Error-Pattern-Correcting Codes: A New Approach to Error Correction Coding for Interference-Dominant Channels

纠错码：一种针对干扰主导信道的纠错编码新方法

• 批准号: 0728676
• 负责人: Jaekyun Moon
• 金额: $ 30万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-10-01 至 2011-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0728676&HistoricalAwards=false
• 关键词: Error; Pattern; Correcting; Codes; New

Error-Pattern-Correcting Codes: A New Approach to Error Correction Coding for Interference-Dominant ChannelsJaekyun Moon, University of MinnesotaAbstractIntersymbol interference (ISI) is a major source of transmission errors in a wide variety of communications channels. This research concerns error correction for such channels. Input-constrained ISI channels are of specific interest in this research, as efficient coding methods for such channels are still largely an open area. This work focuses on 1) developing low-complexity error correction methods for input-constrained ISI channels, with emphasis on high-density data storage applications, and 2) clearing a path to approaching theoretical limits of error correction for such channels with fundamentally improved complexity/performance tradeoffs. One effective way of designing codes for input-constrained ISI channels is to target at dominant error cluster patterns inherent at the channel output, as the structure of the ISI typically give rises to a dominance of a relatively small set of distinct error cluster patterns, regardless of the channel signal-to-noise ratio. This set can easily be found using distance analysis or through empirical observation based on simulation or laboratory measurement. Focusing only on the dominant set, very high rate codes are possible with excellent overall error correction capability. Encoding and decoding are necessarily highly tailored to the few known error patterns. This type of code can be applied either as a stand-alone code or as a critical building block in a larger coding system, such as a turbo equalizer based on iterative soft decision processing. While the research addresses computationally efficient approaches, efforts are also directed to establishing a theoretical framework as well as to understanding theoretical performance limits of the channel-matched codes.

码间干扰（ISI）是各种通信信道中传输错误的主要来源。本研究关注的是这类信道的纠错。本研究对输入受限ISI信道特别感兴趣，因为这种信道的有效编码方法在很大程度上仍然是一个开放的领域。这项工作的重点是1)开发输入受限ISI信道的低复杂度纠错方法，重点是高密度数据存储应用；2)为接近此类信道的理论纠错极限扫清了道路，从根本上改善了复杂性/性能折衷。为输入受限的ISI信道设计编码的一种有效方法是针对信道输出固有的主要错误簇模式，因为ISI的结构通常会导致相对较小的一组不同的错误簇模式占主导地位，而不管信道信噪比如何。通过距离分析或基于模拟或实验室测量的经验观察，可以很容易地找到这一组。仅关注优势集，就可以获得非常高的码率，并具有出色的整体纠错能力。编码和解码必须高度适应少数已知的错误模式。这种类型的代码既可以作为独立代码应用，也可以作为大型编码系统中的关键构建块应用，例如基于迭代软决策处理的涡轮均衡器。虽然研究解决了计算效率的方法，但努力也指向建立理论框架以及理解信道匹配码的理论性能限制。