Through-wall Radar Imaging Pattern And Algorithm Research

By utilizing the strong penetration of EM waves in barriers such as building walls,as well as exploiting the UWB signal for high range resolution and the large aperture forhigh cross-range resolution, Through-the-Wall Radar is developed to implement thelocating, tracking and identifying of the targets behind barriers, which can be widelyused in military and civilian areas such as urban warfare, counter-terrorism and disasterrelief. As a major part of Through-the-Wall Radar, Through-the-Wall Radar Imaging canachieve human position and scene layout detection, and has become the hotspot ofconcealed targets detection technology.The dissertation focuses on the problem of high resolution imaging under thecondition of low SCNR, and studies the Through-the-Wall Radar imaging algorithmbased on multi-channel and multi-view respectively. It mainly comprises:1. Based on the characteristics of electromagnetic waves’ transmission, an echomodel of TWRI is formulated to cope with the closed detecting environment of complexarchitecture, and to analyze the effects of walls on the reflection, refraction, attenuationof signals and the generative mechanism of multipath echoes, which can providetheoretic basis for compensation and multipath suppression.2. Based on the echo model, an improved BP imaging algorithm is used in TWRIwhich can achieve better focusing through using wall cancellation and compensation.3. The theory of Time Reversal Mirror (TRM) in TWRI is described withsimplified Greens function; in addition, the simulation is designed to prove itsfeasibility in imaging in complex multipath as well as its superior focusing performancecompared with BP algorithm.4. Cross-correlation Hyperbola BP algorithm based on multi-channel is proposedwhich can effectively improve the azimuth resolution compared with traditional BPalgorithm and can further improve imaging quality by fusion with traditional BPalgorithm imaging.5. Based on random sparse array, an optimal layout of array is designed and animaging fusion algorithm based on M-N-K detector are proposed which can implement target detection under low SCNR condition effectively and image the building layoutsuccessfully.Simulations and real data are presented to verify the feasibility and validity of themethods above in the practical applications of TWRI, which can also support the futureresearch.