基于联邦增量学习的物联网数据安全共享方法
Data Security Sharing Method for the Internet of Things Based on Federated Incremental Learning
付晓荣1
作者信息
- 1. 湖北国土资源职业学院,湖北 武汉 430090
- 折叠
摘要
为了提升物联网数据安全共享传输精准度,缩短物联网数据安全共享耗时,提出了一种基于联邦增量学习的物联网数据安全共享方法.构建物联网数据共享模型,通过类增量改进联邦学习,对数据共享模型参数进行更新,有效避免局部最优问题;向SN发生共享请求,根据更新的模型参数进行访问控制权限匹配,计算物联网数据安全共享密钥,恢复目标数据内容,实现物联网数据安全共享方法.实验结果表明:当干扰强度为60dB时,本方法的物联网数据安全共享传输精准度为99.1%,当物联网数据量为300GB时,本文方法的物联网数据安全共享耗时仅为0.9s,表明本文方法能够有效提升物联网数据安全共享效果,提升物联网数据安全共享效率.
Abstract
In order to improve the accuracy of IoT data security sharing transmission and shorten the time required for IoT data secu-rity sharing,this study proposes a federated incremental learning-based IoT data security sharing method.It constructs an IoT data sharing model,improves federated learning through class increment,updates the parameters of the data sharing model,and effec-tively avoids local optimization problems.It generates a sharing request to SN,matches access control permissions based on updated model parameters,calculates IoT data security sharing keys,recovers target data content,and implements IoT data security sharing methods.The experimental results show that when the interference intensity is 60dB,the accuracy of secure sharing transmission of IoT data in this method is 99.1%.When the amount of IoT data is 300GB,the time for secure sharing of IoT data in this method is only 0.9s,indicating that this method can effectively improve the security sharing effect of IoT data and improve the efficiency of IoT data security sharing.
关键词
联邦增量学习/物联网数据/安全共享/云端服务器Key words
federated incremental learning/Internet of Things data/secure sharing/cloud server引用本文复制引用
出版年
2024
NETL
NSTL
万方数据