Blind Extraction Array Radar Clutter Suppression Of The Simulation Study

Radar is widely applied in the fields of meteorology, agriculture, biomedical engineering and military. To improve the probability of detection and to reduce the false alarm rate, many advanced theories and methods have been proposed, clutter suppression is one of the most critical technologies. The Radar clutter can be classified into two types: homogeneous clutter and inhomogeneous clutter according to the statistical characteristic. Homogeneous clutter can be suppressed well through conventional space-time adaptive processing (STAP) methods, but there has not been a method to suppress the inhomogeneous clutter ideally. This dissertation concentrates in the technology of clutter suppression in inhomogeneous clutter environment and does the following researches:(1) Analyzing the composition of the Radar echo and concentrating in the impact and challenge brought by the inhomogeneous clutter. Then four different kinds of inhomogeneous clutter are simulated and generated using Zero Memory Non-linear (ZMNL) method. The four different kinds of clutter are: Rayleigh, Log-normal, K and Weibull. A method is proposed to spread the 1-dimension clutter to 2-dimension clutter.(2) Introducing the research of extracting the target echo with the blind source extraction (BSE) method. After summarizing the researching dynamic of BSE, the second-order BSE method proposed by W. Liu which is based on linear predictor is specially introduced. W. Liu's algorithm is adopted to extract the target echo. We can learn form the simulation results that the correlation coefficient between the extracted target echo and the initial target echo is close to 1.(3) A new method of Radar clutter suppression is proposed. This method combines BSE and joint-domain localized generalized likelihood ratio detection (JDL-GLR) together. The specific steps includes: converting the 2-dimension signal that contains target echo, inhomogeneous clutter and noise into 1-dimension vector;extracting the target echo using BSE method; processing the extracted target echo in a imitative Gaussian clutter environment using the JDL-GLR algorithm. We can learn form the simulation results that this assembly processing method can attain an ideal result, when the SINR of the imitative environment is above -10 dB. The feasibility and practicability is affirmed by the simulation result.This research has important reference values in the process of exploring new Radar signal processing method in homogeneous clutter environment.