Object Identification in a Simulated Battlefield using Artificial Intelligence Techniques

This paper examines the use of heuristic techniques in identifying different components (for example, tanks, helicopters, radio sets etc.) that make up the army divisions in a battlefield. Two techniques are used. The first technique examined is a speedup heuristic to perform integration of knowledge from disparate sources (called sensor fusion). This technique is evaluated by determining the amount of time saved in performing the fusion. The second technique involves matching an unidentified object to a known set of parameters (geographically clustered sensor data to divisions). This technique is evaluated by the quality of match achieved. A complete numerical example involving battlefield status information as a function of time is also presented.