基于工业物联网信道属性的动态认证和多级授权技术研究  

Dynamic Authentication and Multi-level Authorization Technology Based on Channel Attributes of the Industrial Internet of Things

作　　者：段馨雅 康海燕[1] 赵广旭 邱晓英 DUAN Xinya;KANG Haiyan;ZHAO Guangxu;QIU Xiaoying(School of Information Management,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学信息管理学院,北京100192

出　　处：《物流科技》2025年第5期59-62,65,共5页Logistics Sci Tech

摘　　要：随着工业联网4.0的推行与落地,用户对通信网络高安全性的需求也日益增加。由于物联网终端传感器资源受限,传统网络协议难以实现海量入网设备的安全认证。基于计算复杂度的传统身份认证技术已经不再适用于新型工业物联网场景。基于物理层属性的安全技术因具有复杂性低和轻量级的优势,而成为上层安全机制的重要补充。在工业4.0通信场景中,已有身份验证方案研究重点放在了识别分类合法和非法设备,没有针对不同安全级别的资源设置不同访问权限。文章基于工业物联网中物理层信道的特异性,提出了一种基于动态认证和多级授权技术的安全认证方案。具体而言首先构建了一种信道估计矩阵的分布函数与卷积映射一一对应的关系网络,其次提出使用支持向量机作为分类器,然后根据恶意设别信道属性的不同来区分身份,最终引入了一个基于多级授权分类的异常检测框架,同时通过仿真分析验证了多级授权技术优越性。With the advancement and implementation of Industrial Internet of Things(IIoT)4.0,the demand for high security in communication networks is increasing.Due to the limited resources of terminal sensors,traditional network protocols face challenges in achieving secure authentication for a large number of connected devices.Traditional identity authentication technologies based on computational complexity are no longer suitable for new scenarios.Security technologies based on physical layer attributes have become an important supplement to upper-layer security mechanisms due to their advantages of low complexity and lightweight.In the communication scenarios of Industry 4.0,existing authentication scheme research has focused on identifying and classifying legitimate and illegitimate devices,without setting different access permissions for different security levels.In this paper,based on the specificity of the physical layer channel in IIoT,we propose a security authentication scheme based on dynamic authentication and multi-level authorization technologies.Specifically,we first establish a network that correlates the distribution function of the channel estimation matrix with convolution mapping,then suggest using support vector machine as a classifier,and differentiate identities based on different malicious channel attributes.Finally,we introduce an anomaly detection framework based on multi-level authorization classification and validate the superiority of multi-level authorization technology through simulation analysis.

关 键 词：工业物联网 动态认证 多级别授权 信道属性 物理层 

分 类 号：F253.9[经济管理—国民经济]