Rejection of narrowband interference in PN spread-spectrum systems using neural networks

A multilayer back-propagation perceptron model is presented as a means of detecting a wideband signal in the presence of narrowband jammers and additive white noise. The performance of the neural network is compared with that of the estimation-type filter, which uses a least-mean-squared (LMS) adaptive filter, in terms of the interference rejection (notching) capability, the bit error probability, and the overall robustness of the system. The nonlinear neural network filter is shown to offer a faster convergence rate and overall better performance than the LMS Widrow-Hoff filter.<
>