混合服务器时变网络环境中潜在威胁挖掘模型  

Potential Threat Mining Model in Hybrid Server Time-Varying Network Environment

作　　者：侯研[1] 

机构地区：[1]长春工业大学人文信息学院,吉林长春130122

出　　处：《计算机仿真》2014年第7期301-304,共4页Computer Simulation

摘　　要：研究多服务器安全服务问题,在多服务器环境下,服务器的结构属性不同,同程度的攻击对各自服务器的威胁也不同,形成时变性动态网络结构。传统的混合服务器威胁检测是假设同一攻击对每个服务器带来的威胁是相同的静态结构理论,缺少对动态网络结果特有抗攻击属性的鉴别,存在漏洞。提出了一种针对混合服务器动态网络结构的潜在威胁挖掘模型。针对混合服务器网络的动态结构,需要选取具有代表性的网络威胁特征,利用Apriori算法运算得到动态环境下支持度达到极小值的潜在威胁特征,获取频繁项的数据集合,迭代计算频繁项的全部元素构成的集合,直至得到全部入侵结果。实验结果表明,利用改进算法进行混合服务器网络中的潜在威胁挖掘,可以极大地提高挖掘的准确性,从而为混合服务器网络安全提供保障。In a multi-server environment, due to the different structural properties of the server, the threat of the same level of attack to each server is different, which results in the time-varying dynamic network structure. The traditional threat detection of hybrid server is based on the static structure theory which assumes that the threat brought by the identical attack is same for every server. It lacks the identification of peculiar anti-attack attribute of the dy- namic network. This paper presented a potential threat mining model for the hybrid server dynamic network structure. Firstly, to the dynamic structure hybrid server networks, the representative network threat characteristic is selected. Then, the potential threat characteristics are obtained by using Apriori arithmetic operations, which minimizes the support degree under the dynamic environment. The data collection of frequent item is acquired, and the collection of all the elements of frequent items is iterated. The experimental results show that the proposed algorithm in threat mining for hybrid server network can greatly improve the accuracy of mining, and thus provide protection for the hybrid server network security.

关 键 词：混合服务器 网络安全 潜在威胁 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]