| Author | Micale, Davide; Costantino, Gianpiero; Matteucci, Ilaria; Fenzl, Florian; Rieke, Roland; Patanè, Giuseppe |
|---|
| Date | 2022 |
|---|
| Type | Conference Paper |
|---|
| Abstract | Software in modern vehicles is becoming increasingly complex and subject to vulnerabilities that an intruder can exploit to alter the functionality of vehicles. To this purpose, we introduce CAHOOT, a novel context-aware Intrusion Detection System (IDS) capable of detecting potential intrusions in both human and autonomous driving modes. In CAHOOT, context information consists of data collected at run-time by vehicle's sensors and engine. Such information is used to determine drivers' habits and information related to the environment, like traffic conditions. In this paper, we create and use a dataset by using a customised version of the MetaDrive simulator capable of collecting both human and AI driving data. Then we simulate several types of intrusions while driving: denial of service, spoofing and replay attacks. As a final step, we use the generated dataset to evaluate the CAHOOT algorithm by using several machine learning methods. The results show that CAHOOT is extremely reliable in detecting intrusions. |
|---|
| Conference | International Conference on Trust, Security and Privacy in Computing and Communications 2022 |
|---|
| Project | Edge enabled Privacy and Security Platform for Multi Modal Transport |
|---|
| Url | https://publica.fraunhofer.de/handle/publica/446183 |
|---|
