A Study On Netted Radar Data Association And Fusion Method In Complex Environment

On the basis of the traditional three-dimensional space of sea,land and air,electromagnetic and aerospace have been added to the scope of modern battlefield,forming a five-dimensional battlefield space.The anti-jamming ability of a single radar is not enough to cope with various kinds of complex electromagnetic interference environment in international war effectively.It has become an inevitable trend to expand the radar’s capabilities in all aspects by networking the radar.With the fidelity of various deceptive jamming signals going up,it is increasingly difficult to identify interference at the signal level.It is inevitable that deceptive signals may enter the data processing layer through the discrimination of the signal layer,so it needs to be discriminated at the data level.In this paper,based on the complex jamming environment that the netted radar may face in the target tracking,the methods of data association and data fusion suitable for the specific interference environment are studied.The purpose of the paper is to identify and counteract the corresponding interference at the level of data processing.The proposed algorithm is verified by simulation combined with the actual scene model.The main contents of this paper are organized as follows:1.The data association and fusion algorithms commonly used in network radar are introduced and analyzed.Firstly,according to the different processing objects,the data association algorithm is divided into point-to-point association,point-to-track association and track-to-track association algorithms.The classification and main applications of the association algorithms are introduced respectively.Secondly,data fusion algorithms are introduced,including centralized and distributed fusion algorithms for location-level fusion and D-S(Dempster-Shafer)theory fusion methods for decision-level fusion.Then the advantages and disadvantages of various fusion algorithms and their applicable scenarios are analyzed.2.The non-cooperative interference countermeasure based on data association and fusion is studied.For the interference of multiple false targets,on the basis of predecessors,considering that the same target will be tracked by each node radar,there will be the same process noise.Therefore,this paper proposes a modified track association method to combat the range multi-false-target jamming for the active/passive radar system deployed in different places.The simulation results show that under different range delays,especially when the range between multiple false targets is small,the modified method can effectively reduce the misjudgment probability of false targets on the basis of guaranteeing the correct probability of physical targets,compared with the method used by predecessor.Similarly,in the heterogeneous active/passive radar network scenario,this paper proposes a method of MCI(Modified Covariance Intersection)track fusion method to resist the range gate pull off/in jamming at the data level.The method modifies the traditional CI(Covariance Intersection)fusion algorithm and can guarantee the consistency of fusion estimation.When the fusion center uses this algorithm for track fusion,the fusion weight of the disturbed radar’s false track can be reduced,and we can obtain the system track with reliable track quality.The validity and reliability of the proposed method are verified in RGPO(Range Gate Pull-Off)and RGPI(Range Gate Pull-In)scenarios respectively.3.The cooperative jamming countermeasure based on data association and fusion is studied.Firstly,the jammer will introduce additional observation noise when trajectory planning is performed.According to the above condition,innovative cumulative distance characteristic of single radar is constructed.Secondly,the paper analyzes that there will be radar site detection errors during the pre-surveillance process of ECAV(Electronic Combat Air Vehicle),and ECAV preset position errors during cooperative control.In view of this,the overall correlation distance characteristic of radar network is constructed.Finally,according to the D-S evidence theory,multiple characteristics are converted into basic probability assignment through membership functions.And data fusion from the first-level to the third-level is carried out until the decision-making method is satisfied,then the identification of real and phantom tracks is completed.In this paper,four simulation experiments are performed to verify the proposed method.The simulation results show that the method can effectively identify real and false target tracks,which has certain theoretical guidance significance for engineering applications.