Moving Target Depth Estimation for Passive Sonar, Using Sequential Resampling Techniques
Abstract
We examine the problem of passive localization of a moving target in a littoral environment, based on its depth and range-rate. We compare performance with the conventional matched field processor which localizes in depth and range. Range-rate localization is more robust with respect to uncertainties in the environment, and with respect to associated uncertainties in the horizontal wave numbers of the channel modes used for the matched field target response. In our approach the complex amplitudes of the modes are treated as nuisance parameters, which comprise a hidden, first-order Markov state process. In lieu of an analytic expression for the evolution of the likelihood function as new snapshots are integrate, we evaluate a method of particle filtering, or sequential resampling.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2000
- Accession Number
- ADA405125
Entities
People
- Jeffrey Krolik
- Shawn Kraut
Organizations
- Duke University