On-line multi-sensor registration for data fusion on airport surface

A block-processing methodology to dynamically estimate and cancel systematic multiple sensor errors is proposed and analyzed. Bias estimators are obtained from sequential consecutive blocks of measurement differences, achieving an accuracy equivalent to that reachable with an extended Kalman filter (EKF) operating over each pair, but with improved computational efficiency. The algorithm is formulated generically first and then applied to solve practical problems identified when fusing data on airport surfaces: global calibration, clock shift and local systematic deviations of sensor references. Performance, with significant improvement in tracking accuracy, is illustrated in simulated representative airport scenarios, fusing data from surface movement radar (SMR) and cooperative sensors.