Research On The Efficient LDPC Encoding/Decoding And Its Combined Adaptive Equalization Techniques For Wireless Communications

The linear block code is called a binary low-density parity-check code if it is based on a sparse parity-check matrix. This sort of code was originally proposed by Dr. Gallager in 1962, which cannot attract a large amount of interest at that time. However, after decades of years, it is now rediscovered by Mackay and Neal, along with the enhancement of computer ability and the development of some relative theory such as graph theory, BP algorithm, turbo codes and so on. It is also shown that the performance of LDPC codes is close to the Shannon limits when combined with BP decoding algorithm.LDPC codes and turbo codes are similar in many aspects, both can extremely approach to the Shannon limits by their unique ways. The LDPC codes are relatively easy defined and can outperform the turbo codes with sufficiently long block lengths. Its decoding complexity is also lower than turbo code. Recently, LDPC code has drawn the worldwide attentions in channel coding community due to its high performance and strong potential applications in practice.This thesis offers a comprehensive study on the performance of LDPC codes. The author first presents a new method for construction of irregular LDPC codes based on generator matrix in order to significantly reduce encoding complexity. The simulation results are obtained over an AWGN channel, and the obtained results are also compared with those acquired by the traditional approaches. In order to further improve the performance, we propose two new kinds of codes, which are LDPC/turbo code, simply denoted by"L/T"code, and the product code with generator-based systematic irregular LDPC codes as its component codes in rows and columns. The factors that can impact the performance of the coded systems are investigated and concluded through the simulations. Finally, a new approach combining equalization and adaptive equalization with LDPC decoding are proposed to combat the intersymbol interference (ISI) and channel noise. Meanwhile, we also study the convergence behavior of iterative detection in different methods, which are the proposed average source entropy and the covariance matrix perturbation characterized by the variations of its dominant eigenvalue as well as the conventional approach based on the mutual information.