Research
Radio-based Localization: From Theory to Practice
With the widespread use of smart devices, radio-based localization has become an essential ingredient for location-based services, e.g., smart navigation and location-based gaming. Numerous localization algorithms have been developed using advanced signal processing, machine learning, and artificial intelligence. On the other hand, their practical usages are doubtful due to various reasons, e.g., complexity, lack of training data, real-time annotation, and et al. In this research, we target to design practical wireless sensing technologies based on the following principles.
Low complexity design operable on hand-held devices.
Self-supervised or semi-supervised learning approach without labeled training data.
Key publications
J. Kang, S.-W. Ko, and S. Kim, "Near-Field Localization with RIS via Two-Dimensional Signal Path Classification," submitted to a possible IEEE journal.
S. M. Yu, K. Han, J. Park, S.-L. Kim, and S.-W. Ko, "Combinatorial data augmentation: a key enabler to bridge geometry-driven and data-driven WiFi positioning," submitted to a possible IEEE journal.
K. Han, S. M. Yu, S.-W. Ko, and Seong-Lyun Kim, "Waveform-guided transformation of IMU measurements for smartphone-based localization," IEEE Sensors J., vol. 23, no. 16, pp. 20379-20389, Sep. 2023.
K. Han, S. M. You, S.-L. Kim, and S.-W. Ko, "Exploiting user mobility for WiFi RTT positioning: A geometric approach," IEEE Internet of Things J., vol. 8, no. 19, Oct. 2021, pp. 14589-14606.