Toward Metacognitive Radars: Concept and Applications

We introduce a metacognitive approach to optimize the radar performance for a dynamic wireless channel. Similar to the origin of the cognitive radar in the neurobiological concept of cognition, metacognition also originates from neurobiological research on problem-solving and learning. Broadly defined as the process of learning to learn, metacognition improves the application of knowledge in domains beyond the immediate context in which it was learned. We describe basic features of a metacognitive radar and then illustrate its application with some examples such as antenna selection and resource sharing between radar and communications. Unlike previous works in communications that only focus on combining several existing algorithms to form a metacognitive radio, we also show the transfer of knowledge in a metacognitive radar. A metacognitive radar improves performance over individual cognitive radar algorithms, especially when both the channel and transmit/receive hardware are changed.