Trellis decoding of linear block codes

Abstract

The purpose of this paper is the investigation of trellis decoding of linear block codes. The intersymbol relationship of convolutional codes makes the Viterbi algorithm ideally suited as a decoding strategy. By representing linear block codes in trellis form, the Viterbi algorithm can also be used to decode linear block codes. A distinct advantage of the Viterbi algorithm above traditional block decoders, is the existence of efficient soft decision algorithms employing channel measurement information. This paper will investigate a method of constructing trellises for block codes as well as the Viterbi decoding of block codes using these trellises. The proposed block trellis decoding technique will be supported with simulation results, comparing its performance with traditional methods such as syndrome decoding.