Fine-Grained Bit-Flipping Decoding for LDPC Codes

被引：11

作者：

Chen, Yuxing ^{[1
]}

Cui, Hangxuan ^{[1
]}

Lin, Jun ^{[1
]}

Wang, Zhongfeng ^{[1
]}

机构：

[1] Nanjing Univ, Sch Elect Sci & Engn, Nanjing 210008, Peoples R China

来源：

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS | 2020年 / 67卷 / 05期

基金：

中国国家自然科学基金;

关键词：

Low density parity check codes; fine-grained classification; bit-flipping; high throughput; low-complexity implementation;

D O I：

10.1109/TCSII.2020.2980846

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

This brief presents a novel class of hard-decision algorithms for decoding low density parity check codes. The new algorithms, named fine-grained bit-flipping (FBF) algorithms, employ a detailed classification of each bit, by introducing the XOR value of its estimated and received value as a subdividing criterion. The fine-grained classification allows the algorithms to strengthen the information utilization during each iteration. Simulation results show that the FBF algorithms can achieve up to 5 times better decoding performance than the state-of-the-art bit-flipping algorithms over the binary symmetric channel. Additionally, a well-optimized hardware architecture is developed for implementing FBF algorithms. Compared to other decoders, implementation results demonstrate that the FBF decoders achieve higher throughput and area efficiency.

引用

页码：896 / 900

页数：5