Abstract

Since the GPS executes poorly indoors, there comes the demand for highly accurate positioning systems that can operate indoors. Among technologies, WiFi-based systems are paid more attention since it can utilize the infrastructure such as Access Points (APs). The WiFi fingerprinting technique is the potential solution in WiFi-based systems. The main goal of this dissertation is to handle the fluctuation of the WiFi signals in different situations, which is one of the main problems of using the WiFi fingerprinting technique. Another goal is to improve the positioning accuracy. The main contributions are described as follows.

(1) A multi-condition WiFi fingerprinting dataset is proposed. The dataset is made considering different environmental conditions in an office room. Different Nearest Neighbor-based algorithms are used to validate the usability of the dataset.  From the experiments, the Chi-Squared distance showed its superior results with the mean positioning error of under 2 m.

(2) An effective random statistical method is proposed. In the offline phase, the standardized database is built and it includes the information of the RPs, the expected value, as well as the covariance matrix of WiFi signals at each RP. In the online phase, the covariance matrix is used to determine the user’s position by comparing the calculated distances between the RSS values from the unknown position and the expected values of the predefined RPs in the standardized database.

(3) A novel WiFi fingerprinting-based with the valued tolerance rough set and decision rules (VTRS-DR) method is proposed. To obtain the user’s position, an advanced positioning algorithm with multi-level classification is presented. Through the experiments, the VTRS-DR method got the mean positioning error of 1.85 m in two cases, which is about 50.49 % better than the results of other methods.

Published paper

[1]. Duong Bao Ninh, Jing He, Vu Thanh Trung, and Dang Phuoc Huy, “An effective random statistical method for Indoor Positioning System using WiFi fingerprinting,” Future Generation Computer Systems, vol. 109, pp. 238–248, Mar. 2020 (SCIE-Q1).

[2]. Duong Bao Ninh, Jing He, Luong Nguyen Thi, Khanh Nguyen-Huu and Seon-Woo Lee, “A Novel Valued Tolerance Rough Set and Decision Rules Method for Indoor Positioning using WiFi Fingerprinting,” Sensors, vol. 22, no. 15, pp. 1-25, 30 July 2022 (SCIE-Q2).

[3]. Duong Bao Ninh, Jing He, Luong Nguyen Thi, and Khanh Nguyen-Huu, “Analysis of Distance Measures for WiFi-based Indoor Positioning in Different Settings,” 2022 2nd International Conference on Innovative Research in Applied Science, Engineering and Technology (IRASET), pp. 1-7, Mar. 2022.

[4]. Duong Bao Ninh, Jing He, Vu Thanh Trung, Luong Nguyen Thi, Do Thi, L., Khanh Nguyen-Huu. “A Multi-condition WiFi Fingerprinting Dataset for Indoor Positioning,” In: Dang, N.H.T., Zhang, YD., Tavares, J.M.R.S., Chen, BH. (eds) Artificial Intelligence in Data and Big Data Processing. ICABDE 2021. Lecture Notes on Data Engineering and Communications Technologies, vol 124. Springer, Cham, 2021, pp. 1-13, May 2022.

[5]. Duong Bao Ninh, Jing He, Luong Nguyen Thi, Seon-Woo Lee, Khanh Nguyen-Huu. “Evaluation of Valued Tolerance Rough Set and Decision Rules Method for WiFi-Based Indoor Localization in Different Environments,”. In: Nguyen, T.D.L., Verdú, E., Le, A.N., Ganzha, M. (eds) Intelligent Systems and Networks. ICISN 2023. Lecture Notes in Networks and Systems, vol 752. Springer, Singapore, pp. 186-194, Aug. 2023.