Radar Pulse Deinterleaving Method Based on Eidos BSB Artificial Neural Network
Maozhong Song · Modern Electronics Technique · 2010
A novel radar pulse deinterleaving method based on Eidos BSB(brain-state-in-a-box) artificial neural network is proposed.Simulation results indicate that radar pulse patterns can be stored and deinterleaved by the parameter optimized unsupervised auto-associative neural network after learning a mass of radar pulse samples with measurement errors.Comparing with other radar pulse deinterleaving methods,this method can adapt more measurement errors and achieve higher deinterleaving accuracy.