|
Densely-Packed Target Data Fusion for Naval Mission-level Simulation Systems
Navy SBIR 2010.1 - Topic N101-101 SPAWAR - Ms. Summer Jones - [email protected] Opens: December 10, 2009 - Closes: January 13, 2010 N101-101 TITLE: Densely-Packed Target Data Fusion for Naval Mission-level Simulation Systems TECHNOLOGY AREAS: Information Systems, Sensors, Battlespace ACQUISITION PROGRAM: Assessments, Simulation-Based Acquisition RESTRICTION ON PERFORMANCE BY FOREIGN CITIZENS (i.e., those holding non-U.S. Passports): This topic is "ITAR Restricted." The information and materials provided pursuant to or resulting from this topic are restricted under the International Traffic in Arms Regulations (ITAR), 22 CFR Parts 120 - 130, which control the export of defense-related material and services, including the export of sensitive technical data. Foreign Citizens may perform work under an award resulting from this topic only if they hold the "Permanent Resident Card", or are designated as "Protected Individuals" as defined by 8 U.S.C. 1324b(a)(3). If a proposal for this topic contains participation by a foreign citizen who is not in one of the above two categories, the proposal will be rejected. OBJECTIVE: To develop a framework and technique for enhancing DON mission-level Modeling and Simulation (M&S) to address Detection and Data Fusion (DDF) under dense target scenarios that include hostile, friendly, and neutral forces (i.e. targets of unknown affiliation/allegiance), defining and parameterizing classes of such scenarios and resulting in metrics for optimizing the decision process that point to clear strategic or tactical Courses Of Action (COAs). DESCRIPTION: Data Fusion is a complex research area, as evidenced for example in Ref [1], which discusses the various levels of data fusion set forth by the Joint Directors of Laboratories of the DOD: Level 1 - Object Refinement Level 2 - Situation Refinement Level 3 - Threat Refinement Level 4 - Process Refinement [1] also documents difficulties in translating these levels into DOD DDF system requirements. Applications of data fusion complexities abound in many fields involving human and biological perception [2-4]. This solicitation seeks new concepts in achieving data fusion and in the M&S thereof. Current capabilities within DON mission-level simulators are 1) subject, under densely-packed target scenarios, to ambiguities in track correlation that are difficult to resolve, This solicitation pertains to the following specific data fusion research areas in multi-INTelligence (multi-INT) collection: 1) all-source fusion An example of a complex problem in MDA and MIO is the case in which the Area of Uncertainty (AOU) and/or field of view of a given sensor include two or more targets of unknown allegiance (a case appropriately described as �densely-packed targets�). A worse case may be one in which targets are so densely packed that a given sensor is only capable of seeing them as one target. To compound such problems later (more recent) sensor information may show that an assumed target correlation (in which two apparently different targets were assumed to be one and the same) in future DDF decisions needs to be de-correlated and de-aggregated based upon the new information. Hence prior history on sensings on that target must be retained and incorporated ASREQ. This solicitation seeks improved methods of M&S in the field of dense-target data fusion to facilitate answering the above and related questions. It is evident that higher levels of data fusion are subjective and will benefit from an analytic framework that extends to structured argumentation for a decision process workflow with non-parametric criteria, and is clearly beyond the simpler topic of data integration. Specifically, the desired framework needs to characterize uncertainty of a decision space based on sensitivity, intelligence accuracy, conflict and ambiguity. Sensitivity can be expressed as the difference in results based on input tolerance. In a dense target DDF scenario, processors are assumed saturated so that potentially significant and defining data may not be fused for consideration in the decision. It is the volume and ambiguity in available information that distinguishes a dense target DDF scenario. Thus, in the decision domain, one might examine a confusion matrix of alternative decisions and adjudication that considers discretionary factors. Conflicting data might be examined by 1) clustering of observations and mitigation by weighting by historical context, Effective disambiguation under conditions of multiple targets tracking in a sensor�s field of view and multiple sensors� fields of view for a netted sensor grid are essential goals, since a dense target DDF decision process must accept numerous, diverse, ambiguous and conflicting sensor inputs from a variety of target