Detection, channel estimation and interference suppression for DS-UWB

• Main Author: Li, Jingjun
• Language: English
• Published: University of British Columbia 2011
• Online Access: http://hdl.handle.net/2429/31889

Ultra-wideband wireless transmission has attracted considerable attention as future technology for Wireless Personal Area Networks (WPAN). One of the proposals for standardization of a UWB physical layser is based on direct-sequence UWB (DS-UWB). The DS-UWB proposal envisages two modulation formats: binary phase shift keying (BPSK) and 4-ary bi-orthogonal shift keying (4BOK). Any DS-UWB compliant device is required to be able to transmit and receive BPSK signals. Due to the large transmission bandwidth, the UWB channel is characterized by a long delay spread which results in severe inter-symbol interference for UWB communications. We therefore study equalization for DS-UWB systems employing BPSK modulation. Since the equalization schemes use UWB channel information, which is not always available in practical applications, we also study channel estimations for DS-UWB systems. Furthermore, DS-UWB systems can be interfered with by other unknown DSUWB systems as well as other relatively narrow-band wireless communication systems that operate in the same band as DS-UWB systems and have a power spectral density (PSD) much stronger than that of DS-UWB systems. We therefore focus our attention on interference suppression for DS-UWB systems. In the first part of this work, we consider equalization for DS-UWB with BPSK modulation. We analyze the performance limits applicable to any equalizer, taking into account practical constraints such as receiver filtering and sampling. Our results show that chip-rate sampling is sufficient for close-to-optimum performance. For analysis of suboptimum equalizer strategies, we derive linear equalization schemes for DS-UWB systems. Moreover, we devise equalization schemes with widely linear processing, which improve performance without increasing equalizer complexity. Simulation and numerical results show that low-complexity (widely) linear and nonlinear equalizers perform close to the pertinent theoretical limit. Finally, to ease the complexity of practical hardware implementation, we derive two low-complexity finger selection schemes for DS-UWB systems and compare their computational costs. Equalizers based on a limited number of fingers with finger selection schemes can achieve performance close to that of equalizers with all fingers within the processing window. In the second part, we investigate channel estimation for DS-UWB systems. To this end, we first study blind channel estimation methods. The simulation results indicate that blind estimation methods are not applicable because of the long delay spread nature of UWB channels. Therefore, training-based channel estimation is preferred. We further derive a low-complexity channel estimation method using training sequences. The simulation and numerical results show that the proposed training-based channel estimation is highly appropriate for practical implementation because it can achieve performance close to the theoretical limit while at the same time causing only little bandwidth wastage. In the third part, we study interference suppression for DS-UWB systems. We first devise structures and methods for interference suppression. In this context, we develop new adaptive filters to remove the multiple-access interference (MAI) by other unknown DS-UWB users or unknown narrow-band interferers (NBI). If the information of NBI is available at the receiver, we propose another filter with a simple structure to largely eliminate the impact of NBI. Simulation and numerical results show that the proposed adaptive filters can effectively suppress interference with moderate computational complexity. The narrow-band interference can be almost entirely removed with the a simple band-stop filter even when the interference power is very large. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate