Novel multiuser ultra-wideband receiver based on adaptive modelling of generalised normal-Laplace distribution

This study takes a new receiving approach to detect an ultra-wideband (UWB) signal in the presence of multiple access interference (MAI). A generalised normal-Laplace (GNL) distribution is exploited for designing a novel multiuser UWB receiver. To more accurately comply with real statistical behaviours of the MAI-plus-noise in time-hopping multiple access UWB systems, a modified parameter estimator, which is an adaptive method of moments estimation (aMME), is proposed. In the existing studies, the analysis about GNL model shows appropriate results for UWB simulations, while it requires heavy complexity to obtain the precise distribution because of no existence of a closed-form probability density function (PDF) expression. To practically evaluate the GNL PDF, the fast Fourier transform, which can significantly reduce the computational complexity is considered. The GNL using the aMME outperforms the conventional matched filter UWB receiver, the soft-limiting receiver, the Gaussian-Laplace mixture receiver, the p-order metric receiver and the previous GNL receivers, especially in high SNR ranges. Furthermore, the presented GNL receiving method is applied to each finger of the conventional Rake receiver for signal detection in IEEE UWB multipath channels. The proposed Rake receiver based on the GNL distribution performs better than the conventional Rake receiver in multipath fading channels.