Design of irregular LDPC codes with uniform check node Degree distributions

Kenta Kasai, Kohichi Sakaniwa

Masoud Ardakani, Frank R. Kschischang

We introduce a new one-dimensional (1-D) analysis of low-density parity-check (LDPC) codes on additive white Gaussian noise channels which is significantly more accurate than similar 1-D methods. Our method assumes a Gaussian distribution in message-passing decoding only for messages from variable nodes to check nodes. Compared to existing work, which makes a Gaussian assumption both for messages from check nodes and from variable nodes, our method offers a significantly more accurate estimate of convergence behavior and threshold of convergence. Similar to previous work, the problem of designing irregular LDPC codes reduces to a linear programming problem. However, our method allows irregular code design in a wider range of rates without any limit on the maximum variable-node degree. We use our method to design irregular LDPC codes with rates greater than 1/4 that perform within a few hundredths of a decibel from the Shannon limit. The designed codes perform almost as well as codes designed by density evolution.

Shinya Miyamoto, Kasai Kenta, Kohichi Sakaniwa

Ryo Shibata, Gou Hosoya, Hiroyuki Yashima

Over the past two decades, irregular low-density parity-check (LDPC) codes have been hardly able to decode information corrupted by insertion and deletion (ID) errors without markers. Surprisingly, in this paper, we bring to light the existence of irregular LDPC codes that approach the theoretical limit of the channel with ID errors even without markers. These codes contain high fractions of low-degree check nodes that do not appear in irregular codes for other channels. This motivates us to investigate the contribution of low-degree check nodes to correcting ID errors. The investigation provides the following interesting result: degree-2 check nodes are critical to approaching the theoretical limit even without markers, codes with only degree-3, 4, or more check nodes provide moderate decoding performance, and codes with only degree-5 or more check nodes can hardly correct ID errors. Finally, we present simulation results that confirm the excellent decoding performance of the irregular codes without markers.

健太 笠井

Mahmoud Rashidpour, S. Hamidreza Jamali

The asymptotic distance distribution of regular low-density parity-check (LDPC) codes derived by Litsin and Sherlev (2002) is generalised for the case of irregular LDPC codes with constant check node degree. Numerical analysis of simple irregular LDPC codes with weight-two variable nodes shows that their minimum Hamming distance can be increased linearly with the code length when the maximum variable node degree of the code is kept below a critical value. This result is consistent with the stability condition of the LDPC codes, at least for the simple irregular case.