Goal-Oriented Communication For Real-Time Tracking In Autonomous Systems

Real-time remote tracking using under-sampled and delayed measurements is considered here. We study an autonomous system where a transmitter monitors the evolution of a discrete Markov source and sends status updates to a destination over an unreliable wireless channel. The destination is tasked with real-time source reconstruction for remote actuation. We introduce new goal-oriented sampling and communication policies, which leverage the significance and effectiveness of messages, as a means to generate and transmit only the most “informative” samples for real-time actuation. Our results illustrate that semantics-empowered policies significantly reduce both the real-time reconstruction and the cost of actuation errors, as well as the amount of ineffective updates.