To solve the problem of classifying the aircrafts using high resolution radar range profiles, the product spectrum which was used in the speech signal processing community was introduced to the radar target recognition community. The product spectrum was defined as the product of the power spectrum and the group delay, this feature combined the information contained in the magnitude spectrum and the phase spectrum of the range profiles and carried more information about the shape of the aircrafts. A multi-layered feed-forward neural network with resilient propagation (RPROP) algorithm was selected as classifier. The range profiles were obtained by step-frequency technique using the two-dimension backscatters distribution data of four different aircraft models. Simulations were presented to evaluate the classification performance with the product spectrum features. The results show that the product spectrum based features can yield good performance even in noisy conditions for the application of radar target recognition.
Oppenheim Alan V, Schafer Ronald W. Digital signal processing . Englewood Cliffs:Prentice-Hall, 1975
Zhu Donglai, Paliwal K K. Product of power spectrum and group delay function for speech recognition . ICASSP . Montreal Quebec Canada:IEEE,2004. I-125~I-128
Riedmiller M, Braun H. A direct adaptive method for faster backpropagation learning:the RPROP algorithm . Proceedings of the IEEE International Conference on Neural Networks . San Francisco:IEEE,1993. 586~591