Two-Phase Deep Reinforcement Learning of Dynamic Resource Allocation and Client Selection for Hierarchical Federated Learning

Xiaojing Chen, Zhenyuan Li, Wei Ni, Xin Wang, Shunqing Zhang, Shugong Xu, Qingqi Pei

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

9 Citations (Scopus)

Abstract

This paper presents a new two-phase Deep Deter-ministic Policy Gradient (DDPG) framework, referred to as 'TP-DDPG', to balance online the learning delay and model accuracy of a federated learning (FL) process in an energy harvesting hierarchical FL (HFL) system. The key idea is to design a DDPG-based approach to learn the selection of participating clients, the CPU configuration, and the transmission powers of the clients, while the other decisions are efficiently optimized by a new straggler-aware client association and bandwidth allocation algorithm. The algorithm evaluates the reward of the DDPG, and substantially improves its convergence rate and stability. Experimental results demonstrate that the proposed TP-DDPG can substantially reduce the training time while achieving a higher test accuracy over the existing benchmarks.

Original languageEnglish
Title of host publication2022 IEEE/CIC International Conference on Communications in China, ICCC 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages518-523
Number of pages6
ISBN (Electronic)9781665484800
DOIs
Publication statusPublished - 2022
Externally publishedYes
Event2022 IEEE/CIC International Conference on Communications in China, ICCC 2022 - Sanshui, Foshan, China
Duration: 11 Aug 202213 Aug 2022

Publication series

Name2022 IEEE/CIC International Conference on Communications in China, ICCC 2022

Conference

Conference2022 IEEE/CIC International Conference on Communications in China, ICCC 2022
Country/TerritoryChina
CitySanshui, Foshan
Period11/08/2213/08/22

Keywords

  • client scheduling
  • DDPG
  • Hierarchical federated learning
  • resource allocation

Fingerprint

Dive into the research topics of 'Two-Phase Deep Reinforcement Learning of Dynamic Resource Allocation and Client Selection for Hierarchical Federated Learning'. Together they form a unique fingerprint.

Cite this