Enable Angle of Arrival in LoRa for Efficient Indoor Localization

Though angle of arrival (AoA) has been recently adopted in many localization systems (e.g., WiFi, Bluetooth), it is still an open issue whether it can be enabled in LoRa for long-range localization. There are mainly two challenges: (1) current LoRa platform is not equipped with multiple RF channels for the array, and (2) the phase information is unavailable from the PHY layer. In this chapter, we propose a new method utilizing the ranging information for efficient AoA estimation, showing that multi-channel capability and phase information are not necessary. Specifically, we design a mobile gateway with a virtual array extended by an RF switch on a single channel. We further modify the ranging procedure by optimizing the hopping mechanism and packet format. To improve the AoA performance, we design a binary classification framework that achieves a theoretical error of 180/n degrees, where n is the number of antennas. Further, we propose a “rotate and follow” strategy to create more “virtual antennas” for the AoA estimation. The prototype localization system, RLoc, is built for both LoS and NLoS scenarios. Experiments show that RLoc achieves an average AoA estimation error of 2.4 degrees and is efficient in kilometer-level area, which outperforms to the existing localization systems.