Studies On Track Fusion In Distributed Passive Radar Networking

Being different from the traditional active radar,passive radar does not radiate electromagnetic wave by itself,but does detection,location and tracking with the help of the third part transmitter.Passive radar system has great power of anti-stealth and in avoiding being attacked by anti-radiation missile,resulting in the strong ability to survive,as well as it has simple structure and low cost,many researchers pay attention to it recently.Widely used passive radar is limited to being equipped with only a receiver,thus has small detection range and low precision with the influence of the geography or the collaborative signal.Drawing lessons from multi-station network detection technology,distributed architecture is the breakthrough point to improve the detection range and precision for passive radar,and furthermore,track fusion is the key to determine the target tracking performance.The paper makes a study of the track fusion for distributed passive radar system.The problems of track association and track fusion based on state estimations are emphasized on in order.Firstly,the framework and improvement performance for track fusion are introduced from the point of the view of data fusion,and the key issues for the application of distributed fusion framework to passive radar system are analyzed in detail.Next,various methods,such as the weighting method,the modified method,the sequential method and the double-threshold method,are discussed as effective ways for solving the track association problem.As the symbols to evaluate the associated performance of different methods mentioned above,possibilities of correct association and error association are compared quantificationally at settings of different target numbers.As for track fusion based on state estimations,algorithms with memory and without memory which are classified according to whether the predicted information is used are derived,and suitable circumstances for various methods are discussed theoretically.Last but not least,the improved decorrelated-distributed Kalman filter(IDDKF)is presented to satisfy the characteristic of time-varying location error existed in distributer passive radar system.This algorithm does not only emit the communication architecture among receivers,but also gives the optimal estimation at fusion moment.The fusion performance for different track fusionalgorithms including the IDDKF is provided analytically based on simulations and actual measurement data.