Soft decision Cooperative Spectrum Sensing based upon noise uncertainty estimation

Spectrum Sensing (SS) constitutes the most critical task in Cognitive Radio ( CR) systems for Primary User (PU) detection. Cooperative Spectrum Sensing (CSS) is introduced to enhance the detection reliability of the PU in fading environments. In this paper, we propose a soft decision based CSS algorithm using energy detection by taking into account the noise uncertainty effect. In the proposed algorithm, two threshold levels are utilized based on predicting the current PU activity, which can be successfully expected using a simple successive averaging process with time. The two threshold levels are evaluated based on estimating the noise uncertainty factor. In addition, they are toggled in a dynamic manner to compensate the noise uncertainty effect and to increase the probability of detection and decrease the probability of false alarm. Theoretical analysis is performed on the proposed algorithm to evaluate its enhanced false alarm and detection probabilities over the conventional soft decision CSS using different combining schemes. In addition, simulation results show the high efficiency of the proposed scheme compared to the conventional soft decision CSS, with high computational complexity enhancements.