Optimizing achievable throughput for cognitive radio network using swarm intelligence

Cognitive radio (CR) technology allows a secondary or cognitive user (CU) to opportunistically access frequency band of a primary user (PU) when it is not in use. However, a CU needs to perform spectrum sensing periodically to avoid causing harmful interference and thereby protecting the quality of service of PUs. A conventional frame structure for CR operation consists of sensing time slot and data transmission time slot which run consecutively. Basically, a longer sensing time produces a higher probability of detection and therefore, better PU protection. However, longer sensing time will reduce the amount of time for data transmission and hence affects the achievable throughput of a CU. In this paper, we study the fundamental tradeoff between sensing time and achievable throughput of the CUs. Based on energy detection scheme, we propose and investigate the feasibility of using Particle Swarm Optimization (PSO) in the design of sensing slot duration to maximize the achievable throughput for CUs for a given frame duration. The results show an encouraging 8.89% increase in throughput with respect to the probability of false alarm, a reduced sensing time by 80% while achieving 97.8% of normalized throughput for CU system with PSO.