Abstract
In this thesis, we propose using two methods to improve the performanceof a noncoherent system using information-bearing pilot and spacial multiplexing. The former is based on EXIT(extrinsic information transfer) chart and the latter is based on the codeword-interleaving strategy. It is known that irregular low-density parity-check (LDPC) code is widely used for optimizing performance in various communication systems. However, in some cases, the Gaussian approximation for the soft values used in the conventional EXIT-chart-based optimization procedure may lose its accuracy, and result in a large prediction error for the convergence threshold. It can be shown that for a noncoherent space-time system using information-bearing pilots and spatial multiplexing, the probability density function of the soft values is far from Gaussian. We have therefore designed an efficient performance optimization method. Compared to the use of conventional optimization method, using the proposed method provides a more accurate prediction of convergence threshold and better system performance. For the codewordinterleaved scheme, which uses extra pilots to help demap and decode, we introduce a method that can be used to extend it to our system. With a different codeword interleaver, the extra information being passed to the adjacent codeword may be different and, hence, produce different results.