摘要
为了防止内容相关型的网络入侵,从网络攻击对象的角度,利用蜜罐技术,提出了信息域自反馈网络入侵防护模型.该模型能够忽略攻击手段和方式而直接判断受到威胁的信息及所属领域,将信息根据其领域、内容等属性划分为信息域,在诱捕环境下将入侵者对信息域攻击的危害程度进行量化,获得域信息敏感度,以此作为评价数据是否需要重点防护的依据.网络场景实验表明,模型实现了对遭受网络攻击或威胁的信息及相关领域的预测,封闭数据集测试模型准确率高达96.56%,开放环境下其平均准确率为84.76%.
To prevent text-related network invasion from the perspective of a network attack object by making use of honey-pot technology, an information domain self-feedback invasion prevention model was presented for contentbased network intrusion which is able to directly judge the threatened information and the corresponding field and ignore the ways and means of the network attacks. First of all, the data for the networks was divided into domains according to the fields, content, and other attributes. In the trapping environment, the damaged extent of the domain suffering from an attack by the intruder would be quantified, through which domain information sensitivity would be achieved, which would be taken as the basis for the evaluation of data protection. Through the experiments in the actual network scene, the results show this to be the successful model which forecasts the information and relative domain suffering from destruction or threat of network attacks. The model attains as high as 96.56 percent accuracy rate for the closed sets test, and the average accuracy rate is 84.76% in the open environment;information hiding technology
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2010年第11期1496-1502,共7页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(60673131)
中央高校基础科研基金资助项目(HEUCF100602)
关键词
自反馈
防护模型
诱捕
信息域
网络入侵
信息隐藏技术
self-feedback
prevention model
trapping
information domain
network intrusion
information hiding technology