Metacognition: Enhancing the performance of a cognitive radio
Manik Gadhiok, Ashwin Amanna, Matthew J. Price, Jeffrey H. Reed · 2011
The application of techniques from artificial intelligence, human decision-making processes, optimization, and learning methodologies have led to significant development in the field of cognitive radios. A cognitive radio improves on a traditional radio by incorporating situational awareness, learning, and decision-making to meet the application goals and adapt to the current operating environment. The field of cognitive radios is a very active, multi-disciplinary research area with significant benefits in many application areas. In this paper, we propose a “metacognitive engine”, where a master process or controller that monitors and adapts the cognition process, generally called the “cognitive engine”. The metacognitive engine utilizes situational awareness and scenario classification and drives the cognitive process. By incorporating radio-level or system-level goals/missions and selecting the most suitable performance metric based on the mission and current state of the cognitive process, the metacognitive controller can adapt/modify the cognitive process to better utilize past experiences and achieve the best performance in the given operating environment. A case study where we apply the metacognitive radio to improve wireless communication systems used for signaling and train control in railroads is also presented.