Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A SYSTEM AND METHOD FOR DETECTION OF ANOMALOUS CONTROLLER AREA NETWORK (CAN) MESSAGES
Document Type and Number:
WIPO Patent Application WO/2020/208639
Kind Code:
A3
Abstract:
A system for detecting Controller Area Network (CAN) messages anomalies, the system comprising a processing resource configured to: obtain a training set including training CAN messages associated with respective one or more vehicles; and train an anomaly detection model, using the training set, the anomaly detection model comprising one or more CAN message type anomaly detection models for one or more respective CAN message types that appear in the training set, each of the CAN message type anomaly detection models characterizing a functional dependency, if any, between one or more parts of values of payloads and interarrival times of CAN messages of the respective CAN message type, the interarrival times being determined using a timestamps associated with the CAN messages of the respective CAN message type; wherein the anomaly detection model is usable for classifying an unclassified CAN message of a given CAN message type as anomalous or non- anomalous.

Inventors:
STEIN YEHIEL (IL)
VARDI YOSSI (IL)
COSTA NADAV GEORGE (IL)
Application Number:
PCT/IL2020/050432
Publication Date:
November 26, 2020
Filing Date:
April 08, 2020
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
SAFERIDE TECH LTD (IL)
International Classes:
G06F21/55
Foreign References:
US20160188876A12016-06-30
US20160381059A12016-12-29
Other References:
TAYLOR: "Anomaly-based detection of malicious activity in in-vehicle networks", DISS. UNIVERSITE D'OTTAWA, 2017, XP055512983, Retrieved from the Internet [retrieved on 20200909], DOI: 10.20381/ruor-20400
Attorney, Agent or Firm:
JENCMEN, Avi et al. (IL)
Download PDF: