Reduced Complexity Super-Trellis Decoding for Convolutionally Encoded Transmission Over ISI-Channels

TL;DR

This paper introduces a matched encoding scheme with a novel trellis structure for convolutionally encoded transmission over ISI channels, enabling joint equalization and decoding with reduced complexity.

Contribution

It presents a new matched non-linear trellis description that significantly reduces the number of states in super-trellis decoding for ISI channels.

Findings

Reduced number of states in Viterbi decoding

Efficient joint equalization and decoding

Complexity reduction via reduced-state sequence estimation

Abstract

In this paper we propose a matched encoding (ME) scheme for convolutionally encoded transmission over intersymbol interference (usually called ISI) channels. A novel trellis description enables to perform equalization and decoding jointly, i.e., enables efficient super-trellis decoding. By means of this matched non-linear trellis description we can significantly reduce the number of states needed for the receiver-side Viterbi algorithm to perform maximum-likelihood sequence estimation. Further complexity reduction is achieved using the concept of reduced-state sequence estimation.