The research questions addressed in this thesis are:
Can 2D positions be estimated in real-time using a machine learning approach on a limited processor in a modifiable indoor environment?
Can we predict the suitability of a given map for the proposed localization approach?