Track parameterization for an acoustic horizontal line array
Rockie L. Ricks · 2005
The Deployable Autonomous Distributed System (DADS) uses a field of bottom mounted, horizontal acoustic sensor arrays. Information is passed to a central node for field level tracking. To reduce vulnerability to trawling, DADS uses acoustic communication instead of cables between sensors. The acoustic communications have limited bandwidth. Therefore, the amount of acoustic track information must be reduced prior to transmission. Acoustic tracks are lists of cone bearing-time pairs. An acoustic track from a constant velocity target can be represented by four parameters, target course time of the closest point of approach (CPA), speed to CPA distance ratio, and depth difference to speed ratio. The depth difference to speed ratio cannot be estimated with significant accuracy from the DADS array, but a rough estimate can be found from the depth of the array and a guess of the target speed. In this paper, the uniqueness and observability of the other three parameters are discussed. The well-known left-right ambiguity is derived. Observability is used to explain difficulty in estimating parameters with data near CPA only. An unresolved issue is how to segment a track to eliminate target maneuvers and non-target measurements. Another issue is how to select the minimum track segment length. If track segments are shorter than the minimum, parameter estimation would not be attempted. The parameters are estimated through nonlinear optimization. A rough initialization for each parameter is given using readily estimated quantities. Using manually determined track segment intervals, the parameter estimation performance is given on four runs of sea test data from a 1999 exercise.