| With their huge bandwidth and correspondingly abundant channel diversity and fine timing resolution, ultra-wideband (UWB) systems have proved able to provide higher communication capacity and better ranging accuracy than other narrowband systems. However, to achieve this, accurate channel estimation and synchronization must be obtained, and efficient diversity combining algorithms must be devised. On the other hand, their GHz wide bandwidth, which provides these advantages, also means they have to operate in narrowband interference. These are the issues that are addressed here: how to better make use of the rich multipath diversity and fine timing resolution of UWB systems in narrowband interference with a low receiver complexity. This dissertation begins with a comparison of transmitted-reference (TR) and stored-reference (SR) UWB systems. By incorporating the channel estimation errors for SR systems, a more realistic comparison between TR- and SR-UWB systems is presented, which provides a better understanding of their respective advantages and disadvantages. Furthermore, a receiver structure is proposed for TR-UWB systems, which not only improves its performance but also preserves a low-complexity structure. The second part of the dissertation considers the problem of signal tracking for SR-UWB systems in narrowband interference, where a tracking loop is optimized to provide more accurate timing information than other traditional loops for the receiver. With a better tracking capability, the receivers are then able to take advantage of the rich multipath diversity offered by the UWB propagation channels. In the last part, different receiver structures and algorithms for SR-UWB systems are proposed and compared in the presence of narrowband interference and under some resource constraint. These receiver structures are designed to better make use of the rich diversity of UWB multipath channels with the lowest complexity possible. Finally, a performance and complexity design trade-off is presented to conclude the dissertation. |
