Abstract
Ultra wide band (UWB) signals are very finely spaced in time domain. Due to very narrow pulse width, the UWB receivers are usually implemented through correlators. The complexity and cost of UWB receiver is determined by the number of correlators used. As UWB channels have very long impulse responses as compared to the pulse width, channel shortening can help in reducing the number of correlators and thus enables simple and cost effective implementation of the UWB receiver. In this paper, we present a new channel shortening algorithm which can shorten the dense multipath channels, such as based on IEEE CM 4 model, to just one significant tap. We compare it to maximum shortening signal to noise ratio (MSSNR) and target impulse response (TIR) based minimum mean squared error (MMSE) algorithms and analyze their complexity, consistency and performance on the basis of number of equalizer taps, captured energy, bit error rate (BER) and signal to interference and noise ratio (SINR) improvement in a multiuser environment. We show that the proposed algorithm is more consistent in shortening the channel with less number of equalizer taps and its performance is well above MSSNR and MMSE algorithms. Channel shortening to an extent of single tap using proposed algorithm eliminates the necessity of instantaneous template waveform evaluation at the receiver for efficient signal detection. This feature further simplifies the receiver implementation along with single correlator design