Modified MSSNR Based Channel Shortening for UWB Receiver Design Simplification

Ultra wide band (UWB) propagation channels are highly rich in multipaths. A major problem encountered in UWB systems is to capture enough multipaths to maintain a sufficient signal to noise ratio (SNR) for further signal processing. This leads to a RAKE receiver with large number of fingers. Channel shortening can help in simplifying the RAKE receiver architecture by reducing the channel taps. The recent developments in channel shortening mainly deal with time varying wired line scenarios like multicarrier modulation (MCM) systems where it is used to suppress few channel taps outside the cyclic prefix (CP) length. In UWB systems, the problem of channel shortening appears in its extreme form where a large number of channel taps are needed to be suppressed and most of the channel energy should be compressed within just few taps. Hence, the underlying theory and assumptions of most of the existing algorithms are invalidated when applied to UWB. The algorithm which addresses the channel channel shortening in the most primitive way and can be applied to any system in general is maximum shortening signal to noise ratio (MSSNR) algorithm. In this paper, we modify the MSSNR algorithm to exploit the UWB characteristics and make it capable to handle the extreme nature of channel shortening needed in UWB systems. The proposed algorithm does not need any training or channel estimation and outperforms MSSNR algorithm in terms of different comparative parameters.