Hybrid Hard-Decision Iterative Decoding of Irregular Low-Density Parity-Check Codes

Time-invariant hybrid (Hscr_TI) decoding of irregular low-density parity-check (LDPC) codes is studied. Focusing on Hscr_TI algorithms with majority-based (MB) binary message-passing constituents, we use density evolution (DE) and finite-length simulation to analyze the performance and the convergence properties of these algorithms over (memoryless) binary symmetric channels. To apply DE, we generalize degree distributions to have the irregularity of both the code and the decoding algorithm embedded in them. A tight upper bound on the threshold of MB Hscr_TI algorithms is derived, and it is proven that the asymptotic error probability for these algorithms tends to zero, at least exponentially, with the number of iterations. We devise optimal MB Hscr_TI algorithms for irregular LDPC codes, and show that these algorithms outperform Gallager's algorithm A applied to optimized irregular LDPC codes. We also show that compared to switch-type algorithms, such as Gallager's algorithm B, where a comparable improvement is obtained by switching between different MB algorithms, MB Hscr_TI algorithms are more robust and can better cope with unknown channel conditions, and thus can be practically more attractive     

Hybrid hard-decision iterative decoding of regular low-density parity-check codes

Hybrid decoding means to combine different iterative decoding algorithms with the aim of improving error performance or decoding complexity. In this work, we introduce "time-invariant" hybrid (H/sub TI/) algorithms, and using density evolution show that for regular low-density parity-check (LDPC) codes and binary message-passing algorithms, H/sub TI/ algorithms perform remarkably better than their constituent algorithms. We also show that compared to "switch-type" hybrid (H/sub ST/) algorithms, such as Gallager's algorithm B, where a comparable improvement is obtained by switching between different iterative decoding algorithms, H/sub TI/ algorithms are far less sensitive to channel conditions and thus can be practically more attractive.     

Time-invariant and switch-type hybrid iterative decoding of low-density parity-check codes

Hybrid decoding is to combine different iterative decoding algorithms with the aim of improving error performance or decoding complexity. This, e.g., can be performed by using a specific blend of different algorithms in every iteration (time-invariant hybrid: H_TI), or by switching between different algorithms throughout the iteration process (switch-type hybrid: H_ST). In this work, we study H_TI and H_ST algorithms both asymptotically, usingdensity-evolution, and at finite block lengths, using simulations, and show that these algorithms perform considerably better than their constituent algorithms. We also investigate the convergence properties of H_TI and H_ST algorithms, under both the assumption of perfect knowledge of the channel, and the lack of it, and show that compared to H_ST algorithms, such as Gallager’s algorithm B, H_TI algorithms are far less sensitive to channel conditions and thus can be practically more attractive.     

Threshold values and convergence properties of majority-based algorithms for decoding regular low-density parity-check codes

This work presents a detailed study of a family of binary message-passing decoding algorithms for low-density parity-check (LDPC) codes, referred to as "majority-based algorithms." Both Gallager's algorithm A (G/sub A/) and the standard majority decoding algorithm belong to this family. These algorithms, which are, in fact, the building blocks of Gallager's algorithm B (G/sub B/), work based on a generalized majority-decision rule and are particularly attractive for their remarkably simple implementation. We investigate the dynamics of these algorithms using density evolution and compute their (noise) threshold values for regular LDPC codes over the binary symmetric channel. For certain ensembles of codes and certain orders of majority-based algorithms, we show that the threshold value can be characterized as the smallest positive root of a polynomial, and thus can be determined analytically. We also study the convergence properties of majority-based algorithms, including their (convergence) speed. Our analysis shows that the stand-alone version of some of these algorithms provides significantly better performance and/or convergence speed compared with G/sub A/. In particular, it is shown that for channel parameters below threshold, while for G/sub A/ the error probability converges to zero exponentially with iteration number, this convergence for other majority-based algorithms is super-exponential.