Study On Wall Clutter Mitigation Techniques For Through-the-wall Imaging Radar

Through-the-wall imaging radar(TWIR) can be used to realize the detection of target behind nonmetal wall based on the penetration characteristic of electromagnetic wave. With various broadband signals and antenna configuration, it can image and localise both moving and static target, which has important applications in military, anti-terrorism and rescuing missions.During the propagation, only part of the signal will penetrate the wall. The left which takes up the major part of signal energy will return directly and thus form the high instantaneous power direct-wave interference. On the other hand, the multiple reflections of signal inside the medium lead to wall ringing interference which has long time duration. All of these wall clutters seriously interfere with the weak target reflection in the echo and hinder target detection.The article first analyses the imapct of wall on signal propagation and the generation mechanism of wall clutters. Then a brief introduction of basic theory of through-the-wall imaging and wall clutters mitigation techniques is presented. The performance of various wall clutters mitigation methods in different transmit/receive models are demonstrated through simulation analysis.Based on the analysis above, Independent component analysis(ICA) algorithm is applied to suppress the direct-wave interference. In order to deal with the lack of observation signal caused by the sparsity of transmit/receive channels, the article propose to apply ICA on multiple down-range observations of each transmit/receive channel to remove wall clutter. The simulation and experimental results show that the new method can separate target and clutter components effectively.At last, the paper proposes to eliminate the residual interference based on the sparse blind deconvoluion(SBD). The general signal model of blind deconvoluion is introduced. The theory of sparse blind deconvoluion and the sparsity of echo are studied emphatically. Based on the discussion above, a SBD algorithm which combines parametric modeling and sparse decomposition is presented. To improve the performance of parameter optimization in global manner, a local optimization strategy is proposed. The effectiveness of different SBD algorithm is verified through simulation analysis.