Nature Inspired Cluster based Optimal Spectrum Sensing in Cognitive Radio Networks

Abstract

The demand for spectrum has increased as the number of wireless devices has grown rapidly. However, despite the huge demand, the available spectrum is underutilized. To resolve issues regarding spectrum requirement and spectrum utilization Cognitive radio (CR) network has been introduced. The ability to share the spectrum among licensed and unlicensed users in the cognitive network gains more attention among researchers to reduce the spectrum scarcity. Specifically, cluster-based cooperative spectrum with optimal sensing introduces tremendous impact with its reliable sensing characteristics compared to cooperative spectrum sensing. Thus, in this research work, cluster-based optimal spectrum sensing is introduced to improve spectrum utilization and minimize sensing time. To achieve this,a simple but efficient K-nearest neighbor algorithm is employed for cluster formation and a bio-inspired firefly optimization algorithm is applied for optimal spectrum sensing. The optimal spectrum bands information is transferred to the fusion center and the final decision is obtained as a hard decision since it provides better sensing performance than soft decision detection.  To validate the performance of the proposed approach, conventional particle swarm optimization, genetic algorithm, and low energy adaptive clustering hierarchy algorithms are compared in terms of false alarm, detection probability, throughput, error rate, energy consumption, spectrum utilization,sensing time, and accuracy.