多维奇偶校验乘积码性能分析

Performance Analysis of Multidimensional Parity Check Product Codes

相对于Turbo码,乘积码在高码率情况下具有性能优势,且无错误平层,其应用广泛。以奇偶校验码为分量码的乘积码译码算法简单,可适应不同的数据大小要求。通过分析多维奇偶校验乘积码的码多样性,得出高信噪比下该码的性能公式;计算机仿真表明:该性能公式结果与最大似然译码算法及迭代译码算法性能相符。通过分析码率、维数、分量码码长等参数对码性能的影响,得出码率、维数相同时,码性能相近;码率相同、维数不同时,高维乘积码具有性能优势。该结果为码参数选择提供依据,可在相同的信息传输有效性下,选择性能更好的码。

Different from Turbo codes, product codes have no error floor problem and have better performance at high rate. In the recent year, product codes are widely used. Parity check product codes, which have the simpler decoding algorithm, can adapt to different required data length. Its performance formula is proposed after analyzing the code multiplicity and information multiplicity. Computer simulations show that the performance of the formula is the same as that of maximum likelihood（ML） decoding algorithm and iterative decoding algorithm. Performance analysis, which is based on different parameters such as code rate, dimension, and code length, shows that multidimensional parity check product codes have close performance at same code rate and dimension, and the codes with higher dimension can get better performance at same code rate.