Development Of Improved Spectrum Sensing Techniques for Cognitive Radio Networks

Abstract

newline Rapid growth of wireless communication traffic and limited spectrum availability to newlinesupport the wireless applications and services, has led to the development of cognitive newlineradio as promising solution to address the problem of spectrum scarcity. Spectrum newlinesensing is one of the fundamental functions towards implementation of the cognitive newlineradio concept in wireless communication. The two major challenges of the spectrum newlinesensing are reliable detection performance at low SNR and optimization of detection newlinesensitivity vis-a-vis complexity of the sensing techniques. newlineAmong the several techniques of spectrum sensing, Energy Detection based method newlineworks without prior information about the primary user signal. This yields sensing newlineperformance in the most efficient manner in terms of low sensing period, low complexity newlineand low computational cost. However, noise uncertainty and low SNR can deteriorate its newlineperformance. Most of the conventional spectrum sensing techniques are capable to give newlinereliable performance but they are complex, take considerable time in decision making, newlineincur high computational cost and need prior information about the primary user signal. newlineIn this research work, the available spectrum sensing techniques are studied with a view newlineto critically identify their problems and limitations and propose the possible solutions to newlinethose. In this thesis, a new fuzzy logic based two stage spectrum sensing technique is newlinedeveloped. An algorithm for this technique is designed based on the rigorous analysis newlineresulting in considerably good and robust performance at low SNR even in the presence newlineof noise uncertainty.