Unified Control-Theoretic Bayesian Multi-Sensor Multi-Target Sensor Resource Management
Department of Defense
Agency Tracking Number:
Solicitation Topic Code:
Small Business Information
SCIENTIFIC SYSTEMS CO., INC.
500 West Cummings Park - Ste 3000, Woburn, MA, 01801
Socially and Economically Disadvantaged:
AbstractOptimal multisensor-multitarget sensor management poses a daunting theoretical and practical challenge. Scientific Systems Company, Inc. of Woburn MA and its subcontractor, Lockheed Martin Tactical Systems (LMTS) of Eagan MN, propose a foundational, joint control-theoretic approach to multisensor, multitarget sensor management. It is based on three innovations: (1) a unified foundation for multisensor-multitarget data fusion based on the multisensor-multitarget recursive Bayes filter; (2) approximations of this usually intractable filter, namely first-order multitarget moment filters and multi-hypothesis correlator (MHC) filters; and (3) integration of these filters with an intuitively meaningful, probabilistically rigorous, and potentially tractable objective functions: the POSTERIOR EXPECTED NUMBER OF TARGETS (PENT), and its extension the POSTERIOR EXPECTED NUMBER OF TARGETS OF INTEREST (PENTI). For PENTI, sensors are optimally directed to preferentially collect observations from targets of current or potential significance. Phase I showed the feasibility assessment for sensor management based on PENT & PENTI and demonstrated (1) tractable objective functions, (2) closed-form formulas for these objective functions, and (3) excellent sensor management performance for simulated multi-target multi-sensor scenarios. The primary focus of the Phase II will be to design and develop advanced approaches for sensor management capable of encompassing a very broad range of applications and emphasize the development of a prototype integrated sensor management algorithm to be tested using realistic multisensor data. Specific Phase II tasks are: (1) Extend PENT to the Multisensor Multistep Setting, (2) Extend PENTI to the Multisensor Multistep Setting, (3) Implement and Test with Different Trackers, (4) Derive/Implement Performance Metrics with Target Preference, (5) Extend the Significance Functions for Targets of Interests, (6) Characterize the Effectiveness, Strengths, and Limitations of the Algorithms, and (7) Phase III, Transfer of Technology, and Commercialization. The project team includes Dr. Ronald Mahler of Lockheed Martin. Lockheed Martin will provide both technical and commercialization support in the application of sensor management technologies.
* information listed above is at the time of submission.