Throughput analysis for cognitive radio networks with multiple primary users and imperfect spectrum sensing

In cognitive radio networks, the licensed frequency bands of the primary users (PUs) are available to the secondary user (SU) provided that they do not cause significant interference to the PUs. In this paper, we analyzed the normalized throughput of the SU with multiple PUs coexisting under any Frequency Division Multiple Access (FDMA) communication protocol. We consider a cognitive radio transmission where the frame structure consists of sensing and data transmission slots. In order to achieve the maximum normalized throughput of the SU and control the interference level to the legal PUs, the optimal frame length of the SU is found via simulation. In this context, a new analytical formula has been expressed for the achievable normalized throughput of SU with multiple PUs under prefect and imperfect spectrum sensing scenarios. Moreover, the impact of imperfect sensing, variable frame length of SU and the variable PU traffic loads, on the normalized throughput has been critically investigated. It has been shown that the analytical and simulation results are in perfect agreement. Our analytical results are much useful to determine how to select the frame duration length subject to the parameters of cognitive radio network such as network traffic load, achievable sensing accuracy and number of coexisting PUs. Index Terms Cognitive radio; normalized throughput; imperfect spectrum sensing; probability of collision; traffic August 23, 2012 DRAFT