Published on Mon Dec 03 2018

ORACLE: Optimized Radio clAssification through Convolutional neuraL nEtworks

Kunal Sankhe, Mauro Belgiovine, Fan Zhou, Shamnaz Riyaz, Stratis Ioannidis, Kaushik Chowdhury

This paper describes the architecture and performance of ORACLE, an approach for detecting a unique radio from a large pool of bit-similar devices. It trains a convolutional neural network (CNN) that balances computational time and accuracy, showing 99% classification accuracy.

0
0
0
Abstract

This paper describes the architecture and performance of ORACLE, an approach for detecting a unique radio from a large pool of bit-similar devices (same hardware, protocol, physical address, MAC ID) using only IQ samples at the physical layer. ORACLE trains a convolutional neural network (CNN) that balances computational time and accuracy, showing 99\% classification accuracy for a 16-node USRP X310 SDR testbed and an external database of $>$100 COTS WiFi devices. Our work makes the following contributions: (i) it studies the hardware-centric features within the transmitter chain that causes IQ sample variations; (ii) for an idealized static channel environment, it proposes a CNN architecture requiring only raw IQ samples accessible at the front-end, without channel estimation or prior knowledge of the communication protocol; (iii) for dynamic channels, it demonstrates a principled method of feedback-driven transmitter-side modifications that uses channel estimation at the receiver to increase differentiability for the CNN classifier. The key innovation here is to intentionally introduce controlled imperfections on the transmitter side through software directives, while minimizing the change in bit error rate. Unlike previous work that imposes constant environmental conditions, ORACLE adopts the `train once deploy anywhere' paradigm with near-perfect device classification accuracy.

Tue Oct 29 2019
Machine Learning
DeepWiFi: Cognitive WiFi with Deep Learning
DeepWiFi hardens the baseline WiFi (IEEE802.11ac) with deep learning. It sustains high throughput by mitigating out-of-network interference. DeepWiFi is interoperable with baseline WiFi and builds upon the existing WiFi's PHY transceiver chain.
1
0
1
Fri Oct 25 2019
Machine Learning
Mobile Traffic Classification through Physical Channel Fingerprinting: a Deep Learning Approach
This paper proposes and validate algorithms to perform this task, at runtime, from the raw physical channel of an operative mobile network. Two datasets are collected through a large measurement campaign: one labeled, used to train the classification algorithms, and one unlabeled, collected from four radio cells in Barcelona.
0
0
0
Fri Jun 25 2021
Machine Learning
A Photonic-Circuits-Inspired Compact Network: Toward Real-Time Wireless Signal Classification at the Edge
Machine learning (ML) methods are ubiquitous in wireless communication systems. ML data processing at a sub-millisecond scale will enable real-time network monitoring. With the proposed model, we achieve 96.32% classification accuracy.
1
1
0
Thu Mar 02 2017
Machine Learning
Wireless Interference Identification with Convolutional Neural Networks
We propose the first WII approach based upon deep convolutional neural networks (CNNs) The CNN naively learns its features through an extensive data-driven GPU-based training process. We show that the CNN outperforms state-of-the-art WII approaches.
0
0
0
Tue Mar 02 2021
Machine Learning
Deep Neural Network Feature Designs for RF Data-Driven Wireless Device Classification
Most prior works on deep learning-based wireless device classification using radio frequency (RF) data apply off-the-shelf deep neural network (DNN) models. RF data possesses unique characteristics that differentiate it from other domains. Our proposed DNN feature designs substantially improve classification robustness.
1
0
0
Fri Dec 01 2017
Machine Learning
Deep Neural Network Architectures for Modulation Classification
We use a deep neural network to learn how to use a radio frequency. The network has an accuracy of approximately 83.8% at high SNR. We also use a network with two dense layers to achieve high accuracy.
0
0
0