Application of Federated Learning in Industrial Internet with Device Identifier

With the accelerated integration of the Internet and traditional industries, it is essential to study the Industrial Internet of Things (IIoT) based on device identification. Among the organizational elements of the IIoT, the device identifier is the key technology to realize the rapid development of the IIoT. However, IIoT faces many challenges (equipment security, communication security, data sharing, data security, etc.). This paper proposes a framework to realize the problem of data sharing and ensures the reliability of communication between network nodes. In the framework, the AI model is designed according to the device identifiers in the network and federated learning (FL) technology is used to realize the problem of data sharing. However, there are some limitations in FL. In this paper, we employ blockchain technology to ensure the security of data and proposed an algorithm to realize the reliability of communication.