A feature detector based on compressed sensing and wavelet transform for wideband cognitive radio

Detection of wideband communication signals is critical for cognitive radio (CR) as it enables secondary users to dynamically access the unoccupied bands. However, accurate and fast spectrum sensing is still a challenge in low signal to noise ratio (SNR) environment. To encounter this problem, a feature detector based on compressed sensing (CS) and wavelet transform (WT) (CS-WT feature detector) is proposed. Feature detector is chosen for its accuracy under low SNR, and CS is introduced to alleviate the sampling bottleneck of wideband sensing. Moreover, noise caused by the CS process is analyzed, and a traditional noise reduction method-two dimensional wavelet transform is utilized to cope with it by treating the spectral correlation function (SCF) as a grey image. It is verified by simulation that WT can effectively reduce the noise introduced by CS, and the proposed detector can achieve 90% detection probability under −10dB, making cyclostationary detection based on CS applicable.