摘要
随着物联网技术的快速发展,智能摄像头由于易用性和低成本,在个人及公共安全方面得到广泛使用.然而,未授权录像问题也引发了人们对于隐私安全的担忧,因此特定环境内隐蔽智能摄像头的检测和识别具有重要意义.现有隐蔽智能摄像头检测方法无法准确地检测出那些将数据延迟传输或保存到本地的摄像头,因为这些方法主要依赖用户查看监控时所产生的摄像头音视频网络流量.针对这一问题,提出了一种基于设备WiFi重连流量的隐蔽智能摄像头检测方法.该方法通过MDK4泛洪攻击使得已接入WiFi热点的所有智能设备下线重连,然后嗅探和分析环境内智能设备重连WiFi过程中产生的加密流量,利用机器学习方法检测出其中的隐蔽智能摄像头设备.实验结果表明,没有接入WiFi的情况下,对于延迟传输或数据被保存在本地的隐蔽智能摄像头设备,该方法仍然具有较高的检测准确率.
With the rapid development of Internet of Things technology,smart cameras are widely used in personal and public safety due to ease of use and low cost.However,the issue of unauthorized video recording also raises concerns about privacy and security,so the detection and identification of hidden smart cameras in specific environments is of great significance.Existing covert smart camera detection methods cannot accurately detect cameras that delay data transmission or save data locally,because these methods rely primarily on camera audio and video network traffic generated when users view surveillance.To solve this problem,this paper proposes a covert intelligent camera detection method based on device WiFi reconnection traffic.The method uses MDK4 flooding attacks to make all smart devices connected to WiFi hotspots offline and reconnect,then sniffs and analyzes the encrypted traffic generated during the process of smart devices reconnecting to WiFi in the environment,and uses machine learning methods to detect hidden smart camera devices.The experimental results show that even without WiFi access,this method still has a high detection accuracy for hidden smart camera devices with delayed transmission or data stored locally.
作者
郭回
马骏臣
吴礼发
Guo Hui;Ma Junchen;Wu Lifa(School of Computer Science,Nanjing University of Posts and Telecommunications,Nanjing 210023)
出处
《信息安全研究》
北大核心
2025年第2期173-180,共8页
Journal of Information Security Research
基金
国家重点研发计划项目(2019YFB2101704)。
