A new dual-mode blind equalization based on estimated error variance

Abstract

A blind equalization algorithm may switch to a decision-directed (DD) equalization mode in order to speed up the convergence rate and improve the steady-state performance. The crux for this kind of equalization is the control of the switches between the two equalization mode. A new blind equalization scheme is proposed which automatically controls the switches between the blind equalization mode and DD equalization mode according to the estimated error variance, and the blind equalization mode used in the proposed scheme is the modified constant modulus algorithm. Simulation results shows that the proposed blind equalization scheme can present significant improvements in the convergence rates along with steady-state performance