A Multi-Mode Approach for the Enhanced Decision Adjusted Modulus Algorithm Usage in Blind Equalization of QAM Signals

The paper explores the blind equalization problem of quadrature amplitude modulated signals. The principles of the most widely exploited blind equalization methods, such as the constant modulus algorithm and enhanced decision-adjusted modulus algorithm (EDAMA), are reviewed. The study focuses on the residual error and misadjustment probability balancing. The proposed approach utilizes a generalization of the EDAMA algorithm for the constellations with closely positioned circles and defines criteria for switching between different equalization cost functions. The algorithm aims to minimize the equalized signal dispersion and uses its estimate as a source for the mentioned cost function switching.