Orchestration of Reconfigurable Intelligent Surfaces for Positioning, Navigation, and Timing

Positioning, Navigation, and Timing (PNT) services are exploited by critical infrastructures which are strategic for the functioning of the modern society, such as telecom, energy, finance, and transportation. Though the most popular PNT services' provider is the Global Positioning System (GPS), its performance is often impacted by adverse conditions and different varieties of interference, either intentional or unintentional. In this paper, we exploit the efficient and effective orchestration of Reconfigurable Intelligence Surfaces (RISs) as a means of offering an alternative PNT model, improving accuracy and availability. In particular, we initially introduce a low-complexity reinforcement learning-based approach to enable the various targets under consideration to select the most appropriate set of RISs that, acting complementary to available anchor nodes, will minimize the error in the targets' positioning and timing calculation. Subsequently, the optimal phase shifts of the reflected signals on the selected RISs are determined, in order to further improve the proposed PNT model's accuracy. Finally, an iterative least square (ILS) algorithm determines the targets' positioning and timing in a fully distributed manner. The performance of the proposed PNT model is achieved via modeling and simulation, and indicative numerical results are presented demonstrating its benefits and tradeoffs.