Joint AoA-ToF Estimation for Robust Indoor Direction Finding

Abstract

Locating victims in a disaster situation using autonomous vehicles has gained a lot of interest in the past few years due to an enormous potential in the technology. In a disaster situation, locating victims with minimal exploration time is of vital importance for the survival of victims. Unmanned Aerial Vehicles (UAV), equipped with multiple sensors such as, visual, thermal and wireless can be used for generating exploration maps that can facilitate in search and rescue missions. These multi-layered maps can be used separately or as an integrated map to explore the environment and to select the next best exploration action to minimize the search and rescue time. Visual and thermal sensors cannot work efficiently in poor visibility conditions, such as in dense smoke or if the victim is behind an obstacle. Radio frequency (RF) signals can propagate through visual obstacles, augmenting UAV with a wireless sensor can potentially decrease the exploration time. Most of the smart phones and smart devices support Wi-Fi (IEEE 802.11), Long Term Evolution (LTE) and Bluetooth standards. RF signals emitted, using any of the aforementioned standards, from a smart device of a victim can be used to estimate the Angle of Arrival (AoA) i.e, the direction of a victim. UAV equipped with a Uniform Linear Array (ULA) antenna can estimate AoA which can be used to build a wireless map. However, there are many significant challenges in an indoor environment that can affect the accuracy of AoA such as, Non-line of sight (NLOS) and multipath conditions. In a multipath environment, RF signals are reflected-off several objects in the environment and the received signal is a superposition of all the multipath signals. Although, super resolution methods such as, Multiple Signal Classification (MUSIC) can be used to estimate the AoA. However, the estimated AoA can be severely degraded by the multipath reflections. The main challenge is to distinguish between AoA of the direct path signal i.e., the signals with the shortest time-of-flight (ToF) which gives the true angle of the emitter/victim and the reflected signals. To tackle this challenge, this thesis proposes a framework for estimating and tracking angle of arrival (AoA) of the direct path signal in an indoor environment. A two-step technique is proposed for direction finding (direct path AoA), using Wi-Fi signals emitted from the smart phone of a victim. First, 2D-MUSIC algorithm is used to estimate the time-of-flight (ToF) and AoA of each path using the channel state information (CSI). Afterwards, particle filter (PF) is used to estimate and track the AoA of the direct path signal while the UAV is moving. The proposed system is evaluated using simulations for an indoor environment. Specifically, we consider an indoor environment for tracking the AoA of the direct path signal using Wi-Fi transmission. Simulation results show that the proposed system can track the AoA of direct path with a median error of less than 5o.

Date of Award	May 2020
Original language	American English