Dataset for the paper: Plug and Power: Fingerprinting USB Powered Peripherals via Power Side-channel (Q6718532)

This repository contains data related to Plug and Power: Fingerprinting USB Powered Peripherals via Power Side-channel, by Riccardo Spolaor, Hao Liu, Federico Turrin, Mauro Conti, Xiuzhen Cheng, to appear in Proceedings of the IEEE International Conference on Computer Communications (INFOCOM), 17-20 May 2023. This dataset includes the labels and features extracted from the energy consumption of 82 USB peripherals under different states (i.e., Boot, On) and actions (e.g., Read, Write, Upload, Download). The dataset contains more than 175.000 segments extracted from around 20.000 power traces. We have collected the raw power traces with a National Instruments USB-6210 DAQ at a sampling rate of 10kHz. Each segment is one second long. Please, find more details about the data collection in the paper. We identify a USB peripheral by its type (Device_Type), model (Device_Model), and physical device with such type and model (Device_Id). For each power traces segment, we assign a unique identifier (Segment_Id), and we indicate the action performed (Action) and the activity/inactivity proportions (Activity_Ratio and Inactive_Ratio). The remaining columns (with the prefix EC__) are the features extracted from segments using the tsfresh libraries for python V0.19.0 (https://tsfresh.readthedocs.io) Please, support our work by citing our paper: Riccardo Spolaor, Hao Liu, Federico Turrin, Mauro Conti, Xiuzhen Cheng, Plug and Power: Fingerprinting USB Powered Peripherals via Power Side-channel, In Proceedings of the IEEE International Conference on Computer Communications (INFOCOM), 2023. Contact info:Riccardo Spolaor (rspolaor@sdu.edu.cn, Shandong University, Qingdao, China) and Federico Turrin (turrin@math.unipd.it, University of Padua, Padua, Italy).