Who Opened the Room? Device-Free Person Identification Using Bluetooth Signals in Door Access

Abstract

Rooms and buildings in public and private institutions are usually secured with door access control, which is prone to intrusion, particularly when the key or access card falls to unauthorized parties. Therefore, this research investigates if users can be identified without them carrying any devices or keys. In recently years, the device-free techniques using wireless signals instead of sensors and devices on body show the great potential in human detection and identification. In this paper, we present the RFDoorGuard, a wireless, invisible and device-free door access system that leverage RSSI signals from Bluetooth Low Energy beacons to recognize the person who accesses the door. We evaluate RFDoorGuard in two real scenarios with a total of 16 users: the first is an office room with the key lock, and the second is a meeting room with swipe card access. We exploit the characteristics of the door access and propose a two-step algorithm. We demonstrate that RFDoorGuard is capable of recognizing user actions and identifying the actual user during door access with the accuracy of 95% to 53% from a group of 2 to 10 people respectively.