Performance Evaluation of Indoor Positioning Algorithm using Bluetooth Low Energy

Physical security management in data center is mandatory according to ISO 27001. It includes user access and traffic management Monitoring people positions, in the Data Center, will be more efficient with the indoor positioning system. Beside to security needs, indoor positioning system should be used for emergency evacuation needs in the Data Center. Indoor positioning has its challenges where global positioning system (GPS) becomes ineffective when used indoors because of no line of sight (LOS) with satellite. Currently, the indoor positioning is mainly based on the wireless signal, such as WiFi, RFID, Zigbee, Bluetooth, etc. In this paper, indoor positioning is developed based on Bluetooth 4.0 beacon technology and its RSSI value. As the classifier, k-Nearest Neighbors (kNN) and Fuzzy k-Nearest Neighbors (FkNN) algorithm are evaluated. Impact of Kalman Filter application to improve the accuracy also being tested. The simulation results shows that the performance of FkNN is better than kNN. While, Kalman Filter can improve the performance of kNN significantly.