RTIL: A Real-Time Indoor Localization System by Using Angle of Arrival of Commodity WiFi Signal

Abstract

Recently, localization technology using existing infrastructures such as LTE and WiFi networks has advanced greatly. However, localization and tracking using WiFi signal is still challenging since the existing localization systems using Received Signal Strength Indication (RSSI) have limitations such as requiring extensive labor and time costs to construct database. Even for the recent advanced localization systems using Angle of Arrival (AOA) or Time of Flight (TOF), the computational burden and modified protocol limit their deployment. In this paper, we propose RTIL, a real-time indoor localization and tracking system with high accuracy by using commodity WiFi signal. Firstly, we devise a novel algorithm for AOA estimation under condition that the target moves. The proposed AOA estimation algorithm not only calculates AOA using only three antennas without large computational burden and special hardware but also eliminates big AOA error caused by the target's movement. Secondly, to improve the AOA estimation accuracy further we propose a refinement method based on Kalman smoothing. Our implementation and evaluation on the commodity WiFi devices show that RTIL can achieve the 67% localization accuracy of 1.2 m.