10307036陈燕.docVIP

版权声明 任何收存和保管本论文各种版本的单位和个人，未经本论文作者同意，不得将本论文转借他人，亦不得随意复制、抄录、拍照或以任何方式传播。否则，引起有碍作者著作权之问题，将可能承担法律责任。摘要 低密度校验(LDPC)码作为一种可逼近的信道编码技术，正在被广泛应用于通信系统中关键词Abstract As an approaching Shannon limit channel code, Low-density Parity-check Code (LDPC) has been applied to communication systems widely. This paper has done some research on the choice of quantization rule, the performance of LDPC-16QAM system on the multi-path channel, DSP implementation of an LDPC decoder. The detail contents are as follows: Firstly, we studied the quantization rule. We found when the signal power is equal to 1, the quantization range is (-4,4) and the performances of 5, 6 or 7 quantization bits are all close to ideal performance. Both the code our lab constructed and the code from standard IEEE 802.16e are implemented and compared. Secondly, we presented an implementation of LDPC decoder on DSP. The DSP used is a Texas Instruments TMS320C5409 fixed point processor. In order to use as few cycles as possible, we arranged the messages in memory in a special way. We first calculated the variable nodes and then the check nodes including the check sum. The average decoding speed is about 20.4kbit/s when the iterative number is 10. Thirdly, we constructed LDPC-16QAM system, researched on the associated algorithm of demodulation and decoding. We also applied code our lab constructed, code from IEEE 802.16e and convolutional code in standard DVB-T on several kinds of channels. The result showed that the performances of two LDPC codes were close and both of them were better than convolutional code. Also, we found that whether the channel was good or bad not only depended on the number of paths but also the power distribution among the channels when the SNR is fixed. Key Words: Low-Density Parity-Check codes (LDPC), Belief Propagation, Min-Sum, 16QAM modulation, multi-path channel, DSP目录 第一章 绪论 1 1.1 引言 1 1.2 论文的主要工作 2 1.3 论文的主要结构 2 第二章 LDPC码的基本理论 3 2.1 信道编码技术的发展 3 2.2 LDPC码的定义 5 2.3 LDPC码的译码算法 7 2.3.1 置信度传播译码算