Developing an indoor localization method for the internet of things (IoTs)

In this thesis, a new indoor localization system that uses Raspberry Pi 3 Internet of Things (IoT)-enabled devices is developed. In this context; first, an online Radio Frequency (RF) fingerprinting positioning approach is proposed instead of the traditional offline RF fingerprinting positioning. Using Wi-Fi received signal strengths, and by employing K-Nearest-Neighbour (KNN) algorithm, the effectiveness of the new approach is evaluated on a two-dimensional positioning system through real-life experiments. Next, the applicability and performance of the proposed online RF fingerprinting localization approach is studied on a three-dimensional positioning system. To this effect, seven distance metrics together with three location prediction methods are tested in the KNN algorithm. Additionally, effects of channel interference on both positioning accuracy and system robustness are investigated. The proposed and implemented indoor localization method presented in the thesis is unique when it comes to the approach, devices and the environment of deployment, and it is successfully shown to yield promising results.