Improved Performance Analysis of LDPC Codes Based on a New Parity Check Matrix Construction

Nguyen Tung Hung

The article introduces a new LDPC decoding algorithm based on Equivalent Parity Check Matrix. Simulation results show that the new LDPC decoding algorithm can improve LDPC decoding performance. Compared to some other improvements, the new LDPC decoding algorithm, which is simpler, can detect errors and be applied to LDPC codes for great-length LDPC codes.

B. N. Sindhu Tejaswini, Rajendra Prasad Lal, Vadlamudi Ch. Venkaiah

Li Peng, Guangxi Zhu, Xiao Liu

This paper presents a new family of low-density parity-check (LDPC) code, the sparse parity-check matrix of which is constructed by self-defining non-diagonal identity sub-matrix that is a solution of the "n-queen problem". So this sub-matrix is called the Q-matrix and these LDPC codes are called the Q-matrixes LDPC codes. The Q-matrixes LDPC codes are good or very good codes with iterative decoding and their Tanner graphs are free of 4-lines cycle. Furthermore, they can be created in cycle form. Their encoding can be achieved in linear time. Especially, their code length and code rate can be flexible and quickly adjusted according to the practical situation, and the performance of high rate is also very good. The other unique excellence is that the large sparse parity-check matrixes of long Q-matrixes LDPC codes require very small storage space. The result of this paper is very significant not only for designing low complexity encoder, improving performance and reducing the complexity of the sum-product iterative decoding algorithm, but also for building practice system of encodable and decodable LDPC code.

Bahram Honary

Lizhi Zhao

A method on construction of LDPC Parity-check matrix was proposed,by removing loops with girth 4 in a semi-random parity-check matrix.Simulation results show that the proposed method can effectively avoid the influence on the performance of LDPC codes exerted by short loops,so the decoding performance is distinctively improved.Furthermore,the better performance can be obtained as the increasing of code lengths.