The RTPS Attack Dataset was created to support research on the cybersecurity of autonomous ground vehicles. The dataset was collected by introducing attacks on a Stella N1 robot, which uses the ROS2 framework and the RTPS protocol for communication.
The test-bed environment consisted of a victim robot (Stella N1), a controller, an attacker PC, and a router. Two types of attacks were performed: Command Injection and ARP Spoofing. The Command Injection attack involved injecting malicious commands into the robot's control packets, either through a flooding or fuzzing approach. The ARP Spoofing attack aimed to intercept the communication between the controller and the robot by manipulating the ARP table.
The dataset includes 240 packet dump files collected from the robot during normal operation and attack scenarios, as well as 2,948 attack packet dump files. Additionally, 8 labeling files provide information on the type of attack (Command Injection or ARP Spoofing) for each packet. The dataset is organized into folders based on the data collection duration (180, 300, 600, or 1200 seconds).
This dataset can be used to develop security solutions for ROS2-based systems, such as anomaly detection models and attack mitigation techniques. The comprehensive nature of the dataset, covering both normal and attack scenarios, makes it a valuable resource for researchers and practitioners working on the security of autonomous ground vehicles.
In eine andere Sprache
aus dem Quellinhalt
arxiv.org
Wichtige Erkenntnisse aus
by Dong Young K... um arxiv.org 04-02-2024
https://arxiv.org/pdf/2311.14496.pdfTiefere Fragen