Performance Analysis of Prediction Based Sensing in Energy Harvesting Cooperative CRN

Abstract

Energy efficiency and spectral utilization are two key factors that are driving the wireless communication strategies for future. Energy harvesting enabled cognitive radio network (CRN) promises to solve the pertinent issues of spectral and energy efficient communications. In this paper, the performance of an energy harvesting cooperative CRN under prediction based sensing is evaluated for various fusion rules. Prediction based sensing improves the spectrum utilization and protects the quality of service (QoS) of primary user (PU). On the other hand energy harvesting enabled Cognitive Radio (CR) is capable to harvest energy from either of radio frequency (RF) sources and non-RF source based on a combined decision taken at the fusion center in a cooperative network. Based on the combined decision declared by the FC, the CRN either harvests energy from RF or non-RF sources or starts transmission. Impact of prediction on sensing, number of frames, number of cooperative CR nodes, collision probability, splitting parameter on energy harvesting and throughput performance are investigated and analyzed. Analytical expressions for detection probability normalize throughput and harvested energy are derived for various fusion rules under an energy harvesting CRN. A detailed comparison of the results are also presented in this paper.