基于车辆的入侵检测方法、装置、系统及存储介质
By running an eBPF program in the vehicle operating system kernel to capture log data, construct and transform semantic vectors, and dynamically adjust thresholds to detect vehicle intrusions, the intrusion detection problem of intelligent connected vehicles is solved, and the accuracy and adaptability of detection are improved.
CN122120034BActive Publication Date: 2026-07-17ZEBRED NETWORK TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZEBRED NETWORK TECH CO LTD
- Filing Date
- 2026-04-27
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
How to conduct accurate and comprehensive intrusion detection on vehicles to prevent data leakage and loss of vehicle control, and address the data security risks of intelligent connected vehicles.
Method used
By running an eBPF program in the vehicle operating system kernel to capture log data, constructing a sequence to be detected, and calculating semantic distance through semantic vector transformation and preset cluster center vectors, the threshold is dynamically adjusted to determine the type of intrusion behavior.
Benefits of technology
It achieves non-intrusive data acquisition, improves the accuracy of intrusion behavior detection, avoids false judgments caused by fixed thresholds, and adapts to threshold drift of vehicles under different conditions.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122120034B_ABST
Abstract
本发明公开了一种基于车辆的入侵检测方法、装置、系统及存储介质，方法包括：基于运行在车辆操作系统内核中的eBPF程序对目标系统事件的日志数据进行捕获，并基于捕获到的日志数据构建待检测序列；对待检测序列进行语义向量转换，得到目标语义向量；基于预设的目标类簇的中心向量，确定目标语义向量与每个目标类簇的中心向量之间的目标语义距离；基于预设的滑动时间窗口，获取距离当前时刻最近的N个第一历史语义向量，并基于每个第一历史语义向量以及每个目标类簇的中心向量，确定目标阈值，N为正整数；将目标阈值与每个目标语义距离进行比较，并基于比较结果确定待检测序列对应的入侵行为类型。本方案提高了对车辆入侵行为检测的准确性。
Need to check novelty before this filing date? Find Prior Art