The security of wireless routers receives much attention given by the increasing security threats. In the era of Internet of Things, many devices pose security vulnerabilities, and there is a significant need to analyze the current security status of devices. In this paper, we develop WNV-Detector, a universal and scalable framework for detecting wireless network vulnerabilities. Based on semantic analysis and named entities recognition, we design rules for automatic device identification of wireless access points and routers. The rules are automatically generated based on the information extracted from the admin webpages, and can be updated with a semi-automated method. To detect the security status of devices, WNV-Detector aims to extract the critical identity information and retrieve known vulnerabilities. In the evaluation, we collect information through web crawlers and build a comprehensive vulnerability database. We also build a prototype system based on WNV-Detector and evaluate it with routers from various vendors on the market. Our results indicate that the effectiveness of our WNV-Detector, i.e., the success rate of vulnerability detection could achieve 95.5%.

无线路由器的安全性受到越来越多的安全威胁的关注。在物联网时代，许多设备都存在安全漏洞，因此非常需要分析设备的当前安全状态。在本文中，我们开发了WNV-Detector，这是一种用于检测无线网络漏洞的通用且可扩展的框架。基于语义分析和命名实体识别，我们设计了用于无线接入点和路由器的自动设备识别的规则。规则是根据从管理网页中提取的信息自动生成的，并且可以使用半自动方法进行更新。为了检测设备的安全状态，WNV-Detector旨在提取关键身份信息并检索已知漏洞。在评估中，我们通过网络搜寻器收集信息，并建立一个全面的漏洞数据库。我们还构建了一个基于WNV-Detector的原型系统，并使用市场上各种供应商的路由器对其进行了评估。我们的结果表明，我们的WNV检测器的有效性，即漏洞检测的成功率可以达到95.5％。