Remarks concerning the optimization of matrix codes

This paper describes the results of a general theory of matrix codes correcting a set of given types of multiple errors. A detailed study has been made of certain matrix classes of these systematic binary error correcting codes that will correct typical errors of some digital channels. These codes published by Elias,^(2,3) Hobb's,⁽⁵⁾ and Voukalis⁽¹¹⁾ account for this theory and other new families of binary systematic matrix codes of arbitrary size, correcting random, burst and clusters of errors are given here. Also presented here are the basic ideas of each of these codes. We can easily find practical decoding algorithms for each of these codes. The characteristic calculation of the parity check equations that the information matrix codebook has to satisfy are also shown. Further on we deal with the optimum construction of these codes showing their use in certain applications. We answer questions such as: “What is the optimum size of the code?” “What is the best structure of the code?” “What is the probability of error correction and the mean error correction performance?” Consequently, in this paper we also describe the results of an extensive search for optimum matrix codes designed to correct a given set of multiple errors as well as their implementation.