Non-cooperative Target Detection/Identification (ID)

ABSTRACT: To reduce the incidence of fratricide, improve time to engagement, and increase situational awareness, the Air Force is pursuing technologies with the capability of positively identifying non-cooperative targets. However, single-sensor NCTR systems may be unable to resolve all target identification ambiguities. Thus, there is a significant need for a system that can fuse information from multiple dissimilar sensors and produce joint target identification hypotheses. The proposed program will address these needs through the development of a distributed, multi-sensor, non-cooperative target recognition (NCTR) system. A key innovation to be provided by the proposed solution is the ability to integrate the NCTR system with a distributed tracking system to achieve a single integrated ID picture across a network of platforms. Further, the solution will provide a NCTR information compression capability that will enable operations in bandwidth-constrained environment. The effort will also enhance algorithms for runtime efficiency and develop uncertainty management techniques for improved robustness. An advanced software prototype will be implemented to demonstrate the NCTR capabilities on relevant problems. In the development of the prototype, Numerica plans to align NCTR fusion system with needs in the AWACS and NCCT programs to enhance the transition opportunities. BENEFIT: At the present time, most Air Force ISR platforms operate independently from other ISR collectors (i.e., without integrated networked sensing), and the non-cooperative target recognition (NCTR) systems on these platforms operate using data from a single sensor. As such, the target identification capabilities are significantly limited over what could be achieved through aggregation of the information available throughout the network. A fundamental benefit to be provided by this research program is a new capability for distributed (network-centric) NCTR processing. In the envisioned solution, the NCTR algorithms will interface with a network-centric tracker to ensure that NCTR processors throughout the network process feature and classification data in a coordinated way. This will enable a single integrated ID picture to be achieved, a capability that is critical for fratricide reduction and improved time to engagement. Another key benefit to be provided by this research program is a capability to perform NCTR data compression for distribution throughout the network. Feature data from some sensors is liable to be large, thus transmission of all NCTR data throughout a bandwidth-constrained network will be prohibitive. Schemes that compress the NCTR data, yet minimize the degradation in NCTR performance, are needed and this program will pursue the capability. A third benefit of the proposed program is the development of techniques that provide fusion algorithm efficiency enhancements by trading runtime performance for classification estimation precision. Also, algorithms that provide uncertainty management will be developed; the capability is critical for improving the reliability of the NCTR estimate, and thus reducing the incidence of incorrect target identification. Finally, a software prototype will be developed to demonstrate the capabilities of the NCTR fusion system on relevant Air Force scenario data. The overall benefit of this program to the Air Force will be the development of a mature NCTR Fusion System, and the preparation of relevant demonstration results, that is capable of providing a transition solution to Air Force programs such as AWACS and NCCT.