Study Of Bistatic Through Wall SAR Image Formation

When applied in the field of through-wall detection, the high resolution synthetic aperture radar (SAR) shows better performance and flexibility in high resolution building imaging, through-wall position and tracking, compared with other through-wall sensors such as fixed array and life detection radar. The bi- and multi-static SAR system, with its transmitter and receiver being deployed in different independent moving platform, is endowed with more functions than mono-static through-wall SAR and proved to be an important research direction for future through-wall radar. However, caused by the azimuth variance of the two platforms with different velocity vectors, problems arise in system work and imaging processing, which has become an attractive research issue in through-wall SAR detection theory.Focusing on signal processing theory for bistatic through wall SAR system with a fixed receiver, this thesis researches on the system characteristics and imaging methods, which can be useful in system design and configuration.1. The system characteristics of bistatic through wall SAR with a fixed receiver are studied. Based on geometrical and signal model, bistatic slant range and its azimuth variance are studied. The Doppler and synthetic aperture characteristic are deduced, two dimensional (2D) resolution and the influence of wall on them are analyzed by simulation. The influences of permittivity of wall on slant range, Doppler FM rate and resolution are shown to be relatively significant. Due to the azimuth variance, echoes with the same Doppler FM rate are in different range gates in data domain, makings imaging processing more difficult. The simulation of 2D resolution distribution under different configurations indicated by bistatic angle illustrates that the resolution gets worse as the angle increases, which can be a theoretical reference for practical system configuration.2. Based on the through wall propagation model, a wall parameter estimation method using time delay extraction from range profiles is proposed, which is solved numerically or by searching curve combined with transcendental equation. According to the electromagnetic theory, an estimation approach based on correlation of echo envelope is proposed, which can be implemented using echo matrix or echo vector. Taking the time delay and amplitude of echo varies with carrier frequency into consideration, a dual- carrier frequency method combined with extraction of time delay and amplitude is invented. It can give analytical expressions of the estimations. The precision of the proposed methods can achieve centimeter for wall thickness and 1 for relative permittivity.3. The Back-Projection (BP) algorithm under in through wall SAR with a fixed receiver is studied. The over all processing efficiency of algorithm in frequency domain and that in temporal domain are compared in the researched scenario. It indicates BP algorithm is still a good choice. Images of a building before and after wall compensation are simulated under different bistatic configurations.4. A modified RD algorithm is proposed, which compensates receiver range and range cell migration in temporal domain simultaneously. The compressed echoes are recomposed using a mapping procedure, therefore, the azimuth variance caused by the stationary receiver is removed, and geometry correction is not required. It is capable of imaging of a building with sparse-structure inner. Its advantage in efficiency is verified by analyses and simulation.5. The 2D Non-Linear Chirp Scaling (NLCS) algorithm of bistatic through wall SAR with a fixed receiver is studied. Based on the azimuth temporal phase in the derived 2D point target spectrum for this configuration, a new perturbation function is proposed. The data are divided into apertures in azimuth, reducing the negative effects . The proposed azimuth processing method is compared with the classical ones by deduction and simulation of targets in a room.6. Geometry correction, image registration and fusion of bistatic through wall SAR are investigated. Bi- and multi-static configuration for building imaging with aspect angle diversity is proposed. Correction algorithm for image geometric distortion induced by bistatic system and effects from the wall is provided. Based on the extraction of edge line feature of wall and control point on geometrical center of target in image, the bistatic images registration method of different platforms is achieved. To depict the building layout, several image fusion methods at different pixel level are proposed, generating images of an entire building which can be used for building structure reversion.The research results achieved in this paper such as models and imaging processing procedures can be valuable for the system design and real data imaging in bistatic through wall SAR.