A new blind equalizer for high-order QAM system

A blind equalization algorithm based on the unsupervised Gaussian cluster formation technique with a robust adaptive step-size to update the equalizer coefficients is proposed. In order to guarantee the convergence of the algorithm, the algorithm runs in a "stop-and-go" operation mode in such a way that a Godard class of algorithm controls its "stop-and-go" operation. In order to achieve faster convergence speed, the algorithm is switched to a Godard class of algorithm when adaptation is stopped. Simulation results confirm the effectiveness of the proposed algorithm on high-order QAM signals.