Range Side-lobe Suppressing Of Stepped-Frequency Ground Penetrating Radar

Stepped-Frequency Ground Penetrating Rada(rSFGPR)is a continuous wave GPR with good performance both in wide bandwidth and high dynamic range, which can realize high range resolution and deep penetrating detection. However, because of high range side-lobe coming from the truncated spectrum in the final synthetic range profile, weak responses of targets could be covered by side-lobe of radar system direct and coupling signal, which decreases the system capability of detecting and resolving targets, lowers dynamic range, increases false alarm rate. Nevertheless, few works have been done to suppress range side-lobe for SFGPR. There is no systematic theory support and effective suppressing algorithm. This thesis carries out a systematic research in the range side-lobe suppressing of SFPGR system.1,This thesis presents the formation mechanism of range side-lobes in SFGPR and basic principle of suppressing range side-lobe. Different existing algorithms for range side-lobe suppression are analyzed. Some classic linear weighting algorithms, time weighting and spectrum weighting, are compared. The experiment results show that the traditional linear weighting algorithm will widen main-lobe and deteriorate range resolution when suppressing range side-lobe.2,To improve the range resolution of the linear weighting, we propose to apply SVA(Spatially Variant Apodization) to suppress range side-lobe of SFGPR. In this algorithm, different spectral weighting functions for each sample point in time domain are used. It's a nonlinear weighting algorithm which can suppress side-lobe and also maintain main-lobe at the same time. The experiment results show the algorithm has the same range side-lobe suppression ratio (RSSR) compared with linear weighting algorithm, and widens main-lobe by only 5%.3,In order to have a good performance both in suppressing range side-lobe and high resolution, this thesis puts forward the super-resolution method of spectrum extrapolation to reduce range side-lobe. Maximum entropy extrapolation and minimum entropy extrapolation are presented. The range side-lob suppressing theory of maximum entropy and minimum entropy extrapolation are deduced in the thesis, and a new idea is proposed to combine entropy spectrum extrapolation with linear weighting to improve the RSSR. The simulation results indicate that the joint of maximum entropy extrapolation and linear weighting can improve RSSR by 30dB and maintain the main-lobe at the same time. Minimum entropy extrapolation can improve RSSR by 14dB and resolution by 50%.