Self-Calibrated UWB based device-free indoor localization and activity detection approach

Localization of people and assets is increasingly important contextual information for many application areas. In the paper we present a new self-calibrated Ultra-Wide Band (UWB) device free localization (DFL) and activity detection approach. The approach applies automatic "zero touch" system setup applying multidimensional scaling (MDS) algorithm for node self-location and local coordinate system specification. The node-to-node distance estimation is based on a two-way ranging method. The radio tomography imaging (RTI) is applied to estimate the location of a person. The study investigates the sensitivity of the system on the number of nodes. It reveals that in order to achieve the expected location error in a system with a low number of nodes, the signal level measurements in the RTI approach have to be complemented by channel property measurements such as channel impulse response captured by UWB system.