Study On Signal Fusion-Based Target Detection In Multistatic Radar System

Compared with the monostatic radar,the multistatic radar system(MSRS)can observe a target with different angles,frequency bands or polarization modes,which causes great advantages and potential in electronic anti-reconnaissance,jamming proofing and weak target detection.Information fusion is the key to improve the performances of MSRS.According to the fusion level,the information fusion can be divided into three classes,e.g.decision,feature,and signal fusion.Because the decision and feature fusion work with a great information loss,they cannot show satisfactory performances with MSRS.Signal fusion-based methods can directly process echo data and provide weak information that the other two methods do not have,so they have significant advantages in weak target detection.With the development of communication technology,signal fusion-based methods have received a great attention.In this article,the signal fusion-based target detection using MSRS is considered.The main content of this dissertation is summarized as follows.1.A general signal fusion-based target detection framework is studied for MSRS.The traditional grid search-based methods work with a great computational complexity.Thus,elliptic spatial resolution cell(SRC)-based method is proposed,which includes space division methods,data association methods and fusion rules.The size of the elliptic SRC approximates the physical SRC,so the common surveillance region can be covered with the elliptic SRCs efficiently.Simulation experiments verify the effectiveness of the proposed framework.2.The multi-target detection problem is studied in the background of clutter,and a joint moving target detection and ghost target suppression method is proposed.Based on the signal fusion-based framework,a fusion rule is designed to suppress ghost targets and detect physical targets simultaneously,so that CLEAN algorithm can be avoided and the computational complexity can be reduced.The problem of moving target detection and ghost target suppression is built as a ternary hypothesis test and solved with two steps.First,a background discriminator is designed to distinguish ghost targets and clutter.The discriminator is based on the statistical characteristics of ghost target and clutter,that is,the observation data of ghost target contain the target and clutter components.The observation data of clutter only contain local clutter observation data.Then,two moving target detectors were developed under the ghost target and clutter background respectively.These detectors were selected according to the background discrimination.Experimental results show that the proposed algorithm can simultaneously detect targets and suppress ghost target.3.The joint adaptive target detection,position and velocity estimation is studied in the background of clutter,which can solve the problem of straddle loss.The problem of joint target detection and parameter estimation is built as compound binary hypothesis testing,where the clutter covariance matrix,target amplitude,target position and velocity are assumed to be unknown.First,an adaptive detector/estimator was designed based on GLRT and AMF frameworks respectively.After the estimated covariance matrix and target amplitude are substituted into the likelihood ratio,the detection statistics is transformed to a function that is maximized by position and velocity.Then,in order to solve the optimization problem,this chapter proposes two different optimization strategies,i.e.,direct estimation method and mixed estimation method.If the detection statistic is higher than the threshold,the precise estimates of target position and velocity will be output as a by-product.Finally,the performances of the proposed adaptive frameworks are analyzed from the aspects of detection performance and estimation accuracy respectively.The results show that the proposed methods perform better than "detection then estimation" methods.4.The problem of target detection in the presence of active deception jamming is studied,and a joint target detection and jamming suppression is proposed.First,the statistical differences between the target and the jamming observation data are analyzed.Then,according to the differences of statistical characteristics,SRCs are divided into the target dominant SRCs,target relative SRCs,deception jamming dominant SRCs,deception jamming relative SRCs and pure noise SRCs.Target relative SRCs,deception jamming dominant SRCs,deception jamming relative SRCs and pure noise SRCs are considered as the interference.Then,a two-step detection algorithm is designed to distinguish target dominant SRCs,deception jamming dominant SRCs and pure noise SRCs.Then,after clarifying the dominant SRCs type,interference cancellation is developed based on CLEAN algorithm.Finally,the final decision is made based on the non-coherent detector.The results show that the proposed method can simultaneously suppress deception jamming and detect target.