Federated Learning Encounters 6G Wireless Communication in the Scenario of Internet of Things

Abstract

The ultimate goal of Internet of Things (IoT) technology is to evolve into the Internet of Everything. Two key elements of IoT are artificial intelligence (AI) for smart devices and the Internet for communication. Privacy protection has posed as a critical challenge for the next intelligent IoT technology revolution as the rapid development of communication technology and big data. Federated learning (FL) combines the privacy protection with machine data analytic and it balances the needs of huge volume data for AI and privacy protection, which also makes it as a leading position in the field of machine learning. However, the way of communication that adopted in federated learning resulted in several critical challenges, such as limited bandwidth, data security, and inconsistent internet speed. In this article, we introduce a super-wireless-over-the-air federated learning framework based on 6G technology to address these issues. By training private data in wireless communication with interference-resistant solid radio waves, future security, and ultra-high-performance AI technology can be realized, which could drive the development of IoT to be smarter, wider, and faster.