Modified Algorithm for Decoding Block Turbo-Product Codes

Abstract

Ensuring reliable data transmission is an essential requirement for modern communication systems. To implement it, an error-correcting turbo-product code can be used, which can work effectively with small information blocks at high code rates, but requires significant computational costs at the data decoding stage, which can significantly load the communication system, especially in real-time mode and under conditions of limited resources. It should be noted that turbo-product code decoding algorithms are difficult to implement effectively. The most common is the Pyndiah–Chase decoding algorithm, which uses a brute force of test codewords, which poses a global discrete optimization task with exponential complexity. It is important to balance the reliability of error correction and the efficient use of computational resources. Therefore, in this article developed a suboptimal modification of this decoding algorithm, which is based on a heuristic procedure for generating a list of probable codewords.

The effectiveness of the modification is shown in experiments using computer simulation of a digital data transmission system, taking into account random information packets and multimedia data. During the experiments, the bit and symbol error rates were analyzed depending on the signal-to-noise ratio in the communication channel. It is determined that the modified algorithm is slightly outperforms its original version in terms of energy gain, and, taking into account the heuristic nature of codeword brute force, has a complexity lower or, in the most unlikely case, the same as that of the basic algorithm. The article can be useful for communication system designers, since it provides an opportunity to analyze and synthesize turbo-product code decoders for use in different practical tasks.