基于深度学习算法的无线通信网络入侵路径辨识

Wireless communication network intrusion path identification based on deep learning algorithms

由于不同类型的入侵攻击意图差异,在网络中的入侵路径也存在明显的不同,为此,文章提出了基于深度学习算法的无线通信网络入侵路径辨识研究。以入侵意图为基础,将具有相同行为的攻击进行交集计算后,建立以入侵意图为中心的网络入侵行为聚类。对于存在交叉关系的入侵,利用网络入侵行为的节点分布概率进行细分。在入侵路径辨识阶段,引入灰色深度神经网络算法,在隐含层计算入侵意图。结合无线通信网络节点与当前入侵行为所处节点之间的欧氏距离,将对应聚类的最小欧氏距离节点作为入侵路径下一节点的辨识结果,循环上述操作,实现对整体入侵路径的辨识。在测试结果中,整体入侵路径覆盖节点对于辨识结果并未产生影响,能够实现对入侵路径的100%准确辨识。

Due to the different intentions of different types of intrusion attacks,there are also significant differences in the intrusion paths in the network.Therefore,a research on wireless communication network intrusion path identification based on deep learning algorithms is proposed.Based on intrusion intent,attacks with the same behavior are intersected and calculated,and a network intrusion behavior cluster centered on intrusion intent is established.For intrusions with cross relationships,the node distribution probability of network intrusion behavior is used for segmentation.In the stage of intrusion path identification,the grey deep neural network algorithm is introduced to calculate the intrusion intent in the hidden layer.Combining the Euclidean distance between wireless communication network nodes and the node where the current intrusion behavior occurs,the minimum Euclidean distance node corresponding to the cluster is used as the identification result of the next node in the intrusion path.The above operation is repeated to achieve the identification of the overall intrusion path.In the test results,the overall intrusion path coverage nodes did not have an impact on the identification results,and can achieve 100%accurate identification of the intrusion path.