Abstract
A comprehensive study of nonlinear statistical optimum adaptive signal filtering and detection using backpropagation (BP) neural networks is reported. It is shown that the BP neural networks can form nonlinear least mean square adaptive filters and minimum-error-probability adaptive signal detectors. Several variations and extensions of the optimum processors are made. In order to accelerate the convergence of the training, a class of training algorithms for the BP with optimized step-size is introduced. Numerical results are compared with the conventional linear processing methods. >
Full Text
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call