计算机科学
身份(音乐)
认证(法律)
具身认知
计算机安全
互联网隐私
情报分析
多因素身份验证
信息隐私
报文认证码
鉴定(生物学)
环境智能
人机交互
保密
电子邮件
作者
Miaomiao Wang,Jiahua Li,Lingfeng Bao,Yijing Lin,Xuesong Li,Xiaodong Liu,Yinuo Hu
标识
DOI:10.1109/tccn.2026.3717992
摘要
As a core paradigm bridging physical and digital worlds, embodied intelligence is widely deployed in key fields such as industrial manufacturing, however its architecture faces acute security risks. Hampered by poor adaptability and high overhead, traditional authentication mechanisms are incompatible with the dynamic topology and low computational constraints of embodied intelligence. To address these challenges, this paper constructs a collaborative security architecture for embodied intelligence integrating identity authentication and trust evaluation, which adopts a consortium blockchain-based trust framework to enable real-time sharing of authentication information and rapid access for dynamic nodes. A multi-dimensional trust evaluation mechanism based on roles, behaviors, and historical interactions is designed to update trust values in real time. In addition, a lightweight multi-factor identity authentication algorithm that combines device fingerprints and operational contexts is proposed, with the elliptic curve cryptography algorithm adopted to reduce computational overhead. During the mutual authentication process, the embodied intelligent devices and edge nodes negotiate a secure session key. Through formal analysis using BAN logic and informal security analysis, this paper verifies the scheme’s security properties, including mutual authentication, key confidentiality, and forward secrecy. Meanwhile, the performance analysis demonstrates that the scheme can meet the requirements of efficient authentication for low computational devices, thereby providing reliable support for the secure interaction of embodied intelligence.
科研通智能强力驱动
Strongly Powered by AbleSci AI