面向硬件感知的边缘计算卸载和资源分配

doi:10.13190/j.jbupt.2021-143

北京邮电大学学报 ›› 2022, Vol. 45 ›› Issue (2): 22-28.doi: 10.13190/j.jbupt.2021-143

面向硬件感知的边缘计算卸载和资源分配

郅佳琳, 王楠, 满毅, 滕颖蕾

北京邮电大学电子工程学院, 北京 100876

收稿日期:2021-07-13 发布日期:2021-12-16
通讯作者: 滕颖蕾(1983—),女,教授,邮箱:lilytengtt@bupt.edu.cn。 E-mail:lilytengtt@bupt.edu.cn
作者简介:郅佳琳(1999—),女,硕士生。
基金资助:
国家重点研发计划项目(2021YFB3300100);国家自然科学基金项目(62171062)

Hardware Model-Aware Joint Offloading and Resources Allocation Optimization

ZHI Jialin, WANG Nan, MAN Yi, TENG Yinglei

School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China

Received:2021-07-13 Published:2021-12-16

摘要/Abstract

摘要： 传统的边缘计算卸载研究并未涉及计算机硬件实现的细节,计算模型建模粗糙,优化方案精准度低。为此,提出了基于硬件实现的多用户多边缘服务器计算卸载和资源分配联合优化方案,充分考虑了计算过程硬件实现的细节,从计算机指令执行粒度出发,综合计算机输入/输出瓶颈和内存功能模块的能耗,重新建立联合优化模型,并在满足卸载任务时延要求的前提下最小化系统能耗。此外,为解决动作空间高维的问题,采用了基于深度确定性策略梯度的混合在线二部匹配算法。仿真结果表明,计算过程中的内存能耗不可忽略,且所提出的优化算法能够有效学习最优策略,对降低系统能耗具有显著作用。

关键词: 移动边缘计算, 计算机流水线, 在线二部匹配, 深度强化学习

Abstract: The traditional researches on edge computing offloading do not involve the details of computer hardware implementation. In addition, their computing model is rough, and the optimization scheme has low accuracy. To solve these problems, a hardware-based scheme that jointly optimizes computing offloading and resource allocation for multi-user multi-edge server is proposed. In particular, the details of hardware implementation in the calculation process are considered. From the perspective of the granularity of computer instruction execution, the input/output bottleneck and the energy consumption of memory functional modules are calculated first. Then, the joint optimization model is re-established. Finally, the system energy consumption is minimized on the premise of meeting the delay requirement of offloading tasks. To solve the high-dimensional problem of action space, a hybrid online two-part matching kuhn-munkras algorithm based on deep deterministic policy gradient is adopted. The simulation results show that the memory energy consumption cannot be ignored. Besides, the proposed optimization algorithm can effectively learn the optimal strategy and has a significant effect on reducing the system energy consumption.

Key words: mobile edge computing, computer pipelining, online bipartite matching, deep reinforcement learning

中图分类号:

TN911.22

郅佳琳, 王楠, 满毅, 滕颖蕾. 面向硬件感知的边缘计算卸载和资源分配[J]. 北京邮电大学学报, 2022, 45(2): 22-28.

ZHI Jialin, WANG Nan, MAN Yi, TENG Yinglei. Hardware Model-Aware Joint Offloading and Resources Allocation Optimization[J]. Journal of Beijing University of Posts and Telecommunications, 2022, 45(2): 22-28.

