Maximum asymptotic efficiency equalizer with decision feedback

This work extends the maximum asymptotic efficiency (MAE) equalizer to include decision feedback (MAE-DFE). A geometrically-inspired approach is used to design the MAE-DFE and analyze its equalization performance. It is shown that with decision feedback, the MAE-DFE not only provides better performance than the original MAE equalizer, but also significantly reduces the computational complexity. We investigate the effect of feedforward filter length, feedback filter length, and decision delay on the performance of the MAE-DFE. In the design of the MAE-DFE, practical rules to determine the three parameters are derived to optimize its performance, and numerical results are shown to demonstrate their effect. The effect of error propagation on the MAE-DFE is also shown in simulations.