Summary
International Conference on Emerging Technologies for Communications
2020
Session Number:K1
Session:
Number:K1-4
Radio Resource Management Aided Multi-Armed Bandits for Disaster Surveillance System
Amr Amrallah, Ehab Mahmoud Mohamed, Gia Khanh Tran, Kei Sakaguchi,
pp.-
Publication Date:2020/12/2
Online ISSN:2188-5079
PDF download
Summary:
In this paper, the radio resource allocation for cognitive radio network (CRN) through dynamic spectrum access (DSA) [1] system is investigated, for which we proposed two Multi-Armed bandit (MAB) [2] algorithms to tackle this problem, i.e., the \epsilon-greedy algorithm and the Upper Confidence Bond (UBC) algorithm. The idea behind choosing MAB algorithms that they are able to solve NP hard problems with different conflicting parameters need to be jointly optimized. The simulation shows \epsilon-greedy algorithm has a faster conversion rate while UCB algorithm can achieve a higher total system-rate performance.