objects and types of sensors reporting contacts. Probabilities of correct identification may be assigned, although a higher fidelity model would construct probabilities based on raw sensor input. Government Furnished Information (GFI) will be made available to facilitate execution of the M&S. GFI may include raw sensor traffic from a communications model (as an example scenario traffic generator) in which asset nodes, topology and lines of communications are defined. Background sources may be defined under a given environment and also modeled by a communications model. The GFI will aid in representing ground truth with geospatial relationships as well as the confusion for the DDF scenario. Ambiguous and conflicting inputs from typical ISR sources may also be provided as GFI. This solicitation asks for new concepts to enable analysis of the impact of netted sensors to achieve optimal DDF results. Flexibility of approach by bidders is expected and encouraged, as no single solution to Dense Target DDF is as of this writing apparent to the TPOCs, and a novel solution may be the best one. A successful outcome of this solicitation will be improved acquisition capability gleaned from mission-level simulators and their in- or off-line DDF engines, in the areas of improved sensor design, more effective targeting, and more effective Information Operations (IO) training under scenarios of densely-packed targets of unknown affiliation/allegiance. IO is referenced herein because of its heavy reliance on effective DDF for IO COA determination. The solicitation applies to Naval problems in Maritime Domain Awareness (MDA) and Maritime Interdiction Operations (MIO). This effort is expected to fundamentally enhance the state of Modeling and Simulation of campaign outcome based on red/blue decision strategy. PHASE I: Define and develop concepts for improved dense target DDF characterization, leading to an improved common tactical picture (CTP) among surveillance platforms and one or possibly more distributed fusion centers and command facilities, with the ultimate goal of M&S of these concepts. Concepts must address battle space characterization as a factor in target resolution. The concepts shall explore CTP in light of adversary action and shall provide a data model representation of asset topology, sensor product, decision process and exacerbating factors. The simplest M&S techniques showing the largest potential improvement in clear COAs (hence the largest Return on Investment (ROI)) under such situations will be preferred, as run-time of the simulator will almost certainly be impacted. Agent-Based Modeling (ABM) methods and data fusion engines running alongside mission-level simulators must be considered, since such an architecture may be best suited for the stringent needs of IO and related training. PHASE II: Develop, test and demonstrate a pilot representation of the proposed improved data fusion system running with mission-level simulators and possibly with ABMs. Show how disambiguation and COA capability under densely-packed target scenarios is improved, leading to improved M&S, acquisition and more effective IO training. GFI as sample data will be provided to aid in a prototype demonstration. A prototype demonstration in a contractor environment is sought by the Government as the exit criterion for this Phase. PHASE III: Develop an improved distributed data fusion and tracking capability to work with mission-level simulators for operational test and Analysis of Alternatives (AOA). The research should be directed at applications in IO personnel training, as well as at general DON acquisition. Phase III will exit with a full-scale DDF scenario integrating live data feeds in a Military Operations Center (MOC) environment. PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: The resulting product will provide a valuable surveillance data fusion capability for private defense industry and other private sector companies with applications involving distributed sensors with diverse detection characteristics and IO training. This capability will be an enabling technology in valuable products for private industry to sell to Government and other organizations dealing with human perception, human decision-making, and improved data fusion (c.f., [4]). REFERENCES: 2. Klein, Lawrence A., Sensor and Data Fusion: A tool for Information Assessment and Decision Making, SPIE Press, July 2004 3. http://www.nurc.nato.int/news/MSA-2009.pdf 4. http://www.data-fusion.org/article.php?sid=75 5. http://www.fas.org/man/dod-101/sys/ship/weaps/cec.htm 6. http://en.wikipedia.org/wiki/Data_clustering 7. Stone, L.D., et al, Bayesian Multiple Target Tracking, Artech House, Boston, 1999. KEYWORDS: Data fusion; modeling & simulation; sensors; detection; information systems; command and control, Course Of Action (COA) tools
|