Evaluation of a Multi-Sensor Acoustic Tracking Scheme with Realistic Data
Abstract
This report summarizes the evaluation of a tracking algorithm on real data. RMC developed this passive sonobuoy tracking algorithm over the last two years. The algorithm uses the conditional density functions to evaluate the possible location for targets and to predict its future position. This differs from algorithms which try to estimate the source location directly; we are able to localize gradually by eliminating sectors when possible. The method can automate many functions that a human operator might perform, leading to a more effective man-machine interaction. Evaluation results indicate that the algorithm works well with the DREA data, consisting of both narrowband Doppler-shifted frequency and bearing measurements. On realistic data, the source is estimated to an accuracy of 1-2 km when the source is enclosed by the sensor array. Outside of the array, performance is reduced, in part by the lessened resolving power of measurement and also due to algorithmic difficulties, on which research is on-going. The next step is to incorporate this algorithm into a scheme that uses broadband data as well.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1995
- Accession Number
- AD1001533
Entities
People
- J J Towers
- Y. T. Chan
Organizations
- Royal Military College of Canada