An improved fair channel hopping protocol for dynamic environments in cognitive radio networks

Abstract

Rendezvous is a fundamental challenge in cognitive radio networks where users can find each other on a specific channel and hence establish a communication link. Most previous works are based on the strong assumption that users are able to find a set of available channels after the spectrum sensing stage and the status of these channels are stable all the time, which, however, may be unrealistic in some scenarios. As a solution, we design a fair channel hopping protocol with dynamic channel state, by adopting the concepts of Markov process, Jenkins Hash and Josephus recursive. Two protocols (FCH_S, FCH_A) are proposed for synchronous clock and asynchronous clock network model, respectively. The channel activity model is built with the aid of Markov process. By taking advantage of Jenkins Hash and Josephus recursive, the fairness of protocol is guaranteed. We assume that (1) a secondary user, SU$_A$, rendezvous with SU$_B$; (2) corresponding channels available probability are $p_a$ and $p_b$. According to these assumptions, we can prove that expect rendezvous time for FCH_S and FCH_A are $\dfrac{1}{p_{a}p_{b}}$ and $\dfrac{1}{p_{a}}+\dfrac{1}{p_{b}}-1$. Simulation results demonstrate that FCH_S and FCH_A can achieve better performance in contrast to the exiting channel hopping protocols (e.g. H.Tan and HHCH).