摘要
网络撞库攻击是一种从数据库中导出数据的攻击方式,通过网站入侵,非法实现对用户信息的窃取和修改,如何更好提高网络安全,提出一种改进的网络撞库攻击信息特征潜在博弈欺骗鉴别算法。首先构建信号模型,采用博弈论方法,对攻击行为的欺骗性进行鉴别,得到网络区分服务等级的服务质量量化函数,从而实现对撞库攻击信号的欺骗性鉴别,利用非单调性决策博弈方法,给出网络威胁离散度状态方程,得到骗性判别的鉴别函数,统计撞库攻击行为的参与者,构建接入网络的优服务质量函数,提取出有用的规则性异常数据特征,并结合后置分类处理和数据处理,实现对攻击信号的准确检测和欺骗信号的鉴别。仿真实验表明,采用该方法,能有效鉴别出网络撞库攻击信号的实质信息特征和欺骗信息特征,对攻击信号的检测性能优越,提高了鉴别准确率,提高了网络服务的质量。
Network hit the base attack from a database derived data an attack mode, invaded after the website, hackers steal its database, to realize user information stealing and modification. This paper proposes an improved network attack hit feature library information potential game cheating identification algorithm, is of great significance in the design of network security. First, construct the signal model, using game theory, the attack on the behavior of the deception of the identification, obtain service quality quantization function network differentiated service level, uses the game theory to realize the deceptive differential impinging library attack signal method, using non monotonic decision-making game method, given a network threat dispersion equation of state, get the diseriminant function deceive. Identification, participants statistics hit the storehouse of the attack, excellent service quality function construction of the access network, extract the characteristic rules of abnormal data useful, and combined with the post classification processing and data processing, realize the accurate detection and identification of deception signal to attack the signal. Simulation results show that using this method, can effectively identify network attacks hit the base signal substance characteristics of information and deception information characteristics, and improve the quality of network service, to improve the accuracy of identification, the detection performance of the signal to attack the superiority.
出处
《科技通报》
北大核心
2015年第8期144-146,共3页
Bulletin of Science and Technology
关键词
撞库攻击
信息特征
博弈
鉴别算法
检测
hit the base attack
information characteristics
game
identification algorithm
detection
