Low Complexity Optimal Policies for Networked Control Systems

Watch a technical lecture where Manali Dutta, a PhD student at IISc Bangalore, explores optimal scheduling policies for wireless networked control systems (WNCS). Discover how to design low complexity policies for systems with plants, sensors, controllers, and actuators connected through unreliable wireless channels. Learn about three specific scenarios: partially observed communication channels, risk-sensitive cost criteria that penalize both mean cost and higher-order moments, and half-duplex controllers. Understand threshold-based transmission strategies for sensor-controller and controller-actuator communications, including when sensors should attempt transmission based on channel belief states, error magnitude thresholds for data transmission, and plant state thresholds for controller-actuator channel activation. Gain insights from Dutta's research background in networked control systems, stochastic control, and reinforcement learning, developed through her studies at IIT Kharagpur and IISc Bangalore.