Opportunistic Positioning Using Unsynchronized References: Crowd Systems - Smartphones: Signals of opportunity, cellular systems (4G, 5G, WLAN, …)

Abstract

With a set of unsynchronized LTE receivers in known locations we localize another receiver, using only ToA measurements of incoming base station signals. We propose an abstract model for our system and cast it in a Bayesian statistical framework. Using Stan we perform MAP optimization to estimate the parameters and Hamiltonian Monte Carlo sampling to explore the posterior distributions. Using simulations and realworld data we show that the parameters of such a model can be estimated, given suitable parametrization. Our model is a basis for further developments in the Bayesian framework, such as adding non-line-of-sight mitigation.