TECHNOLOGY AREAS: Information Systems, Sensors

ACQUISITION PROGRAM: NAVSEA- Navy Electronic Warfare (EW) Programs (AN/BLQ-10, InTop)

OBJECTIVE: Research and develop advanced Data Fusion algorithms capable of robust in-situ adaptation based on environmental context which utilize information from on-board and off-board sensors to improve situational awareness in a real-time environment.

DESCRIPTION: Electronic Warfare systems aboard US Navy vessels were designed to contend with a very predictable threat environment. The problem of classifying an emitter as a threat, although not trivial, was fairly well understood. Over the last decade, the number and type of RF emitters in littoral environments has grown with increasing rapidity. This has highlighted shortcomings with current EW systems and drastically increased operator workload. In particular, an emerging problem for EW systems is the proliferation of radars that employ solid state amplifiers. This problem is further compounded by the use of wideband coded or chirped waveforms. Conventional Specific Emitter Identification (SEI) algorithms that target the Unintentional Modulation on Pulse (UMOP) characteristics of Travelling Wave Tube Amplifiers (TWTA) fail to discriminate between emitters of these types. As new SEI algorithms are developed which extract non-traditional features (e.g., higher order spectra) to contend with these emitter types, linear or quadratic discrimination techniques that are currently employed in EW systems cannot make use of the output from these multi-modal feature extractors, due to the fact that the resulting decision regions are highly non-linear.

The Navy seeks Adaptive Data Fusion Algorithms (based on Kernel Logistic Regression for example) that are capable of robustly adapting to improve detection, feature extraction, feature selection, and classification by creating and mapping multi-dimensional feature vectors to a non-linear vector space. In general, these algorithms should employ a multi-modal, multi-sensor (including off-board sensors and meta-data) fusion approach to weight features in terms of importance and relevance, depending on environmental context, to further minimize the probability of incorrect classification. Finally, these algorithms should incorporate machine learning, based on feedback (operator or classifier) to adapt to "feature drift" or new (unknown) emitters, by shifting or creating decision boundaries in the non-linear space.

PHASE I: Research and Development of an overall concept and detailed description based on simulated data of the decision space; how the algorithm(s) adapt; on what information the adaptation is based; an estimate of improvement in Pcc; and a simple demonstration.

PHASE II: Extend proof-of-concept algorithm from Phase I to robustly and substantially adapt in a laboratory environment. Evaluate performance using government provided data and develop specifications for transition to system insertion.

PHASE III: Transition the system into a production Navy system.

PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATION: These algorithms are applicable to the telecommunications industry as well as industries requiring surveys, searches, or mapping.

Multi-modal data fusion learning algorithms are sensor independent. They are applicable in any application where disparate information can be vectorized and weighted in such a way as to create a vector space in order to more accurately interpret (classify) real-time sensor data. The learning aspect makes them particularly suited to changing environments. For example, in the telecommunications industry, wireless network planning is highly dependent on the very dynamic electromagnetic environment (EME). Currently, test vans with collectors roam urban areas to attempt to characterize the EME in terms of detecting potential co-channel interference or areas of obscuration. This data is then manually processed and assessed to determine where to place new cell towers and repeaters. An algorithm capable of fusing collected information with other types of sensors (imagery, terrain maps, meteorological information, GPS, etc.) and is adaptable to the dynamic urban environment would be very useful to this industry in order to reduce 1) the search area of the van; 2) automate the classification of the co-channel interference (TV station, other cell tower, communications transmitter); 3) learn via the incorporation of a new training set to adapt to changes (frequency allocations, new communications infrastructure, etc.).

In general, industries where route planning, infrastructure planning and situational awareness are required, algorithms of this type can add value by reducing cost and improving business execution.

REFERENCES:
1. C. Pizzo, G. Powell, C. Brown, J. May, "Modeling and Simulation Support for Answering Commanders� Priority Intelligence Requirements", United States Army Command and Control Directorate (CERDEC), June 2005.

2. R. Wiley, The Analysis of Radar Signals, 2nd ed. London, U.K.: Artech House Press, 1993

3. "Higher-Order Spectral Analysis: A Nonlinear Signal Processing Framework" C. L. Nikias, A. P. Petropulu, Prentice Hall, Englewood Cliffs, NJ, USA (1993).

4. C. Bishop, Pattern Recognition and Machine Learning, 2006

5. B. Zadrozny, "Learning and Evaluating Classifiers under Sample Selection Bias", Proceedings of the 21st International Conference on Machine Learning, 2004.

6. "Adaptive Blind Signal and Imaging Processing: Learning Algorithms and Applications" A. Cichocki, S. Amari, Wiley, New York, NY, USA (2002).

KEYWORDS: electronic warfare, data fusion, multiple target tracking, adaptation, in-situ learning

** TOPIC AUTHOR (TPOC) **

DoD Notice:   Between July 20 and August 16, 2010, you may talk directly with the Topic Authors to ask technical questions about the topics. Their contact information is listed above. For reasons of competitive fairness, direct communication between proposers and topic authors is not allowed starting August 17, when DoD begins accepting proposals for this solicitation.

However, proposers may still submit written questions about solicitation topics through the DoD's SBIR/STTR Interactive Topic Information System (SITIS), in which the questioner and respondent remain anonymous and all questions and answers are posted electronically for general viewing until the solicitation closes. All proposers are advised to monitor SITIS (10.3 Q&A) during the solicitation period for questions and answers, and other significant information, relevant to the SBIR 10.3 topic under which they are proposing.

If you have general questions about DoD SBIR program, please contact the DoD SBIR Help Desk at (866) 724-7457 or email weblink.