Kartika Candra Kirana, Hani Ramadhan, Slamet Wibawanto, Heru Wahyu Herwanto
There are boundary issues in predicting indoor positioning. Based on the Haversine formula, the distance is calculated using earth shape approaches that resemble a sphere, so there is a boundary difference between the haversine calculation and the tesseract-designed building. It leads to misclassification of indoor and outdoor positions in the attendance system. Thus, a novel geometry-based indoor positioning is proposed in the attendance system. In this study, the Haversine formula calculates the distance using the latitude, longitude, altitude, and earth curvature. Then a novel threshold is demonstrated using the limit theorem for indoor position filtering. In this preliminary study, we utilize the fixed longitude and latitude of the Surakarta Education and Training Center's Building Information Model (BIM) generated from the google API database to mark the position of the wall. Based on the test results on six indoor nodes and ten outdoor nodes, our proposed method outperforms the traditional geometry method with 75% accuracy, 60 % precision, 100% recall, and 66% false-positive rate. © 2021 IEEE.
Universitas Negeri Malang, Department Of Electrical Engineering, Malang, Indonesia; Pusan National University, Computer Science, Kumjeong-ku, Busan, South Korea
