Semantics-empowered communication for networked intelligent systems

Wireless connectivity has traditionally been regarded as a content-agnostic data pipe; the impact upon receipt and the context- and goal-dependent significance of the conveyed messages have been deliberately ignored. Nevertheless, in emerging cyber-physical and autonomous intelligent networked systems, acquiring, processing, and sending excessive amounts of distributed real-time data, which ends up being stale, irrelevant, or useless to the end user, will cause communication bottlenecks, increased response time, and safety issues. We envision a communication paradigm shift that makes the semantics of information, i.e., the importance and the usefulness of information generated and transmitted for attaining a certain goal, the underpinning of the communication process. We advocate for a goal-oriented unification of data generation/active sampling, information transmission, and signal reconstruction, by taking into account process and source variability, signal sparsity and correlation, and semantic information attributes. We apply this structurally new joint approach to a communication scenario where the destination is tasked with real-time source reconstruction for the purpose of remote actuation. Capitalizing on semantics-aware metrics, we explore the optimal sampling policy, which significantly reduces the number of samples communicated and the reconstruction error in ways that are not possible by todays state-of-the-art approaches.