Authors: Omar Adel Ibrahim, Savio Sciancalepore, Gabriele Oligeri, Roberto Di Pietro
Date: 31th of July, 2019
This page is dedicated to the release of the source code and data of MAGNETO, our novel system able to uniquely identify either the brand and the model or the specific commercial USB flash drive, by analyzing unintentional magnetic emissions radiated during the execution of the boot procedures.
All the details about the proposed system can be found in the paper “MAGNETO: Fingerprinting USB Flash Drives via Unintentional Magnetic Emissions”, actually submitted and under review.
The file that is downloadable on this page contains two folders, namely “Scenario1” and “Scenario2”.
- The “Scenario1” folder refers to the experiments where the aim is to identify the specific brand and model of the USB Flash Drive under test, consistently with the reference scenario of a private company, interested in maintaining high levels of security for its critical equipment.
The folder contains two scripts:
- Scenario1_classification: classification of 14 different brands of USB flash drives, based on unintentional magnetic emissions.
- Scenario1_ROC: Receiver Operating Characteristic curve generation script for all the USB under test.
Note that the source data for these experiments are also available for download.
- The “Scenario2” folder refers to the experiments where the aim is to identify the specific USB flash drive, within USB flash drives of the same brand and model. This experimental setting is consistent with the scenario of a critical infrastructure, where identifying the specific USB flash drive is of crucial importance to maintain physical security.
The folder contains three scripts:
- Scenario2_classification: classification of different flash drives, based on unintentional magnetic emissions. It is possible to select the Mosdart USB flash drive or the SanDisk Cruzer 64 GB.
- Scenario2_ROC: Receiver Operating Characteristic curve generation script for all the USB under test.
- Scenario2_testVarFrequency: tests for evaluating the False Positives Rate, while increasing the frequency span of the acquired data.
Note that the source data for these experiments are also available for download.
The download of the file is available here.
Enjoy with the data!