Radial basis function neural network for pulse radar detection

摘要: A new approach using a radial basis function network (RBFN) for pulse compression is proposed. In the study, networks 13-element Barker code, 35-element code and 21-bit optimal sequences have been implemented. training these networks, RBFN-based learning algorithm was used. Simulation results show that RBFN has significant improvement in error convergence speed (very low error), superior signal-to-sidelobe ratios, good noise rejection performance, improved misalignment range resolution ability Doppler shift performance compared to other neural approaches such as back-propagation, extended Kalman filter autocorrelation based algorithms. The proposed provides robust mean radar tracking.