可重构的环境自适应RS码软判决译码器研究

Represented by deep-space detection, typical applications of Reed-Solomon (RS) codes have rapidly developed and raised higher requirements for the decoding technologies of RS codes, which include: (1) excellent error correction performance and simple implementation; (2) the ability to automatically adjust decoding parameters to the changeable channel environment with the goal of reasonably balancing the error correction performance, the processing speed and the power dissipation; (3) applicability to different assignments with specific requirements for data transmissions. The previously reported works can't individually achieve all above goals. This project will be conducted as follows. (1) Design a novel low-complexity Chase (LCC) decoding algorithm and its VLSI implementation, which can achieve a high coding gain while further reduce the hardware complexity. (2) Exploit the channel detection and the parameter self-adjusting technology, so that the decoder can always keep the best tradeoff between the processing speed and the energy efficiency on the premise of ensuring the quality of communication under the changeable channel environment. (3) Research on the reconfigurable technology, which can satisfy various applications' requirements with the lowest hardware consumption. The reconfigurable and environment self-adaptive algebraic soft-decision RS decoder based on above studies will demonstrate strong innovation, and can be an effective way to solve the issues faced in the applications of RS codes such as the deep-space detection. Moreover, the proposed decoder can ensure fast and reliable data transmissions under various situations and hence provide a powerful support for the system functioning. The research results are also conducive to expand the application scope of RS codes.

以深空探测为代表的RS码典型应用的快速发展对译码技术提出了更高的要求。包括：（1）纠错性能优异且结构简单；（2）能够根据环境的变化自动调整，保持译码效果与处理速度、硬件功耗的合理平衡；（3）可适应不同任务对信息传输的具体要求。已报道的相关研究成果无法满足上述要求。本项目将开展：（1）新型LCC软判决译码算法与VLSI实现研究，在保证高编码增益的同时，进一步降低硬件复杂度；（2）信道探测与参数自动调整技术研究，使译码器可根据信道环境变化，在保证通信质量的前提下，保持最佳的数据处理速度和能效状态；（3）可重构技术研究，以最低的硬件资源消耗满足多种业务需求。将上述研究有机结合形成的可重构环境自适应RS码软判决译码器创新性强，可有效解决深空探测等RS码现有应用领域所面临的难题，保证不同情况下数据快速可靠传输，从而为系统功能的顺利实现提供保障。研究成果还将有助于进一步拓展RS码的使用范围。

基于RS码译码器的发展趋势和深空探测等具体应用需求，本项目探索了可重构且环境自适应的高性能RS码软判决译码关键技术及相应的译码器。主要研究内容包括：(1) 完善重数分配的理论研究，提出新型重数分配算法及自适应模块，可通过处理从重数分配模块接收到的可靠度信息，对信道质量进行评估，并根据评估结果自动调整参数，实现译码器对环境的自适应；(2) 提出一种低复杂度关键方程求解模块，通过采用新的多项式系数更新方式，显著减少处理单元的数目；(3) 实现一种新的具有更高硬件效率的SPCF模块，该模块能在不同时间段分别实现校验子计算、多项式选择、钱搜索和福尼算法四种功能，大大减小了硬件面积；(4) 在保证高编码增益的条件下，通过形成具有规整阵列结构的核心模块，实现可重构的环境自适应RS码软判决译码器，最后对该译码器进行仿真与验证。通过四年的研究工作，本项目设计出的译码器可译码RS(255, 223)，RS(255, 239)两种码率，数据吞吐率可以达到10.13Gbp/s，而功耗仅为27.6mW。在满足误码率低于10-5的条件下，译码器能够自动调整译码参数，实现最优数据处理速度和硬件功耗，从而有效满足深空探测等具体应用系统对译码器高质量通信、低复杂度和小型化等需求。