Enabling Ubiquitous Occupancy Detection in Smart Buildings: A WiFi FTM-Based Approach

Occupancy information is an essential primitive for a wide range of applications like building energy management, security, and behavior analysis. However, balancing accuracy and cost in occupancy sensing is a longstanding challenge. Traditional sensor-based occupancy sensing either fails in coverage area, or in accurate delivery of occupancy information in certain scenarios, e.g. detecting stationary occupants. We propose WiFine, a device-free solution for occupancy detection by leveraging WiFi Fine-Time Measurement (FTM) signals, with enough resolution to detect stationary and moving occupants on single-antenna WiFi devices. Compared to existing WiFi-based methods, WiFine demonstrates higher accuracy while using less sampling rate. WiFine can be adopted by any set of two or more WiFi IoT devices and turning them into occupancy sensors, enabling ubiquitous sensing without requiring new hardware. In real-world experiments, WiFine achieves 95.8% accuracy in different room setups and occupancy statuses with up to three participants, and eight hours of data, outperforming CSI-based approaches with higher accuracy and lower data rate.