Location
Hilton Hawaiian Village, Honolulu, Hawaii
Event Website
https://hicss.hawaii.edu/
Start Date
3-1-2024 12:00 AM
End Date
6-1-2024 12:00 AM
Description
Confining wireless signals (WiFi) in specific areas of indoor spaces is an efficient way to protect these networks against unwanted access. Unfortunately, these same WiFi signals can be utilized to track the location of mobile handsets. There is an apparent tradeoff between securing the range of such signals and their use for indoor geolocation purposes. The modeling of wireless signal coverage for both security and geolocation purposes in areas where measurements are difficult to record can be a daunting task. We utilized a deep autoregressive model and a convolutional neural network model trained on a synthetic floor plan dataset to accurately extrapolate signal coverage across such spaces without using specific information about antennae placements or floor plan designs. Computational experiments showed that these data-driven approaches were able to fill the gaps in signal coverage maps accurately.
Recommended Citation
Konak, Abdullah; Delattre, Simon; and Bartolacci, Michael, "Wireless Signal Prediction using Deep Learning Models for WiFi Positioning and Security Concerns" (2024). Hawaii International Conference on System Sciences 2024 (HICSS-57). 6.
https://aisel.aisnet.org/hicss-57/li/research/6
Wireless Signal Prediction using Deep Learning Models for WiFi Positioning and Security Concerns
Hilton Hawaiian Village, Honolulu, Hawaii
Confining wireless signals (WiFi) in specific areas of indoor spaces is an efficient way to protect these networks against unwanted access. Unfortunately, these same WiFi signals can be utilized to track the location of mobile handsets. There is an apparent tradeoff between securing the range of such signals and their use for indoor geolocation purposes. The modeling of wireless signal coverage for both security and geolocation purposes in areas where measurements are difficult to record can be a daunting task. We utilized a deep autoregressive model and a convolutional neural network model trained on a synthetic floor plan dataset to accurately extrapolate signal coverage across such spaces without using specific information about antennae placements or floor plan designs. Computational experiments showed that these data-driven approaches were able to fill the gaps in signal coverage maps accurately.
https://aisel.aisnet.org/hicss-57/li/research/6