Channel Estimation Based on Compressed Sensing in UWB Systems

WANG Kai,LIU Yu-fin,MANG Man-yu
DOI: https://doi.org/10.3969/j.issn.1006-9348.2011.06.033
2011-01-01
Abstract:Extremely high-speed analog-to-digital converters(ADCs) are required to sample the extremely narrow pulses in the estimation of UWB channels.It is difficult to meet the requirement of sampling rate under the current level of the hardware production.In order to reduce the excessively high requirement of the sampling rate in UWB systems,a novel method for channel estimation was brought forward in this paper based on compressed sensing.Considering the time-domain sparsity of UWB channel,CS reconstruction model was transformed from channel estimation model by using the Quasi-Toeplitz matrices.The channel tap can be estimated from relatively few observations collected at the receiver.Simulation results show that the Nyquist sampling rate of 1/5-1/10 will be sufficient to channel estimation in UWB systems.Comparing with traditional method,the estimated MSE of the novel method decreased by 6-8dB under the similar SNR.
What problem does this paper attempt to address?