Automatic Calibration of Anchor Nodes in Device-Free Radio Localization and Motion Tracking Scenarios

A device-free localization (DFL) system exploits a wireless network of radio-frequency (RF) anchor nodes to detect human movements and localize human bodies without requiring the monitored subjects to wear any electronic device. Almost all DFL algorithms require an initial calibration procedure to estimate the electromagnetic (EM) effects induced by the environment. Calibration includes fingerprinting RF signal acquisitions or other on-field measurements to estimate the parameters of the radio propagation model adopted for data processing. This calibration phase is typically time consuming, error-prone and difficult to be fully automatized. In addition, changes or modifications in the layout of the wireless network require a re-calibration stage to maintain good localization performances. In this paper, we propose a novel calibration technique based on physical radio propagation models. For an assigned deployment of the anchor nodes, the proposed calibration scheme is designed to estimate the geometrical and the EM propagation parameters of the scenario and their effects on the localization accuracy. The proposed calibration scheme paves the way to a fully automatized system for DFL applications based on EM propagation models.