Indoor Positioning System for Campus Building Based on WLAN Fingerprint

Today, many buildings have many floors and rooms. The building usually provides a conventional map that shows the name and location of the existing rooms. The use of conventional maps is currently considered less effective because ordinary people who visit to find the desired location have difficulty. Global Navigation Satellite System (GNNS) is unreliable because the signal is not strong enough to pierce the building. A solution is proposed to create an Android application-based system that can detect the location of humans in the building by utilizing a Wi-Fi signal. The proposed system uses fingerprint technique and k-NN (Nearest Neighbour) algorithm. It has a level of accuracy in the room-scale, where the system can find out where the user is in which room. The system was tested in the Faculty of Industrial Technology Building, Universitas Islam Indonesia, with an accuracy of 82% on a room scale. This paper also provides a solution for choosing the access point to be used by creating a block system. The level of system accuracy is affected by the device ability to receive signals, and the signal from the access point is not always stable. Overall, the designed system can detect where the user is when accessing the application.