Принципи програмної реалізації біоінспірованого методу декодування алгебраїчних згорткових кодів

Abstract

It is shown that advisable to use algebraic convolution codes for increasing the reliability of the transmission of information in telecommunication networks based on radio technologies. An analysis of existing methods for decoding error correcting codes was carried out. It is shown that the problem of soft decoding of a given algebraic convolutional code can be formally presented as an optimization problem. The expediency of using the bioinspired approach together with the random bias procedure for soft decoding of algebraic convolution codes is substantiated. The main stages of the bioinspired decoding method of these codes are considered. For the effective software implementation of the bioinspired decoding method of these codes, the schemes of the corresponding algorithms have been developed and the features of their implementation have been determined. The received sequence, the parameters of the algebraic convolution code and the parameters of the chosen bioinspired procedure are the input data of the developed algorithm of bioinspired decoding of algebraic convolutional codes with the random bias. The output of this decoding algorithm is to estimate the most probable code sequence. The primary base vector of soft solutions, the most reliable basis of the generator matrix and the parameters of the selected bioinspired procedure are the input data in the presented algorithm of the subprocess of the bioinspired search of the codeword of the algebraic convolution codes. The output of this decoding algorithm is the best basic code sequence. Then this sequence is used in the basic algorithm for decoding algebraic convolutional codes.