We propose a novel method using impulse radar to reduce computer resources
and improve the accuracy of reconstructed signals for correlative target re
cognition. Our method employs discrete wavelet transform and principal comp
onent analysis. Simulation results show that the proposed method can improv
e the accuracy of reconstructed signals and also reduce data storage consid
erably. We also show that the accuracy of the reconstructed signals enhance
s the performance of correlative target recognition.