Tracking with Time-Delayed Data in Multisensor Systems
Abstract
When techniques for target tracking are expanded to make use of multiple sensors in a multiplatform system, the possibility of time delayed data becomes a reality. When a discrete-time Kalman filter is applied and some of the data entering the filter are delayed, proper processing of these late data is a necessity for obtaining an optimal estimate of a target's state. If this problem is not given special care, the quality of the state estimates can be degraded relative to that quality provided by a single sensor. A negative-time update technique is developed using the criteria of minimum mean-square error (MMSE) under the constraint that only the results of the most recent update are saved. The performance of the MMSE technique is compared to that of the ad hoc approach employed in the Cooperative Engagement Capabilities (CEC) system for processing data from multiple platforms. It was discovered that the MMSE technique is a stable solution to the negative-time update problem, while the CEC technique was found to be less than desirable when used with filters designed for tracking highly maneuvering targets at relatively low data rates. The MMSE negative-time update technique was found to be a superior alternative to the existing CEC negative-time update technique. Multisensor systems, Kalman filter, Cooperative Engagement Capabilities (CEC) technique, Minimum Mean-square Error (MMSE) technique.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1993
- Accession Number
- ADA271334
Entities
People
- David A. Martin
- Richard D. Hilton
- W.D. Blair
Organizations
- Naval Surface Warfare Center