Optimal user selection for decision making in cooperative sensing

In enabling spectrum-aware communication protocol of cognitive radio (CR), spectrum sensing becomes the key technology for unlicensed or cognitive users (CUs) to exploit the unused spectrum of primary users (PUs) via opportunistic spectrum access (OSA). A longer sensing time will result in a higher detection probability and hence, a better protection for the PU. However, in a fixed frame size, this leads to reduced data transmission time and hence, decreases the achievable throughput of CUs. In this paper, particle swarm optimization (PSO) formulation is proposed to address this sensing-throughput tradeoff under distributed cooperative sensing scenario using hard combining fusion schemes to determine the presence of PUs. The objective is to optimize the sensing time and achieve maximized throughput for CUs while sufficiently protecting the PUs. The findings show that jointly optimizing the sensing time and the parameter k of k-out-of-N rule outperforms other optimized fusion schemes of OR, AND and Half-Voting significantly in terms of achievable throughput and reliability.