Abstract
D.Ing.
Due to continuous research there is a large variety of new source
and channel coding technologies which are constantly being introduced
and refined. This study presents a new approach that enables
one to use enumerative source coding on variable length codewords.
The technique introduced are shown to provide an effective and fast
decoder. The encoding process depends on the code used and its qualities.
This new technique is illustrated by applying it to Huffman source
coding. As a result, an efficient and fast Huffman decoder was constructed.
The new decoder also resulted in small codebook representations. An efficient source coding mechanism, must be complemented
by a channel and error correction coding mechanism which is just as
efficient to ensure an optimal communication channel. We conclude this study by investigating channel and constrained
coding. We know that the implementation of error correction and detection
codes, like Reed-Solomon codes, are resource intensive with
longer codewords. This problem is circumvented by the introduction
of an alternative channel architecture. In this new architecture, a
channel code is applied to the source data before an error correction
code is applied to the channel data. For long codewords in the channel
code, this new approach is shown to be equal or superior to block and
sliding window codes. This new approach is block based, but unlike
block codes, usable in most types of channels.