Fingerprinting-Based Indoor Localization with Commercial MMWave WiFi - Part I: RSS and Beam Indices

Millimeter-wave (mmWave) communications is an emerging technology expected to bring unprecedented data rates and throughput. WiFi operating at unlicensed 60 GHz range is envisioned to become an ubiquitous technology and the IEEE 802.11ad standard is an initial attempt in that direction. Although spatial and temporal resolution of mmWave signals make them suitable for location estimation, a variety of hardwarerelated issues and commonly encountered difficulties in extracting channel measurements from commercial chipsets, challenge opportunistic use of commercial mmWave WiFi chips for indoor localization. We propose in this paper an indoor localization method that fingerprints transmit beam indices that a pair of WiFi transceivers employ to establish a mmWave link, as well as the resulting received signal strength (RSS). In particular, we develop an algorithm that learns possible probabilistic models from the fingerprint data and leverages them to perform indoor localization in the online stage. The proposed algorithm is experimentally evaluated using commercial 60 GHz WiFi routers in an office space