Radar Pulse Compression Sidelobe Suppression In The Phase Encoding Signal

Pulse compression is an advanced technique applied widely in modem radars. Especially binary phase-coded sequences are widely used with the properties that their periodic or aperiodic auto-correlation function(ACF) has an impulse-like form, and that they can be easily generated and processed. However, their range sidelobes are too high and inadequate for most applications. So, how to minimize the peak range sidelobes of a binary phase-coded waveform is a key to practical applications of phase-coded pulse compression. To begin with, the introduction of the thesis summarizes technique background and recent developments in phase-coded pulse compression and generalizes main approaches used usually. Then the main principle of pulse compression technique and a simple method of sidelobe suppression are introduced. Next, the capacity of pattern classification and learning algorithm of neural networks are stated and a known approach of neural networks applied in range sidelobe suppression is showed. At the same time, a modified learning algorithm of neural networks is presented and better experiment results based on the learning algorithm are obtained. Besides, the concept of time-frequency localization and multi-scale analysis of wavelets is simply introduced and a method that wavelet transforms are utilized in sidelobe suppression networks is proposed and experiment results of some kinds of wavelet bases are discussed. Furthermore, the architecture and learning algorithm of wavelet neural networks are studied, a wavelet function similar to Haar wavelet function, called Pseudo-Haar Wavelet, is presented and a new method that wavelet neural networks based on the wavelet function are used to suppress range sidelobes is explored. In the end, the thesis also predicts the possibility of fuzzy systems employed in range sidelobe suppression.