Blind Eye: Motion and Obstacle Detection Leveraging Wi-Fi

TL;DR

The paper investigates detecting motion and obstacles in confined environments by exploiting RSSI fluctuations in existing Wi-Fi networks, eliminating the need for dedicated sensing hardware. It introduces Blind Eye, a passive, device-free approach that uses non-overlapping Wi-Fi channels and RSSI drops to identify obstruction events. The work surveys prior sensing technologies and Wi-Fi-based methods, highlighting limitations that Blind Eye overcomes, such as reliance on cameras or active devices. The results show that motion can be detected through walls and with consumer devices, suggesting practical, low-cost applications in security and robotics using current Wi-Fi infrastructure.

Abstract

Wireless Fidelity or Wi-Fi, has completely transfigured wireless networking by offering a smooth connection to the internet and networks, particularly when dealing with enclosed environments. As with the majority of wireless technology, it functions through radio communication. This makes it possible for Wi-Fi to operate effectively close to an Access Point. However, a device's ability to receive Wi-Fi signals can vary greatly. These discrepancies arise because of impediments or motions between the device and the access point. We have creatively used these variances as unique opportunities for applications that can be used to detect movement in confined areas. As this approach makes use of the current wireless infrastructure, no additional hardware is required. These applications could potentially be leveraged to enable sophisticated robots or enhance security systems.

Figures (9)

• Figure 1: Signal Power v/s RSSI
• Figure 5: Single Obstruction between Transmitter and Receiver
• Figure 6: Obstruction between Transmitter and Receiver at LOS=5.3m
• Figure 7: Plot of Result when two Different Obstruction in LOS
• Figure 8: Momentary drop in signal when obstruction moves fast