CAREER: A Unified Framework for the VLSI Design of Algebraic Soft-decision Reed-Solomon Decoder

本研究的目的是首先为代数软判决里德-所罗门译码器设计高效的超大规模集成电路(VLSI)系统结构。然后，通过研究结构随设计参数的变化，进行进一步的优化，为代数软判决里德-所罗门译码器的VLSI设计提供一个统一的框架。该方法是对译码算法和VLSI结构进行交互优化，以便根据特定的应用需求设计出最高效的译码结构。在智能方面，该研究为先进的理论和实践搭建了一座桥梁。这项研究的关键创新潜力是算法转换和新颖的VLSI设计技术之间的微妙相互作用。与由算法决定VLSI结构的传统设计方法不同，VLSI将采用交互式算法和结构优化。特别是，从VLSI设计级到算法级的反馈对于实现最高效的设计和开发优化框架至关重要。解决这个项目中具有挑战性的任务需要结合编码理论、数字通信和VLSI设计方面的专业知识。就更广泛的影响而言，拟议的研究可能会给数字通信和存储系统带来革命性的变化。待开发的设计技术也可以外推到其他先进算法的实现。建议的研究将通过开发动态课程作业，让本科生和研究生参与研究，并向初中生推广工程学，将其整合到教育中。此外，还将通过招聘和留住妇女，增加来自任职人数不足群体的学生在工程学领域的参与度。

The objective of this research is to first design efficient very large-scale integrated (VLSI) system architectures for algebraic soft-decision Reed-Solomon decoders. Then, through studying how the architectures change with design parameters, further optimizations will be carried out to develop a unified framework for the VLSI design of algebraic soft-decision Reed-Solomon decoders. The approach is to optimize the decoding algorithms and VLSI architectures interactively, such that the most efficient decoder architecture can be designed for given application requirements.With respect to intellectual merit, the proposed research serves as a bridge connecting advanced theory and practice. The key potential for innovation in this research is the delicate interplay among algorithmic transformations and novel VLSI design techniques. Unlike traditional design approaches in which the VLSI architectures are determined by the algorithms, interactive algorithmic and architectural optimizations will be employed. Particularly, the feedback from the VLSI design level to the algorithmic level are critical to achieving the most efficient designs and developing an optimized framework. Addressing the challenging tasks in this project requires combined expertise in coding theory, digital communications and VLSI design.With respect to broader impacts, the proposed research could potentially revolutionize digital communication and storage systems. The design techniques to be developed can be also extrapolated to the implementation of other advanced algorithms. The proposed research will be integrated into education through developing dynamic course work, involving both undergraduate and graduate students in research, and promoting engineering to middle and high school students. In addition, the participation of students from underrepresented groups in engineering will be increased by recruiting and retaining women.