A Rule-Based Cognitive Radar Design for Target Detection and Imaging

This article handles the topic of cognitive radar (CR) architecture design in the framework of multifunction radar operating in a resource-constrained and spectrum-constrained environment. Despite the advances in this field of research and its relative technologies, the way humans and echolocation mammals are able to interact with the external environment goes beyond the capability of any available man-made system. A CR can be thought as a system in which the transmitter, receiver, and software parameters can be changed over time in response to the observed scene with the aim to optimize radar performances given limited resources and environment constraints. The radar, therefore, has to reason about what is being observed and has to take decisions about how to use its limited resources to improve its performance. Rules may represent the way the system reasons, while performance encodes the information contained into the received echoes, and can be used to control next actions and system memory. A rule-based cognitive architecture is proposed in this article as a way to design a CR that has to manage its resources dynamically while handling several tasks, such as target detection, imaging, and recognition in a complex and changing scenario.