Learning-based p-persistent CSMA for secondary users of cognitive radio networks

Abstract

In this paper, we consider learning-based p-persistent carrier sense multiple access (CSMA) for cognitive radio secondary users (SUs) when the primary user (PU) operates with the conventional CSMA/CA scheme. The learning algorithms are applied to tune the value of p to balance the channel idle time and collision costs due to both the PU and the SUs sharing the same wireless channel. Our proposal comprises two steps: first, the Q-learning scheme is used to adjust the transmission probability p when the PU has imperfect sensing. Then, for the multiple SUs’ case, a mixed strategy access policy is executed to solve the fairness problem in media access control (MAC). Our proposed scheme can maximise the SUs’ throughput and ensure protection of the performance of the PU, as well.