ABSTRACT
User presence detection and movement pattern recognition are primary factors that impact the accuracy and reliability in indoor positioning of dynamic users in Wireless Fidelity (Wi-Fi) systems. Although several models have been proposed to improve the set factors, occupancy characterization, which is largely based on device behavior, still impacts the reliability and accuracy of the results obtained. In this paper, an Adaptive Geolocation using Wireless Fidelity for Localization (AGFiLoc) model is proposed. and its performance in improving reliability of user presence detection and movement pattern recognition is compared against the performance of existing models. The developed AGFiLoc model applies sliding window analysis for user presence detection, while a combination of ARIMA and DBSCAN were utilized for the movement pattern recognition of users. Additionally, decision trees machine learning model was used to extend the AGFiLoc models application in dynamic environments. Furthermore, a ε-differential privacy technique was adopted to protect individual’s data within a dataset.The performance of the proposed AGFiLoc model was compared against MLUPPBALS and CSIHBSRCFI models, while considering accuracy, precision, sensitivity, F1-score, latency, and computational efficiency as performance metrics. With respect to the existing MLUPPBALS and CSIHBSRCFI models, AGFiLoc achieved an 8.24% and 4.55% improvement in accuracy, a 9.88% and 7.23% improvement in F1 score, a 20% and 14.29% improvement in latency, and a 21.43% and 13.33% in computational efficiency. Additionally, the precision and sensitivity the AGFiLoc model respectively showed a 9.79% and 10% improvement against the MLUPPBALS model, and it respectively showed a 5.88% and 4.76% percentage improvement against the CSIHBSRCFI model.
References
- Adikpe, A. O., Iyobhebhe, M., Amlabu, C. A., Botson, I. C., Omojola, B. A., Bashayi, J. G., & Ezugwu, C. (2022). A review on wireless fidelity co-location technology adopted indoors for technology-based contact tracing. Jordan Journal of Electrical Engineering. All rights reserved-Volume, 8(2), 133-150.
- Adikpe, A. O., Agbon, E. E., Giwa, A. M., Adeyemo, H. T., Adegoke, M. A., & Njoku, F. C. (2023). Quick Response Code Technology-based Contact Tracing Solutions to Mitigate Epidemic Outbreaks: A Review. Pakistan Journal of Engineering and Technology, 6(3), 34-43.
- Hao, Z., Duan, Y., Dang, X., & Zhang, T. (2020). CSI‐HC: A WiFi‐Based Indoor Complex Human Motion Recognition Method. Mobile Information Systems, 2020(1), 3185416.
- Liu, F., Liu, J., Yin, Y., Wang, W., Hu, D., Chen, P., & Niu, Q. (2020). Survey on WiFi‐based indoor positioning techniques. IET communications, 14(9), 1372-1383..
- Yang, J., Liu, Y., Liu, Z., Wu, Y., Li, T., & Yang, Y. (2021). A framework for human activity recognition based on WiFi CSI signal enhancement. International Journal of Antennas and Propagation, 2021(1), 6654752.
- Jiang, Y., Hu, H. M., Pu, Y., & Jiang, H. (2019). Wilay: building wi-fi-based human activity recognition system through activity hierarchical relationship. In Proceedings of the 16th EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services(pp. 210-219).
- Yang, X., Cheng, J., Tang, X., & Xie, L. (2023). CSI-based human behavior segmentation and recognition using commodity Wi-Fi. EURASIP Journal on Wireless Communications and Networking, 2023(1), 46.
- Gupta, P., Mehrotra, S., Panwar, N., Sharma, S., Venkatasubramanian, N., & Wang, G. (2020). Quest: Practical and oblivious mitigation strategies for COVID-19 using WiFi datasets. arXiv preprint arXiv:2005.02510.
- Trivedi, A., Zakaria, C., Balan, R., Becker, A., Corey, G., & Shenoy, P. (2021). Wifitrace: Network-based contact tracing for infectious diseases using passive wifi sensing. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 5(1), 1-26.
Download all article in PDF
