Structured Ldpc Codes And Adaptive Coding Techniques

In next mobile communication system to suffice more and more high-speed data service and demand of Qos(quality of service) many new wireless link layer transport technologies are going to be used such as MIMO(multiple input multiple output), OFDM(Orthogonal Frequency Division Multiplexing), channel coding and ACM(adaptive coding modulation) etc.Low density parity check (LDPC) codes were first discovered in 1960's which belong to linear block codes with their parity matrix being sparse. Later in 1990's they were rediscovered and soon became the focus in channel code fields because of their remarkable performance close to Shannon limit with iterative decoding and their structures for parallel decoding. As the random structure of LDPC codes made it difficult to encoding in hardware and few mathematic methods has been found for analyzing these codes, many constructed LDPC codes were investigated to simplify encoding circuit and to reduce the complexity of analyses.This paper made a careful study on a class of quasi-regular LDPC codes which could encode in linear time only making use of their parity matrix such asπ-rotation LDPC codes. Two factors of the minimum distance and girth length were discussed to improve the performance of these LDPC codes. At the same time, a novel approach to construction of LDPC codes based on the permutation matrix was proposed which got a parity check matrix with the shortest girth is 6. These codes keep the advantage of linear encoding and become more flexible. Simulations show that the performance of short length codes of this kind is as excellent as other LDPC codes.Adaptive coding technology is often used to improve frequency efficiency of wireless communication system as we know that the adaptive coding scheme based on Turbo codes has been carried out in 3G systems. Since LDPC codes have a wonderful future, an adaptive coding scheme was developed based on quasi-regular LDPC codes and it encodes the source bits with different code rate by dividing the original parity check matrix properly. The simulation results based on SISO and MIMO systems show that the proposed scheme could effectively eliminate the time-variant fading characteristics of wireless channel and under an acceptable demand of BER, the frequency efficiency is enhanced greatly and besides, dynamic encoding is implemented with low complexity and high efficiency.At the end of this paper SNR estimation algorithm is discussed which plays an important role in adaptive coding technology. Several representative SNR estimation methods were compared through the computer simulations and the results can help to select a suitable SNR estimation algorithm in different wireless communication systems.