SmartEye: Mobile Device Proximity Monitoring via Wireless Signal Analysis

Abstract

Mobile devices are pervasively used by everyone in all aspects of their daily lives. Sensing capability of the mobile devices, using their built-in sensors, is usually limited to their immediate proximity. In this paper, we exploit a technique which enables the mobile device to sense its physical proximity by taking advantage of the WiFi Channel State Information. We define a model to detect the movements of human and non-human objects in the proximity of the device. We have exploited the Fresnel zone model to detect the movement towards and outwards the device in the sensing area. The scheme can be used to alarm the user when device is left unattended. We further consider two models for early-detection of a user leaving her mobile device based on the user requirements and environment conditions. We evaluate our scheme using information from simulated theft attack scenarios in real environment and show that our scheme can achieve an average 84.44% and 77.77% accuracy on detecting the theft attacks for outdoor and indoor environments respectively.