Abstract
i The significant features of ultra-wideband (UWB) technology are broader spectrum and lower power transmission, which can be applied to improve data rate and reduces interference in wireless communications. In order to meet the demand for high data rates, the direct-sequence code-division multiple access (DS-CDMA) technology is applied to UWB systems. The basic operations of a DS-CDMA UWB system are similar to the conventional DS-CDMA system except that the pulses are transmitted rather than carriers. One major problem associated with DS-CDMA is that its capacity is seriously limited by multiple access interference (MAI), and this is the same issue for DS-CDMA UWB systems. To overcome this problem, we have to employ some multiuser detection techniques to deal with the MAI problems for DS-CDMA UWB systems. Unlike other wireless communication systems, a UWB system faces a short range indoor environment and serious multipath channels. In this thesis, we first introduce several conventional multiuser detection techniques, such as parallel interference cancellation (PIC), partial parallel interference cancellation (PPIC), and simplified partial parallel interference cancellation (SPPIC). Then we modify the conventional multiuser detection techniques by re-defining the channel model for UWB systems. To further improve the performance of PPIC, some approaches that adjust the partial weights of PPIC are discussed. One common approach of weight selection is adaptive multistage PIC, but this approach has to be operating at the chip rate and is difficult to implement for an ultra-high rate situation such as DS-CDMA UWB systems. To solve this problem, we propose a new adaptive weighted PPIC by simplifying the adaptive frequency from the chip rate to the bit rate, called bit-level adaptive weighted PPIC. Similar to the original multiuser detection technique, we derive this new adaptive weighted PPIC for UWB channels. Finally, we show some computer simulation results of proposed algorithm and different multiuser detection techniques. The results indicate that this proposed algorithm can actually improve the system performance of DS-CDMA UWB with lower computational complexity.