U.S. Army Intelligence Center and School USAICS, Software Analysis and Management System USAMS, Correlation Algorithm Report.

Abstract : This is one in a series of algorithm analysis reports on work performed at the Jet Propulsion Laboratory for the US Army Intelligence Center and School covering selected algorithms in existing Intelligence and Electronic Warfare (IEW) systems. It focuses on self-correlation algorithms in five ELINT systems. These algorithms test a newly reported sighting against information already gathered to determine whether it represents a battlefield entity previously reported. In the systems surveyed (BETA, TCATA, ELINT PROCESSOR, ITEP, QUICKLOOK, AGTELIS) these decision tests were found to be mainly quasi-statistical using location, detection time, and various signal parametrics. The analysis focuses on the underlying assumptions (such as independence of sightings) which must hold for decisions based on these tests to be valid, the environmental factors (such as information available to the test) necessary for the assumptions to hold and failure modes when some of the assumptions are relaxed.