参考文献

[1] XU F M, YANG F, BAO S J, et al. DQN inspired joint computing and caching resource allocation approach for software defined information-centric Internet of Things network[J]. IEEE Access, 2019(7):61987-61996.
[2] WU Y H, WANG Y H, ZHOU F H, et al. Computation efficiency maximization in OFDMA-based mobile edge computing networks\[J]. IEEE Communications Letters, 2020, 24(1):159-163.
[3] ALE L H, ZHANG N, FANG X J, et al. Delay-aware and energy efficient computation offloading in mobile-edge computing using deep reinforcement learning\[J]. IEEE Transactions on Cognitive Communications and Network-ing, 2021, 7(3):881-892.
[4] ZHAN W, LUO C, WANG J, et al. Deep reinforcement-learning-based offloading scheduling for vehicular edgecomputing\[J]. IEEE Internet of Things Journal, 2020, 7(6):5449-5465.
[5] PANDEY V, JIANG W H, ZHOU Y Y, et al. DMA-aware memory energy management\[C]//The Twelfth International Symposium on High-Performance Computer Architecture, 2006. Piscataway, NJ:IEEE Press, 2006:133-144.
[6] CAI X J, JU L, LI X, et al. Energy efficient task allocation for hybrid main memory architecture\[J]. Journal of Systems Architecture the Euromicro Journal, 2016, 71(6):12-22.
[7] BHATI I, CHISHTI Z, LU S L, et al. Flexible auto-refresh:enabling scalable and energy-efficient DRAM refresh reductions\[C]//ISCA.[S.l.]:ACM, 2015:235-246.
[8] LIU A, LAU V K N, DING W C, et al. Mixed-timescale online PHY caching for dual-mode MIMO cooperative networks\[J]. IEEE Transactions on Wireless Communications, 2019, 18(5):2722-2736.
[9] ZHENG H Z, ZHU Z C. Power and performance trade-offs in contemporary DRAM system designs for multicore processors\[J]. IEEE Transactions on Computers, 2010, 59(8):1033-1046.
[10] KULKARNI V A, UDUPI G R. Instruction level energy consumption estimation of embedded processor\[J]. European Journal of Engineering Research and Science, 2019, 4(2):40-44.

[1]	刘阳, 滕颖蕾, 牛涛, 郅佳琳. 基于深度强化学习的滤波器剪枝方案[J]. 北京邮电大学学报, 2023, 46(3): 31-36.
[2]	谭炜骞, 吴斌伟, 汪硕. 确定性网络跨域传输架构与DRL流量调度算法[J]. 北京邮电大学学报, 2023, 46(3): 37-42.
[3]	杨华, 耿烜, 孔宁. 一种采用dueling-DDQN算法的无线网络MAC协议[J]. 北京邮电大学学报, 2023, 46(3): 25-30.
[4]	孙国玮许方敏朱瑾瑜张恒升赵成林. 算力网络中的确定性调度与路由联合智能优化方案[J]. 北京邮电大学学报, 2023, 46(2): 9-14.
[5]	魏明亮耿绥燕赵雄文胡玮范静怡. 超密集网络中移动边缘计算的资源分配和任务卸载联合优化研究[J]. 北京邮电大学学报, 2023, 46(2): 50-56.
[6]	公雨魏翼飞. 一种集成学习辅助DDPG的资源优化算法[J]. 北京邮电大学学报, 2023, 46(2): 29-36.
[7]	付志远施丽琴叶迎晖卢光跃. 反向散射辅助的无线供能NOMA-MEC网络中公平性优化研究[J]. 北京邮电大学学报, 2023, 46(2): 71-77.
[8]	郭兴康孙君. 基于交替方向乘子法与深度强化学习算法的资源分配[J]. 北京邮电大学学报, 2022, 45(6): 126-130.
[9]	郭令奇褚智贤廖建新王敬宇陆璐. 意图驱动的自智网络资源按需服务[J]. 北京邮电大学学报, 2022, 45(6): 85-91.
[10]	李松王新荣王博文孙彦景陈瑞瑞. 基于随机网络演算的车联网边缘计算性能分析[J]. 北京邮电大学学报, 2022, 45(5): 16-22.
[11]	杜梅, 周军华, 李敦桥, 陈士钊, 魏翼飞. MEC计算卸载与资源分配联合智能优化方案[J]. 北京邮电大学学报, 2022, 45(2): 65-71.
[12]	张雨晴, 李云, 黄鸿锐, 庄宏成. 异构网络中任务卸载与资源分配联合优化算法[J]. 北京邮电大学学报, 2022, 45(2): 91-97.
[13]	郑远鹏, 张天魁, 朱光宇, 沈鸿. MEC系统中面向网络切片的3C联合资源分配算法[J]. 北京邮电大学学报, 2021, 44(5): 41-47.
[14]	康曼聪, 李曦, 纪红, 张鹤立. 基于任务间依赖关系的小小区协作卸载策略[J]. 北京邮电大学学报, 2021, 44(1): 72-78.
[15]	黄浩, 胡智群, 王鲁晗, 路兆铭, 温向明. 基于Sumtree DDPG的智能交通信号控制算法[J]. 北京邮电大学学报, 2021, 44(1): 97-103.

面向硬件感知的边缘计算卸载和资源分配

Hardware Model-Aware Joint Offloading and Resources Allocation Optimization

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价