Learning based spectrum utilization in underlay cognitive radio network

Abstract

Spectrum access strategy plays a critical role in multichannel cognitive radio networks (CRNs). The Cognitive radio is a key solution for radio spectrum under-utilization problem. It aims to increase the QoS of the system. In underlay scheme the primary user (PU) and secondary users (SU) can access the spectrum and communicate concurrently. However, SU’s established the threshold for not initiating interference to PU called interference temperature. In this way the unlicensed user accesses the licensed spectrum to efficiently utilizes the resources without interfering the PU. The SU has to sense continuously the spectrum hole, which can result in delay and energy consumption. In this proposal, we propose an efficient spectrum utilization scheme based on Q-learning to maximize throughput and minimize energy and delay.