Ultra-wideband (UWB) wireless communication systems have seen much research interest and industrial activity. One class of UWB systems uses time-hopped (TH) impulse-radio signalling, in which ultra-short pulses are transmitted at baseband. Such systems typically use a repetition code, by which a number of pulses are used to transmit one source symbol, and are characterized by a low duty cycle, with transmitted signal energy occupying only a small fraction of the available transmission time.;It is established that the multiple-user interference (MUI) in TH-UWB systems is impulse-like and poorly approximated by a Gaussian distribution. Therefore, conventional matched filter receiver designs, optimal for Gaussian noise, are not suitable for MUI-dominant environments. Various receiver structures have been proposed that apply a transformation to the partial decision statistics (PDSs) formed by correlating each received pulse with a template at the receiver. These structures, along with corresponding MUI models, are presented and compared in this thesis, including Rake implementations exploiting the multipath propagation abundant in UWB channels.;In addition, novel Rake receiver structures are provided that sense the presence of MUI in each PDS, and alter PDS processing accordingly. Due to the repetition code and multipath propagation, a Rake receiver has many PDSs available for use in deciding on a transmitted symbol. It is demonstrated that, when MUI is significant, performance is improved by use of a MUI magnitude criterion in selection of PDSs for inclusion in the decision statistic. Receiver structures are also proposed and evaluated which use per-PDS MUI-sensing statistics to weight each PDS in the decision statistic, or to alter a nonlinear transformation applied to each PDS.;The proposed receivers form statistics that contain MUI contributions correlated with the MUI in the corresponding PDS, i.e., the interference which corrupts the desired-user signal. Structures are considered both which operate on each finger of every frame, and which operate on frames after finger combining. The interference-sensing statistics do not rely on assumptions about the distribution of the aggregate interference process, or knowledge of interferer timing or hopping codes, but utilize the low duty-cycle property of the TH-UWB system. |