Study And Realization Of Multipath Data Association For Sky-wave Over-the-horizon Radar

Beyond-the-horizon detection of targets at ranges of thousands of kilometers can be achieved by Sky-wave Over-the-horizon Radar (OTHR) operating in the high-frequency band. This very long coverage is obtained by using sky-wave propagation, which is, reflecting the radar signal from the ionosphere. The most fundamental difference between sky-wave radar and other radars is the existence of multiple propagation paths, with distinct time delay, angles of incidence, doppler shifts and fluctuation properties. The tracking stage must deal with the multiplicity of echoes associated with each individual target and, by extracting and assimilating information about the prevailing ionosphere, infer the number of genuine targets, their true locations and velocities. This dissertation concentrates on track clustering based data association for Sky-wave OTHR. The main contribution can be summarized as follows:1. The planar model of transformation between radar coordinates and geographical coordinates is deduced for Sky-wave OTHR based on given ionosphere states. Converting from radar coordinates to geographical coordinates is referred to as Coordinate Registration (CR). The problem of target tracking can be integrated with the problem of determing the ionosphereic propagation paths. By establishing tracks in geographical coordinates with additional parameters that characterize the propagation path structure, the multipath problem can be formulated and solved.2. A new method of Sky-wave OTHR data association is proposed. The measurements observed from radar are transformed to geographical coordinates considering all possible propagation modes, and the transformed points in geographical coordinates for each measurement are called hypothetic points. Clustering algorithm is adopted to divide these hypothetic points into multi-classes. The feasible classes are selected using the restriction of Doppler and propagation modes. The geometry center of clustering can be calculated by averaging the points of feasible classes. The multiple hypothesis method is used to achieve the correlation of measurements, propagation modes and targets among different scans. The final output of tracks is obtained by the filtering process.3. Software design of Sky-wave OTHR data processing is implemented with VC++ environment. The problem of track association between adjacent detection regions is solved by using reference topology method. The flowchart of data processing is introduced and the human-machine interface is presented.