Application of neural network to the transient system correcting for improved target detection

A neural network with back-propagation algorithm is used to approximate the impulse function of the correcting filter in the time domain. The principle and simulated result of the neural network technique are introduced. The mean-square-error between the output of the neural network and the inverse impulse response of the system can be minimized by choosing appropriate network size, learning rate, and maximal number of iterations during the learning phase of the neural network. The distortion of the signal may be reduced by a correcting filter. When the received signal comprising the direct signal and the scatter passes through the net, the direct signal is compressed, and the distortion of the scatter may be reduced. >