【Author】
Lee, Haemin; Kim, Joongheon
【Source】12TH INTERNATIONAL CONFERENCE ON UBIQUITOUS AND FUTURE NETWORKS (ICUFN 2021)
【Abstract】With the development of communication technologies in 5G networks and the Internet of things (IoT), a massive amount of generated data can improve machine learning (ML) inference through data sharing. However, security and privacy concerns are major obstacles in distributed and wireless networks. In addition, IoT has a limitation on system resources depending on the purpose of services. In addition, a blockchain technology enables secure transactions among participants through consensus algorithms and encryption without a centralized coordinator. In this paper, we first review the federated leaning (FL) and blockchain mechanisms, and then, present a survey on the integration of blockchain and FL for data sharing in industrial, vehicle, and healthcare applications.
【摘要】随着 5G 网络和物联网 (IoT) 中通信技术的发展,生成的海量数据可以通过数据共享改进机器学习 (ML) 推理。然而,安全和隐私问题是分布式和无线网络的主要障碍。此外,物联网对系统资源的限制取决于服务的目的。此外,区块链技术通过共识算法和加密实现参与者之间的安全交易,而无需集中协调员。在本文中,我们首先回顾了联邦学习 (FL) 和区块链机制,然后对区块链和 FL 的集成在工业、车辆和医疗保健应用中的数据共享进行了调查。
评论