Year: 2025
Type: preprint
Abstract: This paper proposes a two-stage Bayesian positioning framework to estimate the three-dimensional position and orientation of a mobile station (MS) within an indoor environment enhanced by a reconfigurable intelligent surface (RIS). The proposed approach models the joint probability distribution of all relevant random variables using a probabilistic graphical model and employs the No-U-Turn Sampler (NUTS) to approximate the posterior distribution based on noisy channel parameter estimates. We propose an outlier detector to identify and mitigate the influence of multiple modes that can occur in Markov chain Monte Carlo (MCMC)-based estimations. To assess the fundamental performance limits, we derive the Bayesian Cramér–Rao bound (BCRB) and use it to evaluate both the position error bound (PEB) and rotation error bound (REB). Furthermore, we extend the framework to dynamic tracking scenarios by incorporating a sequential update mechanism that leverages prior knowledge from previously estimated posteriors. Simulation results demonstrate the effectiveness of the proposed method under realistic channel estimation conditions, confirming its potential for accurate and robust positioning and tracking in RIS-assisted indoor environments.
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4416918647