Fast Convergence Dual Mode Adaptive Decision Feedback Equalization

摘要

In this paper, we extend the blind adaptive equalization scheme of Labat, Macchi and Laot that starts up as a linear equalizer (LE) and switches into a decision feedback equalizer (DFE) once the channel eye is open. The modified scheme converges faster and switches more successfully to the DFE under severe equalization conditions such as short filter lengths and low signal-to-noise-ratios (SNR) even for a high-order modulation scheme, e.g., 64-QAM. This is achieved by employing the multi-modulus algorithm and a superior switching strategy. Monte Carlo simulations of a 64-QAM scheme verify the effectiveness of modified scheme.