Model-based towed array processor: Experimental results
Edmund J. Sullivan, Dieter Brecht, Leif Persson · The Journal of the Acoustical Society of America · 2000
Results of a towed array bearing estimation experiment using a model-based processor are presented. The experiment took place in the southern Baltic using an Atlas towed array. The towship was the Swedish corvette Gavle. Results for the case of a 121-Hz source are presented. The towship ran in a straight line, past the source, where the distance of closest approach was 500 m. A short overview of the model-based approach, which is based on a state space formulation using a Kalman estimator, will be presented. By formulating the problem in state space form, it is possible to include sophisticated models in the processor. This, in effect, is a self-consistent means of including a priori information into the processing. It will be shown that the inclusion of the array’s forward motion provides significant improvement in performance over the conventional processor in terms of the variance of the estimate. The improvement in performance arises from the bearing information contained in the forward motion of the array. This information obtains, from the time scale, change in the signal or, in the narrrow-band case, the Doppler. The experimental results clearly show the improvement over the conventional processor.