论文部分内容阅读
为了提升全光网络的安全稳定性,需要进行全光网络攻击检测的光纤传感器设计,但采用当前方法进行传感器设计时,无法给出攻击节点的具体位置,存在传感器设计误差大的问题。为此,提出用于全光网络攻击检测的光纤传感器。首先采用混沌理论构建全光网络攻击信号混沌系统模型,对网络状态特征集在混沌系统中的相空间进行重构,得到攻击信号的非线性时间序列,设定每类攻击行为的提取比例参数,利用支持向量机的最优检测平面设计光纤传感器,完成对全光网络攻击检测的光纤传感器设计。实验仿真证明,该方法检测精度较高,可以为保障全光网络稳定运行提供可行的依据。
In order to improve the security and stability of all-optical network, optical fiber sensor design for all-optical network attack detection is required. However, when the current method is used for sensor design, the specific location of the attack node can not be given, which poses a problem of large sensor design error. For this reason, put forward the optic fiber sensor used for all-optical network attack detection. Firstly, chaos theory is used to build the chaotic system model of all-optical network attack signal. The phase space in the chaotic system of the network state feature set is reconstructed to obtain the non-linear time series of attack signals, and the extraction proportion parameters of each type of attack are set. Optical fiber sensor is designed by the optimal detection plane of support vector machine to complete the design of optical fiber sensor for all-optical network attack detection. Experimental simulation proves that this method has higher detection accuracy and can provide feasible basis for guaranteeing the stable operation of all-optical network.