Throughput analysis of cognitive radio networks via stochastic network calculus

Cognitive radio is a promising technique in improving the utilization of wireless spectrum by spectrum sensing. However, conceivable spectrum sensing errors may deteriorate the quality of service provided for users. In this paper, we propose a framework to analyze the per-user throughput of both primary users and secondary users. Based on this framework, we further estimate how sensing errors influence the throughput in cognitive radio networks. To this end, stochastic network calculus is employed as a basic tool, which is a powerful theory for quality of service analysis of computer networks. Impairment processes of service are introduced to construct the service models for both primary users and secondary users. Then per-user throughput bounds of both primary users and secondary users are obtained after considering arrival process of primary users. Numerical result shows that sensing errors play a decisive role on the upper bound of primary users´ throughput, while have less impact on the lower bound of secondary users´ throughput.