Decision-Directed Channel Equalizer Scheme Based on the Recursive Non Quadratic Adaptive Algorithm

Abstract

This paper presents a new adaptive decision directed equalizer (DDE) for the receivers of Advanced Wireless communication systems. The proposed approach is based on the use of a recursive non quadratic (RNQ) adaptive algorithm in a DDE system with a training sequence, the new algorithm is denoted DDE-RNQ. The proposed DDE-RNQ equalizer achieves better performances in different situations even when strong noise components are present in the channel. In the proposed DDERNQ, the non quadratic distance criterion between the equalizer output and the decision circuit is used rather than the squared mean square error (MSE) distance for the computation of the branch metrics. Compared to the conventional DDE based normalized least mean square (DDE-NLMS) equalizer, the proposed DDE-RNQ shows superior performances of speed convergence and residual error versus the original DDE-NLMS equalizer in terms of MSE.