VWAN: Virtual WiFi ANtennas for Increased Indoor Localization Accuracy

摘要

Indoor localization techniques, both device-based and device-free, using WiFi signals is advancing with an ever-increasing pace. Despite being limited by the number of physical antennas, smaller bandwidth range (20 MHz/40 MHz) and non-trivial phase noises, decimeter-level localization using commodity WiFi chipsets (802.11n) is becoming a reality. Unlike specialized wireless modules, commodity devices face the accuracy limitation dictated by antenna count for Angle-of-Arrival (AoA) methods and bandwidth for Time-of-Flight (ToF) methods. Promising state-of-the-art localization schemes like Chronos overcomes the bandwidth limitation by emulating a wideband WiFi radio through frequency hopping. In this paper, we investigate if an increase in effective bandwidth also solves the problem of fewer antennas and enables superior resolution of Multipath Component (MPC) separation. Our study has led to a system called VWAN which is a virtual antenna extension solution that entails no hardware changes. Simulation results demonstrate that VWAN can accurately disentangle almost all MPCs. This capability to precisely separate all MPCs and consequently completely characterize wireless channel will not only increase localization accuracy but enormously benefit other applications such as passive tracking, WiFi imaging and gesture recognition.