Analysis of Throughput with Reinforcement Learning of TD-CSMA system in Cognitive Radio Networks
Nisha Kiran, Prabhat Patel · 2012
Cognitive radio technology is widely accepted as an efficient approach to solve the problem of scarcity of the wireless spectrum resulting due to the rapid growth in the ubiquitous wireless applications. Several cognitive medium access control protocols have been proposed for the secondary users (non-licensed users) to take advantage of the vacant channels whenever they are not occupied by the primary users (licensed users). This paper analyses a cognitive radio scenario based on non-persistent carrier sense multiple access (CSMA) protocol for secondary user and time division multiple access (TDMA) for primary users in multi-channel TD-CSMA network. Performance of secondary users is evaluated for a various proportions of non-persistent CSMA and TDMA traffic levels. Simulations results show that the throughput performance of CSMA users improves when multichannel are used. Further, reinforcement learning is applied in conjunction with non-persistent CSMA which also enhances the throughput performance on same proportions.