网络威胁态势预测的抗体虚警概率阈值估计

发布时间：2018-08-03 08:24

【摘要】：对网络攻击信号检测中,需要对网络威胁态势预测的抗体虚警概率阈值估计,提高攻击信号检测性能。传统方法采用伪随机时频跳变的信息容量估计方法实现阈值估计和攻击信号检测,当攻击信号为非线性谐振信号时,检测性能不好。提出一种改进的网络威胁态势预测的抗体虚警概率阈值估计算法。构建多路复用器输入输出网络威胁态势预测算法,通过病毒信息特征预处理为免疫性分析提供信息特征数据基础。得到网络跳变向量和观测向量的标准正态阈值,将网络接收端和发送端数据的信息熵作为信道传递向量函数,在人工免疫进化过程中实现对免疫节点的抗体的虚警概率阈值估计,提高网络攻击信号的检测性能,降低网络威胁态势抗体虚警概率。仿真实验得出,该算法得到的估计结果精确,有效提高了攻击信号的检测性能,确保网络安全。
[Abstract]:In the detection of network attack signal, it is necessary to estimate the threshold of false alarm probability of antibody to predict the situation of network threat, so as to improve the detection performance of attack signal. The traditional method uses pseudo-random time-frequency jump information capacity estimation method to realize threshold estimation and attack signal detection. When the attack signal is nonlinear resonance signal, the detection performance is poor. An improved threshold estimation algorithm of false alarm probability for network threat situation prediction is proposed. An input-output (I / O) network threat prediction algorithm for multiplexer is constructed, which provides information feature data basis for immune analysis through virus information feature preprocessing. The standard normal threshold of the network jump vector and observation vector is obtained, and the information entropy of the data at the receiving end and the sending end of the network is taken as the channel transfer vector function. In the process of artificial immune evolution, the false alarm probability threshold of antibody against immune nodes is estimated, and the detection performance of network attack signal is improved, and the false alarm probability of antibody against network threat situation is reduced. The simulation results show that the estimation result is accurate and the detection performance of attack signal is improved effectively, and the network security is ensured.
【作者单位】：中原工学院软件学院;中原工学院计算机学院;
【分类号】：TP393.08

【参考文献】