Using a knowledge based system for emitter classification and ambiguity resolution

The authors describe a research project aimed at incorporating knowledge-based system technology into EW (electronic warfare) systems to improve system performance and to assist the system operator. The threat assessment function of the EW system was targeted for improvement, and emitter classification was the first portion in which work was done. It was shown that emitter classification can be improved through the use of additional parameters. The tradeoff is increased complexity of the search routine and storage required for the lookup tables. It was found that experts perform classification based on the functional performance of an emitter, rather than lookup. The use of a knowledge-based system is vital in that it can reduce search by allowing the system to look at emitter data from an expert´s point of view (i.e. matching the intercepted parameters to the source emitters function). In addition, it was found that ambiguity resolution can be improved through the use of techniques such as `tie breakers´ to choose from a set of equally likely suspected emitters