Start-up behaviour of a neural network assisted decision feedback equaliser in a two-path channel

The performance characteristics of the neural network assisted decision-feedback equalizer (DFE) have been investigated by extensive simulations using a two-path channel model and 16QAM (quadrature amplitude modulation). The authors report some of the simulation results on the start-up behavior of the novel equalizer structure under distorted channel conditions. The results show that, especially in difficult channels, including both linear multipath and nonlinear distortions, the neutral network assisted DFE structures are superior to the traditional DFE with equal computational complexity.<
>