Cross-Layer Analysis and Optimization on Access Delay in Channel-Hopping Based Distributed Cognitive Radio Networks

In channel-hopping (CH) based distributed cognitive radio networks (CRNs), the time duration that secondary users (SUs) spend for establishing communication links is called access delay. To evaluate access delay, we propose an access delay model by jointly considering imperfect spectrum sensing and multi-channel multi-SU transmission, from the cross-layer perspective. The model considers two typical scenarios: a) the first scenario assumes that SUs do not use contention scheme (CS) which indicates that time slot is relatively shorter to just allow a transmission, and b) the second scenario assumes that SUs employ CS (i.e., modified Distributed Coordination Function #(DCF) based Carrier Sense multiple Access/Collision Avoidance (CSMA/CA) in this paper) which indicates that time slot is long enough to regulate multiple transmissions. We then propose a bio-inspired algorithm for the first scenario and a self-adaptive step-length algorithm for the second scenario to search for the optimal values of spectrum sensing parameters. Theoretical analysis and simulation results validate the proposed access delay model and show that the proposed algorithms can reduce most redundant computation. They also show that the optimization of cross-layer parameters can significantly decrease SUs’ access delay. Moreover, we conduct a cost-benefit analysis to evaluate performance of the two scenarios.