车辆算力网络中异步鲁棒联邦学习方法研究
Research on asynchronous robust federated learning method in vehicle computing power network
尹宏博 1王帅 1张科 2张引3
作者信息
- 1. 电子科技大学信息与通信工程学院,四川 成都 611731
- 2. 电子科技大学信息与通信工程学院,四川 成都 611731;电子科技大学(深圳)高等研究院,广东 深圳 518110
- 3. 电子科技大学信息与通信工程学院,四川 成都 611731;广东省智能机器人研究院,广东 东莞 523830
- 折叠
摘要
传统联邦学习的同步训练机制并不适用于动态的车辆算力网络场景,且在恶意车辆攻击的威胁下,缺乏有效的攻击检测机制.为了解决以上问题,首先,提出一种异步鲁棒联邦学习方法,通过车辆之间异步地执行联邦学习过程,在实现车辆数据隐私保护的同时,提高模型协同训练的效率.其次,有针对性地设计了模型选择方法,并提出潜在恶意模型检测方法和车辆信誉评估方法,进一步增强系统鲁棒性.然后,从概率上详细分析了所提方法的安全性,为各项参数优化提供理论基础.最后,仿真结果表明该方法能够在实现高效异步联邦学习的同时具有较好的鲁棒性.
Abstract
The synchronous training mechanism of traditional federated learning was not suitable for dynamic vehicle computing power network scenarios,and lacked effective detection mechanisms under the threat of malicious vehicle at-tacks.To address the above issues,an asynchronous robust federated learning method was proposed,which achieves ve-hicle data privacy protection while improving the efficiency of model collaborative training through asynchronous execu-tion of federated learning processes between vehicles.Secondly,a model selection method was designed,and potential malicious model detection and vehicle reputation evaluation methods are proposed to further enhance the robustness of the system.Then,the safety of the proposed method was analyzed in detail from a probabilistic perspective,providing a theoretical basis for optimizing various parameters.Finally,the simulation results show that this method can achieve effi-cient asynchronous federated learning while having good robustness.
关键词
车辆算力网络/联邦学习/鲁棒性/异步学习Key words
vehicle computing power network/federated learning/robustness/asynchronous learning引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据