“stop-and-go” decision-directed blind adaptive equalization using the complex-valued multilayer perceptron

A “stop-and-go” decision-directed blind equalization scheme is newly proposed. This scheme uses the structure of complex-valued multilayer feedforward neural networks, instead of the linear transversal filters that are usually used in conventional LMS-type blind equalization schemes. A complex-valued activation function composed of two real functions is used. Each real activation function has a multi-saturated output region in order to deal with QAM signals of any constellation size. Also, the complex backpropagation algorithm is modified for the proposed scheme. Computer simulations are performed to compare the proposed scheme with the conventional “stop-and-go” algorithm in terms of the convergence speed, the MSE value in the steady state and the constellation of QAM signals after the initial convergence. Simulation results demonstrate the effectiveness of the proposed scheme