Multilevel Coding and Unequal Error Protection for Multiple-Access Communications and Ultra-Wideband Communications in the Presence of Interference.

Abstract

Interference is one of the major factors that degrade the performance of a communication system. Various types of interference cause di Kerent impact on the system performance. In this thesis, we consider interference management at the physical layer. In order to enhance the performance, the receiver needs to have the knowledge about the interference. By exploiting the knowledge about interference, such as statistical properties, it can be suppressed to enhance the link quality. This thesis contains two main topics: multilevel coding (MLC) for unequal error protection (UEP) and receiver design for ultra-wideband (UWB) communications to suppress interference. Both topics deal with interference in di Kerent ways, and face di Kerent design challenges.

MLC is a way to provide UEP for different streams of information with different levels of importance in a communication system. It combines coding and modulation schemes to optimize the system performance. The idea is to protect each bit in the modulation constellation point by an individual binary code. We designed and analyzed a DS-CDMA system with asymmetric PSK modulation and MLC using BCH codes in an AWGN channel. The analysis includes probability of bit error of the system, and the capacity and throughput of the MLC scheme combined with 8-PSK modulation. The results show that the MLC scheme can have a higher throughput than the regular coding scheme in the low SNR region in the AWGN channel.

We also analyzed the performance of UWB communications in the presence of MAI and jamming interference. We considered a nonlinear interference suppression technique for impulse radio based UWB systems in the AWGN channel. The technique is based on the locally optimum Bayes detection (LOBD) algorithm, which utilizes the interference probability density function (PDF) for receiver design. This type of receiver has low complexity, and numerical results show that its performance asymptotically approaches that of the optimum receiver. Lastly, we discussed the implementation of the proposed receiver by adaptively monitor and update the interference PDF. The adaptive LOBD algorithm makes the proposed receiver implementation practical to deal with different types of interference.