Applications Of Compressive Sensing Over Wireless Fading Channels

Abstract

Abstract Wireless fading channels exist mixed noise, the most common noises are gaussian noise and impulse noise, to ensure the quality of the received signal, removing the noise in the channel is very important and necessary. WSDN (wavelet soft-threshold de-noising) can suppress low-intensity gaussian noise well; unfortunately, the denoising effect in removing impulse noise using WSDN is not obvious, especially when noise intensity is relatively high. BPDN (Basis Pursuit Denoising) which is a reconstruction algorithm of compressive sensing can control high-intensity gaussian noise in the channels well preceding WSDN, but also the effect of BPDN denoising the common impulse noise is not obvious, to denoise impulse noise, we adopt IBPDN (Improved Basis Pursuit Denoising) that changes L2 form to noise to L1 form, which was proposed in 2006 by Guosheng Bing. The experimental data show that IBPDN has good anti-noise ability to both gaussian noise and impulse noise; furthermore, more satisfactory results are obtained using IBPDN to mixed noise than those with BPDN.