Fully adaptive radar cost function design

Abstract

Cognitive radar promises to improve radar performance over traditional feed-forward systems by emulating animals' cognitive processes. The fully adaptive radar (FAR) framework provides a generalized approach to implementing the cognitive perception-action cycle for radar systems. This paper offers a multi-objective optimization based methodology for designing the cost functions which drive the FAR framework's operation. The proposed general cost functions were applied to waveform adaptation for single target tracking. Experimental results were used to demonstrate how tailoring the cost functions make the FAR performance amenable to particular radar applications.