Random Matrix Method For Target Detection In Spaceborne Cluster Flight Netted Radar

Space-based cluster flight netted radar integrates the advantages of cluster flight spacecraft and space-based netted radar and has a series of features such as convenient spacecraft orbit control and relative position perception,high sensitivity,high survivability,high reliability,and strong anti-stealth ability.However,the time-varying and uncertainty of the geometry of space-based netted radar brings great challenges to the target detection performance.Therefore,the paper carries out the research on the random matrix of target detection of space-based netted radar and discusses the relationship between radar geometry and target detection performance,in order to solve some problems of radar geometry design and provide theoretical and technical support for engineering practice.The main work of this paper is as follows:(1)Based on the dual-satellite companion flight mode,a space-based bistatic radar pair movement model is constructed,the radar ambiguity function,range resolution formula and velocity resolution formula are derived,and the upper bound of range resolution and velocity resolution is proved.The MCS-HA algorithm of HA fusion MCS is proposed,and the matching optimization conditions of radar pairs are obtained.The conclusion is proved by numerical calculation and analysis.(2)Analyze the robustness of target detection.The signal model of space-based cluster flight netted radar is constructed,and the covariance matrix of target detection is derived.Through the analysis of the correlation channel gain matrix,the basic conclusion that the target detection performance depends on the geometry of space-based cluster flight netted radar is obtained.Using the MME detection method,the false alarm probability and missed detection probability of the target are defined,and the constraint conditions of the robustness of the MME detection are calculated,and then the bounded relative distance satisfying the robust MME detection is analyzed.The validity of the constraint conditions is proved by numerical calculation and analysis.(3)Target Detection Algorithm Based on Random Matrix Spectrum Theory.Based on the proposed new concept of space-based cluster flight netted radar and geometrical configuration characteristics,the target detection problem caused by the uncertainty of the geometrical configuration of space-based cluster flight grouping radar is investigated by applying the random matrix theory.Firstly,the asymptotic and non-asymptotic spectral theories of the random matrix are analyzed.Secondly,two different target detection algorithms are proposed for the asymptotic and non-asymptotic spectral theories of random matrices for space-based cluster flight grouping radar geometrical configuration uncertainties.Finally,the two algorithms are numerically analyzed.(4)Target detection with application of Gerschgorin circle theorem.For the problem of difficulty in obtaining CFAR detection algorithms for multi-static or netted radars,the space-based cluster flight grouping network target detection problem is investigated by applying the K out of N fusion criterion.By analyzing the known,time-varying channels of space-based cluster flight netted radars,the uncertainty detector of the relevant space-based cluster flight netted radar geometry is transformed into a noise uncertainty detector,which lays the foundation for designing a CFAR detector for bi-static radars;based on Gerschgorin circle theorem,a target detection algorithm for space-based bi-static radars is proposed.A new concept of false-verdict probability,which combines the probability of missed detection and false alarm,and a method to minimize the false-verdict probability are proposed for space-based cluster flight netted radar target detection by using a weighted summation method.Numerical calculations verify the validity and reasonableness of the proposed method.