ZTE Communications ›› 2022, Vol. 20 ›› Issue (3): 43-53.DOI: 10.12142/ZTECOM.202203006

• Special Topic • Previous Articles     Next Articles

Neursafe-FL: A Reliable, Efficient, Easy-to- Use Federated Learning Framework

TANG Bo1, ZHANG Chengming1, WANG Kewen1, GAO Zhengguang2, HAN Bingtao2()   

  1. 1.ZTE Corporation, Shenzhen 518057, China
    2.The State Key Laboratory of Mobile Network and Mobile Multimedia Technology, Shenzhen 518055, China
  • Received:2022-06-18 Online:2022-09-13 Published:2022-09-14
  • About author:TANG Bo received his master's degree from Northwestern Polytechnical University, China in 2005. Currently, as the system architect of ZTE Corporation, he is mainly responsible for federated learning and AI security solutions. His current research interests include container and container cloud, heterogeneous resource scheduling, and AI security and trustworthy AI. He holds several patents in the above research areas.|ZHANG Chengming received his BS degree from Yangzhou University, China in 2011, and ME degree from Nanjing University of Posts and Telecommunications, China in 2015. He is a senior R&D engineer of ZTE Corporation. His current research interests include AI platform, container cloud, resource scheduling, federated learning, and AI security.|WANG Kewen received his BS degree from China University of Mining and Technology, China in 2015, and the ME degree from Beijing Institute of Technology, China in 2017. He is an AI senior algorithm engineer of ZTE Corporation. His current research interests include AI Systems and AI security, such as federated learning, adversarial attack, and defense in deep learning.|GAO Zhengguang received his PhD degree from Beijing University of Posts and Telecommunications, China in 2020. He was a visiting PhD student in High Performance Networks Group, University of Bristol, UK from 2018 to 2019. After graduation, he was selected for “LAN JIAN” program of ZTE Corporation as an algorithm researcher. His current research interests include 5G/6G communication technologies, mobile networks, and machine learning for future communications.|HAN Bingtao (han.bingtao@zte.com.cn) received his BS degree from Tianjin University, China in 2001, and MS degree from Nankai University, China in 2004. He is the deputy director of the State Key Laboratory of Mobile Network and Mobile Multimedia Technology, and the leader for “Adlik” project of the LF AI & Data Foundation. Currently, he is the AI system architect of Central R&D Institute, ZTE Corporation. His current research interests include deep learning algorithms, AI systems, and network intelligence. He is the author and co-author for numbers of patents and related monographs.

Abstract:

Federated learning (FL) has developed rapidly in recent years as a privacy-preserving machine learning method, and it has been gradually applied to key areas involving privacy and security such as finance, medical care, and government affairs. However, the current solutions to FL rarely consider the problem of migration from centralized learning to federated learning, resulting in a high practical threshold for federated learning and low usability. Therefore, we introduce a reliable, efficient, and easy-to-use federated learning framework named Neursafe-FL. Based on the unified application program interface (API), the framework is not only compatible with mainstream machine learning frameworks, such as Tensorflow and Pytorch, but also supports further extensions, which can preserve the programming style of the original framework to lower the threshold of FL. At the same time, the design of componentization, modularization, and standardized interface makes the framework highly extensible, which meets the needs of customized requirements and FL evolution in the future. Neursafe-FL is already on Github as an open-source project1.

Key words: federated learning, privacy-preserving, Neursafe-FL