1.重庆邮电大学通信与信息工程学院, 重庆 400065
2.重庆邮电大学软件工程学院, 重庆 400065
李云(1974年生),男;研究方向:无线移动通信;E-mail:liyun@cqupt.edu.cn
网络出版日期:2024-09-26,
收稿日期:2024-05-31,
录用日期:2024-08-06
移动端阅览
李云, 南子煜, 姚枝秀, 等. 面向DAG任务的分布式智能计算卸载和服务缓存联合优化[J/OL]. 中山大学学报(自然科学版)(中英文), 2024,1-12.
LI Yun, NAN Ziyu, YAO Zhixiu, et al. Joint optimization of distributed intelligent computation offloading and service caching for DAG tasks[J/OL]. Acta Scientiarum Naturalium Universitatis Sunyatseni, 2024,1-12.
李云, 南子煜, 姚枝秀, 等. 面向DAG任务的分布式智能计算卸载和服务缓存联合优化[J/OL]. 中山大学学报(自然科学版)(中英文), 2024,1-12. DOI: 10.13471/j.cnki.acta.snus.ZR20240181.
LI Yun, NAN Ziyu, YAO Zhixiu, et al. Joint optimization of distributed intelligent computation offloading and service caching for DAG tasks[J/OL]. Acta Scientiarum Naturalium Universitatis Sunyatseni, 2024,1-12. DOI: 10.13471/j.cnki.acta.snus.ZR20240181.
建立了一种有向无环图(DAG,directed acyclic graph)任务卸载和资源优化问题,旨在应用最大可容忍时延等约束实现系统能耗最小化。考虑到网络中计算请求高度动态、完整的系统状态信息难以获取等因素,最后使用多智能体深度确定性策略梯度(MADDPG,multi-agent deep deterministic policy gradient)算法来探寻最优的策略。相比于现有的任务卸载算法,MADDPG算法能够降低14.2%至40.8%的系统平均能耗,并且本地缓存命中率提高3.7%至4.1%。
A directed acyclic graph(DAG) was developed for task offloading and resource optimization, aiming to minimize system energy consumption under constraints such as maximum tolerable delay. Considering that computing requests are highly dynamic in the network and it is difficult to obtain complete system state information, the multi-agent deep deterministic policy gradient(MADDPG) algorithm is used to explore the optimal strategy. Compared to existing task offloading algorithms, the MADDPG algorithm can reduce the average system power consumption by 14.2% to 40.8%, and improve the local cache hit rate by 3.7% to 4.1%.
移动边缘计算多智能体深度强化学习计算卸载资源分配服务缓存
mobile edge computingmulti-agent deep reinforcement learningcomputation offloadingresource allocationservice caching
ALE L, KING S A , ZHANG N, 2022.D3PG:dirichlet DDPG for task partitioning and offloading with constrained hybrid action space in mobile-edge computing[J]. IEEE Internet Things J,9(19):19260-19272.
CUI Q, ZHAO X, NI W, et al, 2023.Multi-agent deep reinforcement learning-based interdependent computing for mobile edge computing-assisted robot teams[J]. IEEE Trans Veh Technol, 72(5): 6599-6610.
GAO T, TANG Q, LI J, et al, 2022. A particle swarm optimization with lévy flight for service caching and task offloading in edge-cloud computing[J]. IEEE Access, 10: 76636-76647.
GUO Y, MA D,SHE H,et al, 2024.Deep deterministic policy gradient-based intelligent task offloading for vehicular computing with priority experience playback[J]. IEEE Trans Veh Technol, 73(7): 10655-10667.
KAR B, YAHYA W , LIN Y D, et al, 2023.Offloading using traditional optimization and machine learning in federated cloud–edge–fog systems: a survey[J].IEEE Commun Surv Tutor, 25(2): 1199-1226.
KHAN L U, YAQOOB I,TRAN N H, et al, 2020. Edge-computing-enabled smart cities: a comprehensive survey[J]. IEEE Internet Things J, 7(10):10200-10232.
KONG X, WU Y, WANG H, et al, 2022. Edge computing for internet of everything: a survey[J].IEEE Internet Things J, 9(23): 23472-23485.
LIU S,YU Y,LIAN X,et al,2023. Dependent task scheduling and offloading for minimizing deadline violation ratio in mobile edge computing networks[J]. IEEE J Sel Areas Commun, 41(2): 538-554.
LIU Z, HUANG L, GAO Z, et al, 2024. GA-DRL: graph neural network-augmented deep deinforcement learning for DAG task scheduling over dynamic vehicular clouds[J]. IEEE Trans Netw Serv Manag, 2024: 1.
LUO Q , HU S , LI C, et al, 2021. Resource scheduling in edge computing: a survey[J]. IEEE Commun Surv Tutor, 23(4): 2131-2165.
LV X, DU H, YE Q, 2022.TBTOA:A DAG-based task offloading scheme for mobile edge computing[C]//IEEE International Conference on Communications, Seoul, Korea:IEEE: 4607-4612.
NATH S,WU J, 2020.Deep reinforcement learning for dynamic computation offloading and resource allocation in cache-assisted mobile edge computing systems[J]. Intell Converged Netw, 1(2): 181-198.
QIU T, CHI J, ZHOU X , et al, 2020. Edge computing in industrial internet of things: architecture, advances and challenges[J]. IEEE Commun Surv Tutor, 22(4): 2462-2488.
SAHNI Y,CAO J,YANG L, et al, 2021.Multihop offloading of multiple DAG tasks in collaborative edge computing[J]. IEEE Internet Things J, 8(6): 4893-4905.
SHEN Q, HU B J , XIA E, 2022. Dependency-aware task offloading and service caching in vehicular edge computing[J].IEEE Trans Veh Technol, 71(12): 13182-13197.
SONG T, TAN X, REN J, et al, 2023. DRAM: A DRL-based resource allocation scheme for MAR in MEC[J]. Digit Commun Netw, 9(3): 723-733.
SU S,YUAN P,DAI Y,2024.Reliable computation offloading of DAG applications in internet of vehicles based on deep reinforcement learning[J]. IEEE Trans Veh Technol, 99: 1-13
WANG J,HU J, MIN G, et al,2019. Georgalas, Computation offloading in multi-access edge computing using a deep sequential model based on reinforcement learning[J]. IEEE Commun Mag,57(5): 64-69..
WANG X,HAN V,LEUNG V C M,et al, 2020. Convergence of edge computing and deep learning: A comprehensive survey[J].IEEE Commun Surv Tutor,22(2): 869-904.
YAO Z,LI Y,XIA S,2022.Attention cooperative task offloading and service caching in edge computing[C]// IEEE Global Communications Conference.Rio de Janeiro, Brazil: IEEE: 5189-5194.
ZHANG Y, FENG B, QUAN W,et al, 2020. Cooperative edge caching: A multi-agent deep learning based approach[J]. IEEE Access, 8: 133212-133224.
ZHOU H, ZHANG Z, LI D,2023.Joint optimization of computing offloading and service caching in edge computing-based smart grid[J].IEEE Trans Cloud Comput,11(2): 1122- 1132.
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构