Lattice based soft-constraint satisfaction multi-modulus blind equalization algorithm

Abstract

In this paper, a method of accelerating the speed of Convergence of a blind equalization algorithm is examined. It is shown that as in conventional equalizers the orthogonalizing properties of lattice algorithms make them appear attractive in blind equalization of a channel. The lattice is applied to a newly proposed blind equalization algorithm [I], [2], known as soft-constraint satisfaction multi-modulus algorithm (SCS-MM-I). Experiments show that the introduction of a stochastic gradient lattice structure in SCS-MM-I results in an increase of convergence rate by an order of magnitude.