An SDR implementation of reliable spectrum sensing using fuzzy logic and genetic algorithm

Abstract

This paper presents the design of a reliable spectrum sensing method using fuzzy logic and genetic algorithm. Fuzzy logic acts as a decision-making system. It is used to select an optimal sensing method based on the current wireless channel state, the required accuracy of sensing, and computational complexity. Three sensing algorithms named as energy detection, feature detection and cyclo-stationary detection are used. The fuzzy decision-making system analyses the present state of SNR, availability information about the primary user, application demand on delay and accuracy. It selects optimal spectrum sensing techniques based on that analysis result. A genetic algorithm is used to choose optimal parameter selection for the selected optimal algorithm to improve the reliability further. The entire scheme is implemented on the USRP RIO SDR hardware. Performance evaluation on experiments shows that the proposed system outperforms with the probability of detection 99.7%.