中文摘要：

云计算应用日趋广泛,其安全性问题也在日益凸显。由于网络安全攻击行为的升级,传统的攻击检测技术已不适用于云计算安全防护。本文基于云计算平台特点和安全需求以及K-means攻击检测方法,提出了一种基于进化算法的聚类技术,并设计了一种面向云计算平台的分布式攻击检测改进型方法。使用该方法,可以更好的获得聚类初始中心值,提高攻击检测率,并且加快攻击检测速度。最后,进行了实验仿真并分析了测试结果,验证其具有较好的实用价值。

英文摘要：

With the cloud computing applications are used widely, its security issues are also become increasingly important. Due to the upgrading of network security attacks, it was not proper to use the traditional attack detection technology as the security method for cloud computing. Therefore, this paper provides an new method for the cloud calculation protection, which based on platform charac- teristics and security requirements, named the improved K - means attack detection method. It also gives a clustering technique based on evolutionary algorithm, and then designs a test measure, which can find distributed attack detection for computing platform. Using the method, it can obtain much better initial clustering center value; it also can increase the rate of attack detection. It （：an accelerate the speed of attack detection. Finally, the method is simulated and analyzed, and the test results are analyzed, it shows that the im- proved method has a good practical value.