LDPC Simulation Software

This LDPC software is designed to introduce computer-based simulation of LDPC codes. The pseudo-random irregular low-density parity-check matrix implementation is based on Radford M. Neal's program collection, which can be referenced in [1]. Although Neal's collection is well-documented, the C language source code may present challenges for those unfamiliar with C programming. Our software is implemented in MATLAB, offering more readable code structure with clear matrix operations and algorithm flow. The implementation includes key functions for matrix generation, encoding procedures, and iterative decoding algorithms. You may also reference another MATLAB-based LDPC source code in [2], which employs a different coding style (notably, Arun's log-likelihood decoder contains implementation errors that have been corrected in our version).

Furthermore, we recommend gaining fundamental knowledge of LDPC coding principles to better understand the implementation and simulation processes. The software demonstrates practical applications of parity-check matrix construction, message passing algorithms, and error correction mechanisms. Recommended readings include classic LDPC texts such as "Modern Coding Theory" and "LDPC Codes: Construction, Decoding and Performance Analysis." These resources help establish comprehensive understanding of LDPC coding theory, enabling more effective use of this software for simulation and performance analysis tasks. The code includes detailed comments explaining the sum-product algorithm implementation and performance optimization techniques.