Duration: 6 months
Cognitive Radars (CRs) were introduced to enhance the information-processing power of traditional radars. They optimize their performance by developing rules of behavior in a self-organized manner, through learning from experience. A CR embodies the four building blocks of cognition, namely, the perception-action cycle, memory, attention, and intelligence. Over the past decade, Cognitive Radar concepts have been studied and applied to ground-based, naval, and air-borne radars. The paradigm of CR for space-borne Synthetic Aperture Radar (SAR) paves the way toward a more efficient usage of on board resources. The foundations of CR are a base for the functional model of a space-borne Cognitive SAR, adapting everything to the space environment. The roadmap to develop such a model should be elaborated on two levels: the system concept level and the technologies level. The system concept of a Cognitive SAR can be based on the functional blocks of cognition. This includes the perception-action cycle as feedback from the receiver to the transmitter, an on board memory for information handling, an attention manager for task scheduling, and an intelligence, taking into account observables at the receiver, actions at the transmitter, and optimal policies. The technologies level comprises the definition of the complex SAR payload architecture, capable of implementing the perception-action cycle. Cognitive capabilities should enhance the performance of this closed-loop, which links the receiver to the transmitter. Therefore, another requirement is to select highly efficient on board processing units, to dynamically adapt the SAR payload in (near) real-time. Innovative data analysis techniques should be evaluated and implemented. Artificial intelligence, machine learning, and optimal control algorithms are all feasible candidates as enabling technologies in the Cognitive SAR framework.
