Accession Number:

AD1104524

Title:

Optimal Sensor Tasking for Space Situational Awareness

Descriptive Note:

Technical Report,15 Jul 2015,14 Jul 2019

Corporate Author:

RESEARCH FOUNDATION OF STATE UNIVERSITY OF NEW YORK THE BUFFALO United States

Personal Author(s):

Report Date:

2020-01-31

Pagination or Media Count:

28.0

Abstract:

This report documents the major project findings during the duration of the AFOSR Award titled Optimal Sensor Tasking for EnhancedSpace Situational Awareness. Continuing work on Gaussian mixture model GMM, we have developed an adaptive mechanism to automatically select the architecture of the Gaussian mixture model. The developed method is the only method in the literature which constrained the GMM approximation to satisfy the Fokker-Planck-Kolmogorov equation. We have developed computationally efficient semi-analytical approaches for uncertainty propagation while making use of tools from convex optimization. The CUT algorithm has been used to derive information theoretic sensor tasking framework. Our work clearly shows the benefit of using mutual information as a tasking metric as opposed to Fisher information generally used in the literature. Also with the Conjugate Unscented Transform for uncertainty propagation, these approaches provide a computationally efficient framework to simultaneously task sensors and track multiple targets.

Subject Categories:

  • Miscellaneous Detection and Detectors
  • Military Operations, Strategy and Tactics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE