Title
Partial Product-LDPC Codes Without Rate Loss
Abstract
In this paper, we present a new construction of product codes, where all rows of the information array are encoded using the low-density parity-check (LDPC) codes while only partial columns of the information array are encoded using the algebraic codes. Also distinguished from the conventional product codes, the actual code rates of the presented partial product codes are the same as those of the row component LDPC codes, without any rate loss. This is achieved by using the free-ride coding to transmit the additional column checks along with the LDPC codewords. For decoding, an iterative decoding algorithm, which is scheduled as free-ride decoding, row decoding and column decoding, is proposed. At the column decoding stage, the messages associated with those successfully decoded columns are hardened and exploited to improve the free-ride decoding and the row decoding. By assuming a binary symmetric channel model after row decoding, we present an approximated bound on the word error rate (WER) of the doubly-protected (by both the row code and the column code) information bits. To predict the performance of the proposed codes, a genie aided (GA) lower bound derived from the original LDPC coded system with partial known bits is presented. These bounds and the simulation results reveal that the presented product-LDPC codes can have a significant performance improvement over the row component LDPC codes (about one dB at the WER of 10– <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</sup> ). They also reveal that the presented product-LDPC codes can have a flexible trade off between the error-floor and the waterfall performances, by using different row codes and column codes.
Year
DOI
Venue
2022
10.1109/ICCC55456.2022.9880620
2022 IEEE/CIC International Conference on Communications in China (ICCC)
Keywords
DocType
ISSN
Free-ride codes,LDPC codes,product codes
Conference
2377-8644
ISBN
Citations 
PageRank 
978-1-6654-8481-7
0
0.34
References 
Authors
6
5
Name
Order
Citations
PageRank
Qianfan Wang100.34
Suihua Cai200.34
Fanhui Meng300.34
Li Chen4367.36
Xiao Ma548764.77