Anticodes and error-correcting for digital data transmission

The work reported in this thesis is an investigation in the field of error-control coding. This subject is concerned with increasing the reliability of digital data transmission through a noisy medium, by coding the transmitted data. In this respect, an extension and development of a method for finding optimum and near-optimum codes, using N.m digital arrays known as anticodes, is established and described. The anticodes, which have opposite properties to their complementary related error-control codes, are disjoined fron the original maximal-length code, known as the parent anticode, to leave good linear block codes. The mathematical analysis of the parent anticode and as a result the mathematical analysis of its related anticodes has given some useful insight into the construction of a large number of optimum and near-optimum anticodes resulting respectively in a large number of optimum and near-optimum codes. This work has been devoted to the construction of anticodes from unit basic (small dimension) anticodes by means of various systematic construction and refinement techniques, which simplifies the construction of the associated linear block codes over a wide range of parameters. An extensive list of these anticodes and codes is given in the thesis. The work also has been extended to the construction of anticodes in which the symbols have been chosen from the elements of the finite field GF(q), and, in particular, a large number of optimum and near-optimum codes over GF(3) have been found. This generalises the concept of anticodes into the subject of multilevel codes.