Abstract:

Fog computing is usually applied to industrial control, Internet of vehicles, intelligent medical and other real-time scenarios. These scenarios require dynamic, adaptive and low delay access control methods. Traditional access control methods are not suitable. This paper proposes a dynamic access control method based on trust value. This method extracts device behavior features from log files. And then calculates the trust value of active devices using machine learning models, while the trust value of inactive devices is calculated using grayscale method. The method proposed in this paper can not only realize fast and dynamic access control, but also can be easily migrated to multiple scenarios. The author have implemented the method based on equipment and data set provided by the research group.

Key words: fog computing, trust value, access control, machine learning