Using Target Range Rate Data in Distributed Data Fusion Systems

Current data fusion architectures are evolving from single-platform, standalone systems to multi-platform, integrated systems. By taking advantage of the favorable geometry of a distributed data fusion system, target tracking performance can be greatly increased. Sensors can typically provide an estimation of a target's position, but only the radial component of the target's velocity. The target's velocity is typically obtained by tracking a target over time and estimating both position and velocity by putting the correlated sensor reports through a kalman filter. This filtering can take on the order of minutes (depending on sensor update rates) in which time the target may have maneuvered, possibly causing the filtering process to restart. If the range rate information from multiple platforms is combined, target velocity can be estimated quicker and with greater accuracy. This velocity information can then be fed back to the individual platforms, allowing smaller correlation gates which will enable tracking in higher density scenarios. The drawback to this approach is an increase in the amount of data that needs to be communicated among the platforms. This paper considers the performance implications of four different multi-platform architectures: sharing the best track, fusing all of the tracks, fusing all of the sensor reports, and fusing sensor reports that contain range rate information. The metrics that will analyzed are: track initiation time, velocity accuracy, and communication message bandwidth.