Vehicles Position Tracking in Parking lots Using K-Nearest Neighbor and Fingerprinting Based on RSSI Bluetooth

Abstract

Location-based services have become very popular in recent years, among these services are navigation and location determination both indoor and outdoor. Determining an object location is often an interesting topic for study, various studies have been carried to determine the position of the object using several methods using either GPS, CCTV, Wireless or Bluetooth. One of the interesting cases to study is determine vehicle position in the indoor parking lots. It is important to record the parking location because it will make it easier for newcomers to locate their vehicles in the parking area. Various studies have suggested that Bluetooth is a promising wireless technology in indoor positioning system because of its cost-effectiveness and user-friendly features. In this study, We use several methods to record parking positions, including the K-Nearest Neighbor and Fingerprinting techniques based on the RSSI Bluetooth (Received Signal Strength Indicator) with distance calculation parameters using Manhattan Distance. The results of our study, it was found that the accuracy about 77% with Precision, Recall and F1-Score pessimistic approach reaching 87%, 85% and 85%.