中文摘要：

针对最优映射下低密度奇偶校验（LDPC）编码调制系统无迭代增益的问题，采用因子图和互信息方法对系统进行了研究，提出了一种基于联合信息的重加权置信传播算法。和传统迭代接收算法相比，该算法改进了迭代结构，不仅增加了调制器，获得了格雷映射下的迭代增益，而且在解调器的外信息中引入加权符号信息，在译码器的外信息中引入指数型先验信息，合成为联合信息，增大解调器和解码器之间的平均互信息值，进一步提高了系统性能。从互信息角度说明该算法的合理性，同时分析了权值和自适应指数对算法性能的影响。给出了在瑞利衰落信道下的仿真结果，验证了该算法的有效性和优越性。

英文摘要：

The methods of factor graph and mutual information were utilized to analyse the performance of the low-density parity check （ LDPC ） coded modulation system under optimal mapping to solve its problem of no iteration gain, and then a reweighted belief propagation algorithm based on joint information was proposed. Compared to conventional iterative receive algorithms, the new algorithm modifies the iterative scheme by adding a modulator and the performance gains in Gray mapping are achieved. Furthermore, the algorithm introduces the reweighted symbol information into the demodulator＇ s extrinsic information and adds the exponential priori information into the extrinsic information of the decoder. Thus, the joint information is formed to increase the average mutual information between the demodulator and the decoder, and the system performance is further improved. The rationality of the proposed algorithm is substantiated from the perspective of mutual information. The simulation results for Rayleigh fading channel and different decoder algorithms illustrate the effectiveness and performance improvement of the proposed algorithm compared with conventional iterative algorithms.