kumoh national institute of technology
Networked Systems Lab.

Sanjay Bharwdaj, Jae-Min Lee and Dong-Seong Kim ,"Dynamic Multichannel Access for URLLC in Industrial Wireless Networks," 2020 Korean Institute of Communication and Sciences (KICS) Summer Conference, August 12-14, 2020, Yong Pyong Resort, Pyeongchang, Gangwon Province, Korea (A).
By :
Date : 2020-06-07
Views : 145

In industrial wireless networks, real-time transmission increases probability of collision and congestion thus degrading their performance in terms of reliability and latency. In this paper, dynamic multichannel access in industrial wireless networks for ultra reliable low latency communication (URLLC) is considered. For which, partially observable Markov decision process (POMDP) based formulation is carried out for unknown system dynamics of the industrial wireless networks. For the challenges of unknown system dynamics, double deep Q-learning

(DDQL) is proposed which deals with two key requirements: reliability and latency Performance evaluation of DDQL is compared with Myopic and Whittle Index based heuristic policy. Simulation results clearly indicate that proposed DDQL finds better and optimal policy that maximizes the rewards in terms of reliability and latency.