Adaptive Double Threshold Based Spectrum Sensing to Overcome Sensing Failure in Presence of Noise Uncertainty

A major breakthrough has been observed in wireless communication technology over the past few years. The advent of new wireless applications has choked the available bandwidth and pleads for a more efficient strategy for its allocation to users. Among the various proposed methods of spectrum de-congestion, Cognitive radio seems to be a promising solution to spectrum scarcity. Spectrum sensing is the first and most crucial step in establishing cognitive radio system. However, the available spectrum sensing techniques are severely limited by noise power fluctuations, fading, multipath propagation and low signal-to-noise ratio. These factors affect the sensing functionality in terms of increased missed detection and false alarm rates, reduced probability of detection and large number of samples. This paper proposes an adaptive threshold method to overcome sensing failure at very low SNR with uncertain noise power using a check parameter and double threshold concept. The double threshold concept tapers the width of the uncertainty zone and makes the detection process robust. Simulation results validate the new findings and improve the detection probability by 19.95% at a low SNR of −12 dB.