Orientation Estimation using Wireless Device Radiation Patterns

TL;DR

This paper presents a method to estimate the orientation of wireless devices by analyzing their radiation patterns using RSS measurements, achieving low error without device cooperation and demonstrating resistance to spoofing.

Contribution

It introduces a novel approach to estimate device orientation from radiation patterns without cooperation, applicable to common wireless systems in real-world scenarios.

Findings

Mean orientation error as low as 7.6 degrees

Method resistant to spoofing attacks

Applicable to typical consumer wireless devices

Abstract

Wireless devices inherently have a non-uniform distribution of energy from their antenna or antennas. The shape that this forms is commonly called a radiation pattern or antenna pattern. We demonstrate that orientation can be estimated without the cooperation of the target device despite only having a small number of RSS measurements per packet. We do this by applying bounds to the amount of rotation in the time interval between packets. Using simulations, we show that this method can achieve a mean orientation error as low as 7.6{\deg}. We then perform a security analysis to demonstrate the method's resistance to spoofing. This paper focuses on consumer wireless devices where patterns are not deliberately highly directional and scenarios that cannot rely on contrived movement patterns of the entities involved, which is unrealistic or impractical in many settings. Our work concentrates on existing wireless systems and infrastructure common in domestic, office, and commercial.