Efficient feature extraction from time-frequency analysis of transient response for target identification

In this letter, we present a new feature extraction method, for radar target identification, that use both an early-time and a late-time response simultaneously. A time-frequency analysis is used for feature extraction from the transient target response, and the dimension of feature vectors is reduced by a novel method that we propose. Simulation for target identification is performed with three different swept-wing aircraft models. Results show that the proposed feature extraction method has good identification performance in the presence of white Gaussian noise. (C) 2000 John Wiley & Sons, Inc.