摘要
网络入侵时间存在较强的无序性,与网络信道受损的关系比较复杂。传统的方法只能通过简单的线性关系对两者之间的关系进行描述,一旦入侵时间复杂程度增加,将会造成入侵时间非线性关联性急剧增加,导致模型的可信度降低。为此提出一种基于约束聚类算法的网络入侵时间与网络信道受损关系建模方法。根据模糊聚类相关原理建立网络入侵时间约束模糊聚类的目标函数,对该目标函数进行最小化变换处理,计算不同入侵时间次数的权重和对应的置信度,根据计算结果建立两者关系的模型。实验结果表明,利用改进算法进行入侵时间与信道受损关系建模,能够获取不同入侵次数对于信道受损程度的影响,保证了网络的安全。
Strong disorder exists in network intrusion time, thus damaged relationship with network channel is more complex. A modeling method of relationship of network intrusion time and network channel damaged was presented based on a clustering algorithm with constraint. According to the principle of fuzzy clustering, the objective function of fuzzy clustering constrained by network intrusion time was built. Then, the objective function was minimized, the weight and the number of different invasion time corresponding confidence level were calculated, and relationship between the models was established according to the calculation results. Experimental results show that the improved algorithm can obtain the influence of different number of invasion damage for the channel and ensure the safety of net-work.
出处
《计算机仿真》
CSCD
北大核心
2015年第1期327-330,共4页
Computer Simulation
关键词
网络入侵
模糊聚类
受损
关系建模
Network invasion
Fuzzy clustering
Damaged
Relationship modeling