A novel multi-user detector in UWB communication systems

A novel multi-user detector in ultra-wideband multi-access (UWB-MA) communication systems is proposed in this paper. It is divided into three parts in the order of data flow. Data pretreatment part executes Gaussian type noise suppression based on convolution type wavelet packet transformation. Then wideband space-time spectrum estimator based on improved spectral correlation direction of arrival (DOA) matrix with smoothness algorithm is applied to recognize the number of users and realize distance-angle super-resolution of signal sources. With the information of users an improved robust-M detector based on wavelet packet transformation complete the detection, which has good performance in the presence of multiple-access interference (MAI). The results of analysis and simulation both show this kind of signal detection system has good detecting ability, fine noise immunity and especially be suited to apply in dense multi-path environments such as indoor UWB communication. Comparison between the proposed detector and the conventional is also given.