A Spectrum Sensing Method and Performance Analysis Based on Stochastic Resonance under Low SNR

Rui GAO,Zan LI,Li-ping WU,Qun-wei LI,Pei-han QI

DOI: https://doi.org/10.3969/j.issn.0372-2112.2013.09.002

2013-01-01

Tien Tzu Hsueh Pao/Acta Electronica Sinica

Abstract:According to the spectrum sensing problem under low signal-noise ratio (SNR) and dynamic noise in cognitive radio(CR)networks ,this paper introduced nonlinear stochastic resonance(SR)of physics into spectrum sensing ,and proposed an en-ergy detection(ED)based on generalized stochastic resonance(GSR) .For the proposed algorithm ,SR noise was added to make non-linear system ,signal and noise matched ,which modifies the probability distribution of the detection statistics ,and confirms the exis-tence of the signal effectively .This paper drive the probability density unction (PDF)of matched noise ,and get some significant con-clusions about the performance ,effect of noise uncertainty and sensing time of the proposed algorithm .The simulation results vali-date the theory ,and show that the proposed algorithm can improve the performance of existing energy detection at least 3dB under low SNR .The proposed algorithm also has less sensing time ,low complexity and can effectively overcome the influence of the noise uncertainty .