基于GA-Elman神经网络的网络态势预测模型研究

发布时间：2018-03-20 00:33

本文选题：网络安全　切入点：态势感知　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年移动互联网高速发展,给人们生活带来了极大的便利,同时也带来了严重的网络安全问题。由于现在人们的工作生活对网络的依赖性很强,网络攻击带来的损失也越来越难以承受。传统的网络安全设备都是各自独立工作的,没有综合分析各个设备之间信息的相关性,不能从整体上对网络安全态势做出评估。为了从宏观上掌握网络的安全状况以及预测未来网络态势的发展趋势,网络安全态势感知(Network Security Situation Awareness,NSSA)已成为当前网络安全领域的热门研究方向。NSSA在对网络安全的各种因素提取的基础上,对网络当前的状态进行定量评估,并利用历史态势信息预测网络态势的未来状态。能够明显提高网络管理人员对网络的把控。本文在详细研究了 NSSA模型及相关技术理论的基础上,针对网络态势信息具有时序性的特征,对标准的Elman神经网络结构进行改进,建立了基于GA-Elman的态势预测模型。由于Elman神经网络的承接层仅仅保存了隐藏层神经元的反馈信息,本文通过增加反馈节点的层数来改进Elman神经网络的结构,以使其能存储各层节点的反馈信息。在学习参数的算法方面,为了解决误差逆传播算法容易陷入局部最优的缺点,使用了具有全局搜索策略的遗传算法来对网络进行训练。最后实现了基于遗传算法的Elman网络模型来进行态势预测,通过实验证明模型能够准确的对未来的态势进行预测。
[Abstract]:In recent years, the rapid development of mobile Internet has brought great convenience to people's life, but also brought serious network security problems. The losses caused by network attacks are becoming more and more difficult to bear. Traditional network security devices work independently, without comprehensive analysis of the correlation of information between each device. In order to grasp the security situation of the network and predict the development trend of the network situation in the future, we can not evaluate the network security situation as a whole. Network Security Situation Awareness (NSSA) has become a hot research direction in the field of network security. On the basis of extracting various factors of network security, the current status of the network is evaluated quantitatively. And using historical situation information to predict the future state of network situation. It can obviously improve the network control of network managers. This paper studies the NSSA model and related technology theory in detail. In view of the temporal characteristics of network situation information, the standard Elman neural network structure is improved, and a situation prediction model based on GA-Elman is established. Because the Elman neural network's accepting layer only preserves the feedback information of hidden layer neurons. In this paper, the structure of Elman neural network is improved by increasing the number of layers of feedback nodes, so that it can store the feedback information of each layer node. The genetic algorithm with global search strategy is used to train the network. Finally, the Elman network model based on genetic algorithm is implemented to predict the situation of the network. The experimental results show that the model can accurately predict the future situation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;TP393.08

【相似文献】