Researching the Accuracy of Indoor Positioning using NTNU's Wireless Network

Abstract

Indoor positioning can be based on several positioning techniques and systems. The possibilities for indoor positioning systems are endless. They can provide useful information in complex indoor environments such as hospitals, train stations and office buildings. Existing indoor positioning techniques and systems using these techniques and their corresponding accuracies were explored.This thesis researched the accuracy of NTNU's wireless network through studying the effects of including outer and internal walls in a model and if the location estimates provided room accuracy. A test was performed to study the accuracy of the wireless network. Outer and internal walls were included in five different phases and the locations of a network of control points were estimated based on Cisco's Wireless Location Appliance system. Other indoor obstacles, such as stairwells and lift shafts, and the material of the obstacles were not included due to technical limitations. The insertion of outer and internal walls in the model had limited effect on the accuracy and the precision of the measurements. The most accurate location estimates were of locations between triangulated access points with close proximity. These were also the only locations obtaining room accuracy. Further work should consider the inclusion of other obstacles and material type in the model and signal strength data.