Application of Pattern Recognition Techniques for Early Warning Radar
Abstract
The purpose of this research was to develop and evaluate an ensemble of simple data processing and pattern recognition techniques that could provide the Early Warning Radars (EWRs) with an enhanced target discrimination capability. The PAVE PAWS and the Ballistic Missile Early Warning System (BMEWS) radars are phased array radars which support the Early Warning System (EWS). These radars operate at UHF and were designed to detect and track large numbers of objects, as part of the then perceived threat, i.e., a massive ballistic missile attack of hundreds of ICBMs/SLBMs. Discrimination and tracking of individual objects was not optimized in their design. Our aim was to examine discrimination methodologies that could be implemented without modifying the EWR hardware. Any enhancements would be installed by software upgrades only. Pattern recognition techniques represent one approach to extracting additional discrimination information from the radar cross section (RCS) data that is currently generated by the radar. For example, given approximately 100 seconds or more of radar cross section (RCS) data displayed as a function of time, the global signature of the first or second stage of the missile, or tank object, results in a visual pattern that is very different from that of a reentry vehicle (RV). Moreover, there may be other structural components to the various patterns that could be used as discriminates. Our basic objectives were to develop and assess the mathematical techniques to extract these structures in a systematic way and to implement these techniques in a computer program. To this end we surveyed, in a systematic way, the RCS times histories for a large number of objects.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 29, 1995
- Accession Number
- ADA299735
Entities
People
- Thomas Foster