An Energy Efficient Adaptive Switching Spectrum Sensing (ASSS) Technique With Optimal PU Node Detection for CR‐WSN

ABSTRACT Cognitive Radio—Wireless Sensor Network (CR‐WSN) plays a vital role in spectrum utilization by allowing secondary users (SUs) to utilize the under‐used licensed bands in an opportunistic manner. However, spectrum sensing accuracy is often affected by various channel perturbations such as multipath fading channel, noise, and interference. In this paper, we propose an Adaptive Spectrum Sensing Switching (ASSS) technique where the SU Cluster Head (SU‐CH) in each cluster adaptively switches between Energy Detection (ED) and Matched Filter (MF) sensing methods to improve the detection accuracy of SU nodes. The proposed ASSS method uses Received Signal Strength Indicator (RSSI) as a metric where the SU nodes report their sensing results to their cluster heads, and then the SU‐CHs take an appropriate decision based on the channel conditions. During poor channel conditions, the SU‐CH nodes employ MF‐based detection for its robustness against fading and noise, leading to optimal PU node detection. On the other hand, during good channel conditions, ED is employed, resulting in reduced energy consumption and computational complexity. The simulation results prove that at low SNR, the proposed ASSS method without Bayesian threshold significantly improves detection probability on average by about 78% for ED and inferior by slightly about 14.28% for MF and 40% for the proposed ASSS method with Bayesian threshold, thereby overcoming the trade‐off in energy consumption by achieving an energy efficiency of 60.79% lesser than ED, 35.95%, and 18.87% more energy efficient on average compared to MF and proposed ASSS method with Bayesian threshold based approaches.