Hierarchical Bayesian-based Indoor Positioning Using Distributed Antenna Systems

This work proposes and evaluates hierarchical Bayesian-based localization methods to estimate the position of a target node in indoor deployment scenarios. The measurements are acquired through a distributed antenna system which is connected to a common master anchor node. Each antenna head is affected by different channels parameters, what makes the estimation more difficult. The proposed method combines received signal strength and time of flight measurements to estimate the target location. In our investigations, we also consider a one-level hierarchical Bayesian network model, which introduces conditional interdependencies to the model parameters, resulting in less susceptibility to local variations. The Markov Chain Monte Carlo sampling method is used to approximate the posterior distribution of the two-dimensional target's location coordinates. The root mean square error is used to evaluate the performance of the proposed solution in indoor scenarios. Our results show that by combining hybrid measurements or increasing conditions between the parameters by a hierarchical approach, the proposed mechanisms outperform the classic Bayesian model when estimating the target node using even fewer measurements.