An adversarial sample generation method for vehicle-mounted Ethernet flow space
By designing an adversarial example generation method in the vehicular Ethernet traffic space and using the particle swarm optimization algorithm to generate the optimal attack strategy, the problem of insufficient portability and practicality of adversarial examples in the vehicular environment in the existing technology is solved, and the efficient generation of adversarial examples to evaluate the robustness of intrusion detection systems is realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEASTERN UNIV CHINA
- Filing Date
- 2025-09-09
- Publication Date
- 2026-07-21
AI Technical Summary
Existing adversarial example generation methods cannot effectively evade intrusion detection systems in vehicular Ethernet environments, have poor portability and practicality, and the generation time increases linearly with the scale of malicious traffic, making it difficult to generate large-scale adversarial examples under limited prior knowledge.
A method for generating adversarial examples for the vehicular Ethernet traffic space is designed. Through an adversarial environment construction module, an optimal attack strategy generation module based on an optimization algorithm, and an optimal attack strategy generalization module, adversarial examples are generated in the traffic space using the particle swarm optimization algorithm to ensure that they can effectively evade the detection of intrusion detection systems.
It improves the portability and practicality of adversarial examples when evading different types of intrusion detection systems, increases generation efficiency, provides a method for generating adversarial examples that conforms to the threat capabilities of attackers, and evaluates the robustness of intrusion detection systems.
Smart Images

Figure CN121037849B_ABST