WiMap: Autonomous Wi-Fi Mapping for Device-free Tracking in Smart Homes

Tracking the locations of users and Wi-Fi-enabled devices, such as smart home appliances, can facilitate more intelligent home automation services. Typical Wi-Fi human tracking systems also depend on the known Wi-Fi device layout, highlighting the essential role of Wi-Fi device mapping. While progress has been made in autonomous Wi-Fi devices mapping, these methods generally rely on the unreasonable assumption that any pair of Wi-Fi devices can communicate with each other. It is impractical because the communication pattern in a home should be centered around Wi-Fi routers, i.e. , all Wi-Fi appliances should communicate with the router rather than the other Wi-Fi appliances. This paper proposes WiMap , an automatic Wi-Fi mapping system suitable for the router-centric communication pattern in smart homes. The basic idea is as follows: though cleaning robots cannot directly communicate with smart home appliances (except routers), their movement can passively affect ambient Wi-Fi signals similarly to humans. Consequently, the Wi-Fi mapping problem can be regarded as the inverse problem of device-free tracking. To resolve cumulative deviation and environmental interference, we propose a multi-trajectory fusion approach and the Velocity-to-AoA (V2A) Wi-Fi mapping model. The V2A model theoretically reveals the method to solve the angle of arrival from velocity without multiple antennas, enabling accurate Wi-Fi mapping. We implement the WiMap with commercial Wi-Fi devices, and the results demonstrate that our angle-based model achieves median localization errors of 0.10m.