| 摘要: |
| 随着近场无线通信技术的发展与物联网设备的普及，物联网无线接入面临日益严峻的安全威胁，危及用户隐私与系统安全。传统基于密码学的无线接入认证机制难以有效应对依赖物理层特性的安全威胁（如设备身份伪造与非法接入）。而现有多因子（如生物特征或硬件指纹）认证机制在资源受限的物联网环境中存在诸多局限，如依赖专用传感器或需用户主动配合等。为了应对这些局限，本文提出一种基于节点运动状态的物联网无线接入安全辅助认证方案MOSPA（Motion State Perception for Au-thentication）。MOSPA首先通过感知节点无线信号的方向、强度及其变化趋势，推测节点的运动状态（包括位置、速度、加速度与方向）。然后，该方案基于节点运动状态的合理性，判定其接入行为是否存在风险。MOSPA无需额外硬件支持，为传统基于密码学的无线接入认证提供额外的判定依据，有效降低恶意连接的入侵风险。实验结果表明，MOSPA在典型物联网接入场景中以较低资源开销实现高准确率的接入行为风险判定，有效识别非法接入行为。该方案为物联网无线接入提供了一种非侵入式和轻量化的安全辅助认证新思路。 |
| 关键词: 无线物联网 安全接入 辅助认证 节点运动状态 两阶段CNN 变分自编码器 |
| DOI： |
| 投稿时间：2025-06-06修订日期：2026-01-21 |
| 基金项目: |
|
| MOSPA: Perceiving and Learning Node Motion for Wireless Authentication Security Enhancement in IoT |
|
Ma Jianhui, Wei Songjie
|
| (Nanjing University of Science and Technology) |
| Abstract: |
| With the rapid development of near-field wireless communication technologies and the widespread deployment of Inter-net of Things (IoT) devices, wireless access in IoT systems is facing increasingly severe security threats, which endanger user privacy and overall system security. Traditional cryptography-based wireless access authentication mechanisms have difficulty effectively addressing security threats that exploit physical-layer characteristics, such as device identity spoof-ing and unauthorized access. Meanwhile, existing multi-factor authentication mechanisms (e.g., those based on biomet-rics or hardware fingerprints) exhibit significant limitations in resource-constrained IoT environments, including reliance on dedicated sensors, high deployment costs, or the need for active user participation. To enhance the security of IoT wireless access while compensating for the shortcomings of existing multi-factor authentication approaches, this paper proposes a security-assisted authentication scheme named MOSPA (Motion State Perception for Authentication). MOSPA infers the motion state of a node by sensing the direction, strength, and temporal variation of its wireless signals. Specifi-cally, the motion state includes key physical attributes such as position, velocity, acceleration, and movement direction. Based on the inferred motion state, MOSPA evaluates the rationality and consistency of a node’s access behavior and de-termines whether the access attempt poses potential security risks. Unlike conventional authentication mechanisms, MOSPA does not require additional hardware support or specialized sensors, making it well suited for resource-limited IoT devices. Instead, it leverages existing wireless signal characteristics to provide supplementary decision-making in-formation for traditional cryptography-based wireless access authentication. By integrating motion state perception into the access control process, MOSPA effectively reduces the risk of malicious connections and unauthorized device access. Experimental results demonstrate that MOSPA achieves high accuracy in access behavior risk assessment with relatively low computational and communication overhead in typical IoT access scenarios. The proposed scheme can effectively distinguish legitimate access attempts from illegal or malicious ones, even under dynamic wireless environments. Overall, MOSPA provides a non-intrusive, lightweight, and practical security-assisted authentication solution, offering a novel perspective for improving the security of wireless access in IoT systems. |
| Key words: Wireless Internet of Things Secure Access Auxiliary Authentication Node Motion State Two-Stage CNN Variational Autoencoder |
