摘要
随着人工智能应用场景的集中式爆发,移动应用对数据通信和计算的需求日益增长,位于远端的传统云计算处理方法难以满足快速响应的要求.因此,整合利用海量的用户侧终端设备算力(包括计算、存储、通信等)的端侧算力网络,通过分布式协作合理地利用终端算力完成计算任务成为一种新的处理方法.鉴于单台终端设备的资源受限,高企的通信开销限制任务协同效果,导致终端很难高效协同完成高度复杂的计算任务.本文提出利用点对点(Device-to-Device,D2D)通信辅助终端节点协同计算,并设计了基于有向图卷积网络(Directed Graph Convolutional Network,DGCN)的协作拓扑和资源分配决策算法(Multi-Agent Soft Actor-Critic,MA-SAC),将有向无环图(Directed Acyclic Graph,DAG)任务中包含的子任务部署到多个终端进行协同计算,满足DAG子任务部署在多个不同节点间的跨节点传输需求,降低子任务间数据传输在基站侧的网络通信开销.仿真结果显示,所提算法能够在保证业务时延要求下,降低38.2%的网络通信开销,有效提升31.9%的端侧资源利用率.
Driven by the concentrated surge of AI application scenarios,the increasing requirements on data commu⁃nication and computation in mobile applications is growing,the traditional cloud computing which relies on remote process⁃ing,often fails to meet low-latency requirements.Therefore,a new paradigm has emerged:terminal-side computing power that aggregate the vast terminal devices(including computing,storage,communication,etc)through distributed collabora⁃tion to efficiently execute computational tasks.However,constrained by the limited resource of standalone device and pro⁃hibitive communication overhead that impairs task coordination,such terminals still face significant challenges in achieving efficient collaboration for highly complex computing tasks.This paper presents device-to-device(D2D)communication as⁃sisted terminal devices collaborative computing,and a multi-agent soft actor-critic(MA-SAC)based on directed graph con⁃volutional network(DGCN)is designed to solve this problem.The subtasks included in directed acyclic graph(DAG)tasks were deployed to multiple terminals for collaborative computing,it is introduced to cater to the exigencies of task transmis⁃sion between disparate nodes within the DAG,and reduces the communication overhead when data transmission in the net⁃work.Through the simulations,the efficacy of the proposed scheme is demonstrated.The proposed scheme reduces network communication overhead by 38.2%and effectively improve resource utilization by 31.9%.
作者
顾健华
冯建华
许辉阳
刘佟佟
周婷
GU Jian-hua;FENG Jian-hua;XU Hui-yang;LIU Tong-tong;ZHOU Ting(Department of Computer Science and Technology,Tsinghua University,Beijing 100084,China;China Mobile Group Device Co.,Ltd.,Beijing 100053,China)
出处
《电子学报》
北大核心
2025年第6期1771-1783,共13页
Acta Electronica Sinica
关键词
端侧算力
终端协同
多跳D2D
端算力分配
有向图卷积网络
terminal-side computing power
terminal collaboration
multi hop D2D
terminal-side computing power allocation
directed graph convolutional network
