- 综合Fog computing + ML + IoT attack/anomaly detection 三者的研究论文方向:
- system architecture: 提供三者的组成关系
- ML/DL methods: detection的方法有不同
- study cases: 针对不同的应用或者网络环境,例如smart city, vehicle; cellular network, sensor network, low power wide area networks
- fog computing management: how to dynamically orchestrate system components/ manage computing resources/ security problem of edge servers themselves.
Paper: Dynamic management of a deep learning‑based anomaly detection system for 5G networks This paper points out that new technologies should consider new features: efficient, automatic, seamlesss =>
- resource management according to the number of users and their generated traffic
- hot upgrade of detection models
- dynamic deployment of new resources on demand
- deployment of specific analysis tools to extract detailed information
- 除了KDD 99 和 NSL-KDD之外的一些anomaly detection的数据集:
- ISSNIP dataset (Melbourne IoT data)
- IBRL dataset
- s12
- Banana
- 一些有名的安全方面的会议:
- Oakland,
- USENIX Security,
- CCS,
- NDSS …