A Noise-Resistant UWB Channel Estimation Scheme Based on Compressed Sensing

摘要

Compressed sensing (CS) is a promising candidate to lower the sampling rate of ultra-wideband (UWB) signals and CS-based UWB channel estimation is an emerging topic. Due to the noise folding (NF) effect of CS, it is difficult to achieve high channel estimation accuracy. In this paper, a new CS-based channel estimation scheme is proposed to combat the NF effect. The key idea of the proposed scheme is to modify the random measurement matrix based on the statistical prior knowledge of the UWB signals such that most of the noise components can be removed. The orthogonal matching pursuit (OMP) algorithm is employed to recover the channel waveform which is then used by a correlator-based receiver to detect the data symbols. Simulation results confirm that the proposed scheme can be used to combat the NF effect of CS effectively and thus gain better channel estimation accuracy and BER performance.