Efficient design of indoor positioning systems based on location fingerprinting

This paper presents results of systematic study to improve designing of indoor positioning systems based on location fingerprinting technique. Thus far, researchers have been focusing on improving this type of positioning systems by arbitrarily selecting one of available pattern classification techniques. On the contrary, this study suggests a modeling framework which can be used as a tool to efficiently design such systems. Using the proposed framework, this paper quantifies accuracy and precision improvement resulted from adjusting system parameters. Consequentially, deployment and performance tuning time which is the most time-consuming process of the fingerprinting technique can be reduced. A system model based on proposed framework is validated with a prototype in which both model and prototype utilize simple Euclidean distance as a technique for location determination.