Reed-Solomon码的符号级软判决译码算法

Symbol-level soft-decision decoding of Reed-Solomon codes

为了在译码性能和复杂度之间获得更好的折中,提出两种Reed-Solomon码的符号级软判决译码算法:一种将置信度排序译码集成到Chase译码的组合译码算法;另一种则采用逐级选择测试序列集的广义Chase-2译码算法。二者在一定的场合能够充分利用软判决信息。仿真结果表明,第一种算法在译码复杂度较低时,能够获得很好的综合性能;而第二种算法则具有很好的灵活性,且在译码复杂度可接受的范围内具有很好的综合性能。两种算法在一些场合中都能获得比目前其它一些常用的符号级软判决译码算法更好的综合性能。

To achieve good tradeoffs between error performance and decoding complexity, two types of symbol-level soft-decision decoding algorithms for Reed-Solomon codes are proposed. The first algorithm integrates order statistic decoding into Chase decoding while the second algorithm employs Chase-2 algorithm in a step-by-step version. Both algorithms make good use of the received soft-decision information. Simulation results show that the first algorithm performs well when the decoding complexity is low while the second algorithm provides good flexibility and performs well when the decoding complexity is affordable. Both algorithms outperform some popular symbol-level SDD algorithms in some cases.