End-to-End Radio Fingerprinting with Neural Networks

TL;DR

This paper introduces a neural network-based radio fingerprinting method that accurately classifies device identity and transmission distance from RF signals, achieving high accuracy in a previously challenging task.

Contribution

It presents a novel end-to-end neural network architecture with residual connections for RF device and distance classification, demonstrating significant accuracy improvements.

Findings

88.33% accuracy on 16 devices over multiple distances

Effective pre-training for large-scale, subtle RF classification tasks

Successful discrimination of device identity and transmission distance

Abstract

This paper presents a novel method for classifying radio frequency (RF) devices from their transmission signals. Given a collection of signals from identical devices, we accurately classify both the distance of the transmission and the specific device identity. We develop a multiple classifier system that accurately discriminates between channels and classifies devices using normalized in-phase and quadrature (IQ) samples. Our network uses residual connections for both distance and device classification, reaching 88.33% accuracy classifying 16 unique devices over 11 different distances and two different times, on a task that was previously unlearnable. Furthermore, we demonstrate the efficacy for pre-training neural networks for massive data domains and subtle classification differences.