Feasibility of Indoor Localization using Angle of Arrival with Low Complexity Hardware

Owing to the enormous growth of the number of devices with wireless connectivity, for example smartphones and tablets but also healthcare monitors, object and device localization has become important. Besides the integration of Global Positioning System (GPS) sensors to localize objects and find locations outdoor, the number of applications collecting indoor location related information is uncountable. Currently developed applications are designed to incorporate indoor location depending functionality. For example one can use a mobile guide to accompany visitors in an exhibition, monitor elderly people at home or trace goods in a warehouse. The needed localization accuracy is strongly application dependent.
This PhD intends to demonstrate the feasibility of the Angle of Arrival localization algorithm to perform object localization within buildings. Therefore, a linear phased antenna array operating at 2.435 GHz is designed, with the intention to develop low complexity hardware based on off-the-shelf components. The collected data samples are processed using different direction finding algorithms. The beamscan, Multiple Signal Classification (MUSIC) and Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) algorithms are elaborated.
Both the implemented algorithm as well as the antenna array architecture introduce inaccuracies when determining the incident angle of a transmitting source. On one hand, the estimated direction will deviate from the real angle, it is called the angular measurement error. On the other hand, the main lobe of the total radiation pattern has a certain beamwidth making the defined transmitter direction a marked zone instead of an exact direction. Both inaccuracies lead to undesired, but inevitable localization errors.
Within the context of this research two main tracks are followed. Firstly, this study investigates what interspacing is needed between multiple coherent sources to be able to distinguish among them. Hence, the resolution of the measurement system is determined, based on the knowledge of the beamwidths. Through theoretical simulations and practical measurements, in the anechoic room, the feasible resolution is defined and shown to be related to the room dimensions and the source position. Secondly, it is intended to predict theoretically what localization error can be expected due to the angular measurement error whenever estimating the incident transmitter angle. Practical measurement results obtained in the anechoic room and an empty room introducing reflections, must verify these predictions. They allow us to make statements concerning common angular measurement errors and thus localization errors feasible with Angle of Arrival localization deployed onto low complexity hardware.