Abstract:

The current frequent occurrence of cyberspace security incidents has resulted in huge losses to national security and the real economy, demonstrating that the information security threats confronting nations have transcended the traditional concept of invasion warfare. Therefore, network security vulnerability scanner is an important means to prevent network attacks. Vulnerability scanners currently on the market are usually designed using brute?force scanning, which has problems such as limited detection dimension, slow speed and low accuracy. This paper proposes a distributed multi?dimensional assessment and detection model using Docker technology for multi?node deployment and simultaneous information collection. It divides information into multiple dimensions and quantifies them. The model introduces a fuzzy hierarchical evaluation method to assess the vulnerability values of target systems, and enhances the attention to corresponding systems based on their vulnerability levels. It combines fingerprinting technology with vulnerability detection methods. Tests conducted using a scenario?based Combat Network Shooting Range (CFS) show a significant improvement in detection efficiency compared to commonly used enterprise?level network scanners, outperforming traditional one?dimensional vulnerability detection methods in terms of hit rate and efficiency.

Key words: Network security, Vulnerability scanning, Multi?dimensional information collection, Distributed, Docker

摘要：

网络入侵已经超出传统战争入侵的概念，频发的网络空间安全事件不仅使国家安全受到威胁，而且给实体经济造成巨大损失。网络漏洞扫描器是防范网络遭受攻击的重要工具，目前市场上的漏洞扫描器通常采用暴力扫描方式设计，存在检测维度有限、速度慢和准确率低等问题。因此，提出一种基于分布式的多维度评估检测模式，采用目前最新的应用容器引擎（Docker）技术实现多节点部署，通过并发式方法差异化收集信息的同时，将信息划分为多个维度并量化；引入模糊综合层次评价法对检测目标系统进行脆弱值评定，并根据脆弱值评定结果提升对应系统的关注度，采用脆弱等级与指纹识别相结合的方法进行漏洞检测；使用场景化实战网络靶场（CFS）进行测试，以评估其在真实攻防环境下的检测性能与适用性。测试结果表明，该检测模式对单个系统的检测效率与最常用的企业级网络扫描器相比有显著提升，其主要技术指标优于传统单维度漏洞检测方法。

关键词: 网络安全, 漏洞扫描, 多维度信息收集, 分布式, Docker