可信区块链联邦学习技术在无线通信网中的研究与应用

Research and Application of Trustworthy Blockchain-based Federated Learning Technology in Wireless Communication Networks

随着无线通信技术的快速发展,越来越多的数据产生并在网络中传输,包括个人隐私信息和商业机密等敏感数据。数据安全和隐私保护引发通信行业的广泛关注,并已成为人工智能大规模落地的关键挑战点。可信区块链联邦学习从保障信息安全的角度发展人工智能,被誉为最有前景解决这一挑战的新方向。归纳了当前行业面临的关键问题和挑战,分析了可信区块链联邦学习的国内外主要研究状况。讨论了无线通信网中典型的区块链联邦学习的系统架构,并阐述了其系统功能和实现的关键技术。在此基础上,给出该系统下的典型应用场景,为可信人工智能从理论走向实用提供了可能性。

With the rapid development of wireless communication technology,the increasing data is being generated and transmitted in the network,including sensitive data such as personal private information and commercial confidential information.Data security and privacy protection has atracted extensive attention in the telecom industry and has become key challenges for the large-scale commercialization of artificial intelligence(Al).From the perspective of ensuring information security,the trustworthy blockchain-based federated learning(TBFL)is considered as a promising new direction to develop the artificial intelligence and solve the challenges.This paper summarizes the key issues and challenges faced by the current industry and analyzes the main domestic and foreign research status on the TBFL.The system architectures for the typical TBFL in wireless communication networks are discussed,and some system functions and key implementation technologies are described.Finally,the typical application scenarios under the system are given,providing the possibility of the trustworthy AI from theory to reality.