Random filter based ultra-wideband channel estimation

Ultra-wideband (UWB) channel estimation based on compressive sensing (CS) is developed. A random filter is exploited in the system to produce the measurement matrix. A new filter which is “all pass” in frequency domain and both real in time and frequency domain is established. Then subsampling and reconstruction algorithm are utilized for channel estimation. The Cramer-Rao bound of the sparse channel estimation is obtained. Numerical results indicate that the estimation algorithm with quite a low sampling rate can get a perfect performance. The results also reveal that the algorithm´s Mean Square Error (MSE) approaches the CRB at high Signal-Noise Ratio (SNR).