In an era where vehicle connectivity is rapidly expanding, safeguarding in-vehicle networks against cyber threats has become paramount. The Controller Area Network (CAN) protocol, the backbone of in-vehicle communication, unfortunately lacks inherent security features, making vehicles susceptible to malicious attacks. Message injection attacks, where fabricated messages disrupt ECU functionality, are a significant concern. To address these vulnerabilities and propel automotive security research, we introduce the Chd Car Hacking Tool and dataset – a vital resource for cybersecurity professionals and researchers.
The CHD Car Hacking Dataset: A Deep Dive
Our publicly available car-hacking dataset is meticulously designed to facilitate the development and validation of intrusion detection systems. It encompasses a range of attack scenarios, including Denial of Service (DoS), fuzzing, and spoofing attacks targeting critical vehicle functions like drive gear and RPM gauge. These datasets were created by capturing real-world CAN traffic via a vehicle’s OBD-II port during active message injection attacks. Each dataset contains 300 instances of intrusion events, each lasting between 3 to 5 seconds, within a comprehensive 30 to 40 minutes of recorded CAN traffic.
Attack Types Featured in the CHD Dataset:
-
DoS Attack: This dataset simulates a Denial of Service attack by injecting ‘0000’ CAN ID messages at a rapid rate of every 0.3 milliseconds. The ‘0000’ CAN ID is intentionally chosen due to its dominant nature in CAN communication, effectively mimicking a flood of high-priority messages.
-
Fuzzy Attack: To represent unpredictable attack vectors, the fuzzy attack dataset includes injections of completely randomized CAN ID and DATA values every 0.5 milliseconds. This simulates a scenario where attackers attempt to disrupt the CAN network with malformed and arbitrary messages to identify vulnerabilities.
-
Spoofing Attacks (RPM/Gear): These datasets focus on targeted manipulation of vehicle operations. Spoofing attacks are executed by injecting messages with specific CAN IDs related to RPM and gear information at a consistent interval of 1 millisecond. This demonstrates how attackers can fabricate sensor readings to deceive vehicle systems.
Understanding the CHD Dataset Attributes
Each entry in the CHD car hacking dataset is structured with the following attributes, providing a detailed view of the CAN bus traffic:
-
Timestamp: Records the precise time of each CAN message in seconds, enabling temporal analysis of attack patterns.
-
CAN ID: The CAN Identifier, represented in hexadecimal format (e.g., 043f), crucial for identifying message types and priorities within the CAN network.
-
DLC: Data Length Code, indicating the number of data bytes in the CAN message, ranging from 0 to 8 bytes.
-
DATA[0-7]: Up to 8 bytes of data payload for each CAN message, representing the actual information transmitted across the network.
-
Flag: A critical indicator distinguishing between injected attack messages and regular CAN traffic. ‘T’ denotes a Transmitted (injected) message, while ‘R’ signifies a Received (normal) message, essential for training intrusion detection systems.
Leveraging the CHD Car Hacking Tool and Dataset for Enhanced Security
The CHD car hacking dataset serves as an invaluable resource for researchers and developers working on intrusion detection systems, security validation, and threat analysis in the automotive domain. By utilizing this dataset in conjunction with appropriate CAN bus analysis tools – effectively making the dataset itself a “CHD car hacking tool” for research purposes – the automotive security community can:
- Develop robust algorithms for detecting anomalies and malicious activities within in-vehicle networks.
- Evaluate the effectiveness of different intrusion detection techniques against realistic car hacking scenarios.
- Gain deeper insights into the vulnerabilities of CAN protocol and develop mitigation strategies.
This dataset empowers the cybersecurity community to proactively address the evolving threats in connected vehicles, ultimately contributing to safer and more secure transportation systems. By providing this comprehensive and meticulously curated dataset, we aim to foster innovation and collaboration in the critical field of automotive cybersecurity.
