COMPARISON OF METHODS SYNDROME AND SOFT DECODING FOR HAMMING CODE

Information technology and computer technology are rapidly developing, which leads to increased requirements for information transmission systems. To meet current security requirements, it is necessary to ensure a high level of information integrity and accessibility. One of the most effective ways to ensure the integrity of information is noise-resistant coding. The speed of data acquisition and the correctness of their recovery largely depends on the choice of the decoding method. Many decoding methods have a high complexity of implementation, so their use is not always effective. This article discusses a comparison of syndromic decoding and soft decoding methods for a linear Hamming code. The paper presents a comparison of the speed and correctness of the soft decoding algorithm described in patent RU 2738724C1 with hard decoding methods – syndrome decoding and the Viterbi algorithm. The implementation of the decoding algorithms and algorithm of parallel syndromes table computation without thread blocks on Java programming language has been carried out in this work. The paper provides a detailed description of the Hamming code, syndrome decoding and soft decoding, and a comparison of these decoding methods is demonstrated and analyzed in detail. Conclusions are presented showing that the proposed probabilistic decoding method has better characteristics in terms of speed and correctness than all other methods, and hard decoding methods have on average more errors than the soft decoding algorithm for a certain number of iterations.