Design and implementation of an ad-hoc indoor positioning system based on a distributed routing algorithm
Pontus Borg, Arvid Wall · Chalmers Publication Library (Chalmers University of Technology) · 2019
Compared to positioning outdoors, which to a large degree has been resolved with the use of GPS, positioning indoors suffers from multipath and attenuation effects, mainly due to building structures causing bad line of sight (LOS).With this aspect in mind, the thesis investigates the possibility of implementing a relative positioning system with mobile nodes to locate conference, team or meeting rooms using Zigbee technology.A wireless sensor network (WSN) is constructed using the received signal strength indicator (RSSI) as a measurement of distance.Results show that RSSI used to represent distance is highly dependant on a number of factors: The direction of the antenna for both the transmitter and the receiver, the channel on which the sensor network is operating on and fabrication differences between devices have all been noted to affect measurements.Tests show that variance of transmission power affect the measurements to a degree such that it is not possible to relatively sort devices based on signal strength located at a distance shorter than three meters apart.In order to locate rooms based on signal strength the signals need to be processed with a Kalman filter before the positioning can take place.Zigbee devices rely on hop-count as a metric of distance and nodes implementing Zigbee does not store the shortest path to every other node in the network.To create a network better suited for locating remote nodes in the network a shortest path algorithm called DV-distance is implemented.It relies on the Bellman-Ford shortest path algorithm and uses RSSI measurements instead of hop-count as a measurement of distance.Successful results, devices sorted in the correct order, are achieved when the network is set-up in the correct way.