Computing Task Orchestration in Communication and Computing Integrated Mobile Network.

Ziqi Chen, Yi Ren,Qi Sun ,Nan Li, Yantao Han,Yuhong Huang, Hongtao Li,Chih-Lin I

GLOBECOM (Workshops)（2023）

引用 0|浏览0

暂无评分

摘要

Communication and Computing Integrated Mobile Network has the potential to enable next generation communication system services through providing communication and computing services in a holistic way. Due to the computing and communication resource limits, it is challenging to provide an efficient computing task orchestration scheme. In this paper, we present the concept of computing task curve, which describes the computing power needs, transmission speed requirements and end-to-end delay relationship of computing tasks. Based on computing task curve, a deep reinforcement learning based task orchestration algorithm is presented which jointly optimise PRB usage and computing cost through intelligently allocating tasks to computing nodes in the network. Finally, the algorithm is verified in NS-3 simulator, showing better overall performance than baseline methods.

查看译文

关键词

Mobile Network,Computation Tasks,Deep Learning,Computational Cost,Power Calculation,Computational Resources,Joint Optimization,Deep Reinforcement Learning,Transmission Speed,Computing Nodes,Wireless,Value Function,Deep Neural Network,Data Center,State Space,Cloud Computing,Types Of Tasks,Feed-forward Network,Task Execution,Transmission Delay,Deep Reinforcement Learning Agent,Mobile Edge Computing,Offloading Decision,Greedy Policy,Baseline Solution,Task Offloading,Deep Reinforcement Learning Algorithm,Head Acceleration,Brute Force,State Action Space

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要