Improving Blind Equalization Algorithm for Wireless Communication Systems

Blind equalization technique is an effective approach for canceling the intersymbol interference (ISI) induced by multipath effect in digital wireless communications. The convergence behavior as well as the residual mean square error (MSE) of a blind equalizer is an interesting issue in its design. Perhaps the main drawback of the traditional constant modulus algorithm (CMA) is its relatively slow convergence, especially in the time varying multipath wireless channels. In order to overcome this difficulty, a novel combined blind equalization algorithm based on a normalized iterative method is proposed in this paper. At the beginning of the iteration process, the normalized super-exponential iterative algorithm is utilized to perform an initial equalization, and then the iteration process switches to a decision-directed (DD) algorithm at appropriate values of the mean square error. This switch method can obtain a great improvement in both convergence behavior and residual MSE. The performance of the proposed algorithm is verified through computer simulations.