Abstract
M.Ing. (Electrical Engineering)
As the demand for high-speed data communications over conventional channels such as coaxial cables and twisted pairs grows, it becomes neccesary to optimize every aspect of the communication system at reasonable cost to meet this demand effectively. The choice of a line code is one of the most important aspects in the design of a communications system, as the line code determines the complexity, and thus also the cost, of several circuits in the system. It has become known in recent years that a multilevel line code is preferable to a binary code in cases where high-speed communications are desired. Apart from ternary codes, not many multilevel codes are available. Some of the existing line codes also suffer from serious drawbacks regarding a lack of complying to input restrictions, small values of efficiency, and great code complexity.
In this study, Markov models and values of channel capacity are presented for several
classes of restricted multilevel sequences which are thought to be of practical importance in view of the channel input restrictions that these codes satisfy. Different coding methods are used to construct low-complexity encoders and decoders for generating and decoding these sequences with high values of efficiency, good error behaviour and favourable power spectral densities