-2023) white paper[EB/OL]. https://www.cisco.com/c/en/us/solutions/collateral/executive-per-spectives/annual-internet-report/white-paper-c11-741490.html,2023.

[3] Nguyen H, Nawara D, Kashef R. Connecting the indispensable roles of IoT and artificial intelligence in smart cities:a survey[J]. Jour- nal of Information and Intelligence, 2024, 2(3):261-285.

[4] Hu F, Deng Y, Saad W, et al. Cellular-connected wireless virtual reality:Requirements,challenges,and solutions[J]. IEEE Commu- nications Magazine, 2020, 58(5):105-111.

[5] Siriwardhana Y, Porambage P, Liyanage M, et al. A survey on mo- bile augmented reality with 5G mobile edge computing:architec- tures,applications,and technical aspects[J]. IEEE Communications Surveys ＆ Tutorials, 2021, 23(2):1160-1192.

[6] Garikapati D, Shetiya S S. Autonomous vehicles:evolution of artifi- cial intelligence and the current industry landscape[J]. Big Data and Cognitive Computing, 2024, 8 ( 4 ):42,doi:10.3390/bd-cc8040042.

[7] Saeik F, Avgeris M, Spatharakis D, et al. Task offloading in edge and cloud computing:a survey on mathematical,artificial intelli- gence and control theory solutions[J]. Computer Networks, 2021, 195:108177,doi:10.1016/j.comnet.2021.108177.

[8] Islam A, Debnath A, Ghose M, et al. A survey on task offloading in multi-access edge computing[J]. Journal of Systems Architecture, 2021,118:102225,doi:10.1016/j.sysarc.2021.102225.

[9] Ning Z, Dong P, Kong X, et al. A cooperative partial computation offloading scheme for mobile edge computing enabled Internet of Things[J]. IEEE Internet of Things Journal, 2018, 6 (3): 4804-4814.

[10] Du J, Than I, Feng I, et al. Computation offloading and resource allocation in mixed fog/cloud computing systems with min-max fairness guarantee[J]. IEEE Transactions on Communications, 2017, 66(4):1594-1608.

[11] Zhu C, Tao J, Pastor G, et al. Folo:latency and quality optimized task allocation in vehicular fog computing[J]. IEEE Internet of Things Journal, 2018, 6(3):4150-4161.

[12] Ketykó I, Kecskés L, Nemes C, et al. Multi-user computation off- loading as multiple knapsack problem for 5G mobile edge compu- ting[C]// European Conference on Networks and Communica- tions, 2016:225-229.

[13] Guo H, Liu J, Zhang J. Computation offloading for multi-access mobile edge computing in ultra-dense networks[J]. IEEE Commu- nications Magazine, 2018, 56(8):14-19.

[14] Qu G, Wu H, Li R, et al. DMRO:a deep meta reinforcement learn- ing-based task offloading framework for edge-cloud computing[J]. IEEE Transactions on Network and Service Management, 2021, 18(3):3448-3459.

[15] Lu H, Gu C, Luo F, et al. Optimization of lightweight task offload- ing strategy for mobile edge computing based on deep reinforce- ment learning[J]. Future Generation Computer Systems, 2020, 102:847-861,doi:10.1016/j.future.2019.07.019.

[16] Zhu X, Luo Y, Liu A, et al. Multiagent deep reinforcement learning for vehicular computation offloading in IOT[J]. IEEE Internet of Things Journal, 2020, 8(12):9763-9773.

[17] Jošilo S, Dán G. Wireless and computing resource allocation for selfish computation offloading in edge computing[C]// IEEE IN- FOCOM-IEEE Conference on Computer Communications, 2019: 24672475.

[18] Mao Y, Zhang J, Letaief K B. Dynamic computation offloading for mobile-edge computing with energy harvesting devices[J]. IEEE Journal on Selected Areas in Communications, 2016, 34 (12): 3590-3605.

[19] Yuan J, Li X. A multi-source feedback based trust calculation mechanism for edge computing[C]// IEEE INFOCOM-IEEE Con- ference on Computer Communications Workshops, 2018:819-824.

[20] Kong W, Li X, Hou L, et al. A reliable and efficient task offloading strategy based on multifeedback trust mechanism for IoT edge com- puting[J]. IEEE Internet of Things Journal, 2022, 9 (15): 13927-13941.

[21] Aghapour Z, Sharifian S, Taheri H. Task offloading and resource allocation algorithm based on deep reinforcement learning for dis- tributed AI execution tasks in IoT edge computing environments[J]. Computer Networks, 2023,223:109577,doi:10.1016/j.com-net.2023.109577.

[22] Zhao X, Liu M, Li M. Task offloading strategy and scheduling opti- mization for internet of vehicles based on deep reinforcement learn- ing[J]. Ad Hoc Networks, 2023,147:103193,doi:10.1016/j.ad-hoc.2023.103193.

[23] Dab B, Aitsaadi N, Langar R. Q-learning algorithm for joint com- putation offloading and resource allocation in edge cloud[C]// IF- IP/IEEE Symposium on Integrated Network and Service Manage- ment, 2019:45-52.

[24] Evendar E, Mansour Y. Learning rates for Q-learning[J]. Journal of Machine Learning Research, 2003, 5:1-25,doi:10.1007/3-540-44581-1_39.

[25] Koenig S, Simmons R G. Complexity analysis of real-time rein- forcement learning[C]// Association for the Advance of Artificial Intelligence, 1993:99-105.

[26] Mechalikh C, Taktak H, Moussa F. PureEdgeSim:a simulation framework for performance evaluation of cloud,edge and mist computing environments[J]. Computer Science and Information Systems, 2021, 18(1):43-66.

基金资助

国家自然科学某金项月(62372174)