算法生成恶意域名的实时检测

摘要

Abstra ct:The current detection of algorithmically generated malicious domain has long detecting cy-cle, and has no sufficient use of known and pub-lished malicious domain. Aimed at this problem, we propose a real time detect method based on relation-ship between new domain name and known mali-cious domains combined with live pattern of algo-rithmically generated domains. The experiment result shows that the method can quickly and effectively filter out algorithmically generated domains used by malware, but this method consumes a large amount of computation and memory resources that needs to be solved in the following work.%　　当前对算法生成域名技术的检测，检测所用时间周期过长，无法对算法生成的恶意域名进行快速检测。针对此问题，本文基于新增域名与已分类恶意域名之间的关联关系，提出一种算法生成域名的实时检测方法，并在某省运营商DNS服务器机房部署本系统，实验验证本检测方法。实验表明与已有方法相比，本方法能够快速筛选用于恶意网络行为的算法生成域名。但本方法需要消耗大量的计算资源和内存资源，需要在后续的工作中研究解决。