Fractal and Wavelet Signal Processing Techniques for Radar Guided Missiles

In this project FASTMAN will determine the feasibility of using fractal and wavelet signal preprocessing in conjunction with neural network pattern classifiers on high-range resolution radar data for detecting cold stationary ground targets in a clutter-rich environment. Previous researchers have demonstrated that measurement of the fractal dimension of high-range resolution (HRR) returns can be used to distinquish stationary targets in clutter rich environments. Other groups of researchers have shown that the Wavelet Transform can be used to accurately and efficiently measure the fractal dimensions of signals by examining the behavior of the transformed signals across scales. We propose to marry these two concepts into a signal preprocessor in which the Wavelet Transform will compute the local fractal dimensions of HRR returns, and those fractal dimensions will form a feature vector which will be used by neural network pattern classifiers to distinquish targets in the backscattered signal. We will also investigate the applicability of two other Wavelet processing techniques for radar signal processing: the use of the Adaptive Wavelet Transform to extract key target features and the use of doppler-tolerant wavelet radar pulses.