基于相似系数的DDOS攻击检测研究

RESEARCH ON SIMILARITY COEFFICIENT-BASED DDOS ATTACK DETECTION

针对服务器的DDOS攻击,在分析流量基础上,提出一种利用相似系数检测DDOS攻击的算法。该算法以单位时间的数据量作为计算项,统计实验数据值,通过与阈值的比较检测DDOS攻击。此算法具有在线检测、实时性强的特点,对插入式DDOS攻击检测尤其有效。实验表明,在阈值参数取值适当的情况下,DDOS攻击的检测准确率有较高的提升。

For the DDOS attacks against servers,we propose an algorithm to detect DDOS attack by using similarity coefficient based on traffic analysis. This algorithm takes the data traffic per unit time as the calculating items and detects DDOS attack by comparing the statistical values of experimental data with the threshold values. The algorithm has the features of online detection and high real time performance. In addition,it is especially effective for the plug-in DDOS attack detection. Experiments show that the accuracy of DDOS attack detection has better improvement when the threshold parameters are properly selected.