算力网络中面向计算重用的任务调度优化  

Optimization of task scheduling for computing reuse in computing power network

作　　者：马云霄 吴忠辉 徐祖云 衷璐洁[2] 许长桥[1] MA Yunxiao;WU Zhonghui;XU Zuyun;ZHONG Lujie;XU Changqiao(State Key Laboratory of Networking and Switching Technology,Beijing University of Posts and Telecommunications,Beijing 100876,China;Information Engineering College,Capital Normal University,Beijing 100048,China)

机构地区：[1]北京邮电大学网络与交换技术国家重点实验室,北京100876 [2]首都师范大学信息工程学院,北京100048

出　　处：《通信学报》2023年第11期129-142,共14页Journal on Communications

基　　金：国家自然科学基金资助项目(No.62225105);北京邮电大学博士生创新基金资助项目(No.CX2021108)。

摘　　要：为应对未来算力需求爆炸性增长所带来的挑战,将计算重用技术引入算力网络中,通过重用计算任务结果,来缩短服务时延并减少计算资源消耗。在此基础上,提出基于服务联盟的上下文感知在线学习算法。首先,设计重用指数来减少额外查找时延;然后,基于服务联盟机制进行在线学习,根据上下文信息及历史经验做出计算任务调度决策。仿真实验结果表明,所提算法在服务时延、计算资源消耗等方面均优于基准算法。To cope with the challenges posed by the future explosive growth in computing power demand,computing reuse technology was introduced into the computing power network to reduce service latency and computational resource consumption by reusing the results of computational tasks.Based on this,a service federation-based context-aware online learning algorithm was proposed.First,the reuse index was designed to reduce the extra lookup latency.Then,online learning was performed based on the service federation mechanism to make computational task scheduling decisions according to contextual information and historical experience.The experimental results show that the proposed algorithm outperforms the baseline algorithms in terms of service latency and computational resource consumption.

关 键 词：算力网络 计算重用 任务调度 在线学习 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]