A New Filtering Algorithm for Generating Path Loss Model to Improve Accuracy of Trilateration-Based Indoor Positioning System

Global Positioning System (GPS) is used to determine someone's or something's position. It is very accurate. But, GPS cannot be used in an indoor environment as its signal is blocked by an object (i.e. roof). To overcome this problem, Indoor Positioning System (IPS) is used in the indoor environment. There are two methods that IPS can use to determine the position. The first method is pattern matching. This method is highly accurate. But, the computation cost is expensive. The second method is the path loss model. Its computation cost is cheap. However, it is fairly accurate. In this paper, a new filter algorithm is proposed to improve path loss model accuracy. The algorithm works by filtering the received signal strength index (RSSI) in the training phase. The algorithm works by replacing the RSSI value at a certain distance by considering the current and the previous RSSI value. The experiment result shows that the proposed method has 3% better accuracy than the existing method.