Quantization based reliable equalization algorithm for turbo DFE and dual-mode blind equalizers

Abstract

In this paper we propose a new adaptive equalization technique for Turbo Decision Feedback Equalizers (TDFE) and dual-mode blind equalizers, that takes the advantage of quantization in the probabilistic scheme. In generating the error signal the proposed technique considers the likelihood between the quantized equalizer output and all the constellation points, thus making the receiver insensitive to the decoding errors. The proposed technique achieves lower BER with less computations than other TDFE algorithms. Also a more reliable soft transition between start-up and Decision Directed (DD) mode in blind equalizers is proposed. This technique has been compared with Benveniste-Goursat, Stop-and-Go and Lim's soft transition algorithm. It is shown that the proposed approach is superior and more robust under different channel and SNR conditions.