Adaptive blind equalization criteria and algorithms for QAM data communication systems

A new error criterion for blind equalization is proposed, and it is shown that, after the startup period, the new error criterion converges to the conventional decision-directed MSE criterion, such that the transition between the two modes is smooth. It is proved that the error criterion improves the ill-conditioning probability. It is also shown that the convergence of the new criterion is independent of the initial equalizer tap settings. Its implementation complexity is similar to that of the conventional MSE equalizer algorithm. Comparisons are made with other blind equalization criteria and algorithms. Computer simulations of 16-point QAM support the theoretical analysis of the criterion convergence. A better performance is shown in the eye-pattern reopening test and the average error probability measurement