基于Merkle哈希树的通信网络异常数据流自动识别方法  

Automatic identification method for abnormal data flow in communication networks based on Merkle hash tree

作　　者：孙树垚 SUN Shuyao(Information Center of National Radio and Television Administration,Beijing 100866,China)

机构地区：[1]国家广播电视总局信息中心,北京100866

出　　处：《无线互联科技》2025年第1期15-18,23,共5页Wireless Internet Science and Technology

摘　　要：为解决通信网络在高异常数据比例下识别模型性能退化、F1值偏低的问题,文章设计了一种基于Merkle哈希树的通信网络异常数据流自动识别方法。对通信网络数据流进行预处理,以消除噪声和冗余信息;通过设定合理的阈值对异常评分进行比较,从而提取出关键的异常数据流特征;利用Merkle哈希树,建立一个高效的异常数据识别模型,该模型通过哈希树节点的连接机制,在层级间有效传递信息,实现对通信网络中异常数据流的自动识别。实验结果表明,相较于传统方法,基于Merkle哈希树的通信网络异常数据流自动识别方法在各个数据集上的表现更为优越,特别是在数据集4中,取得了0.93的F1值,证明该方法识别通信网络异常数据流的准确性和可靠性,几乎无遗漏。In order to solve the problem that the performance of the identification model of communication network is degraded and the F 1 value is low under the high proportion of abnormal data,an automatic identification method of abnormal data flow in communication network based on Merkle hash tree is designed.The data stream of communication network is preprocessed to eliminate noise and redundant information.By setting a reasonable threshold to compare the abnormal scores,the key abnormal data flow characteristics are extracted.Using Merkle hash tree,an efficient abnormal data identification model is established.The model effectively transmits information between levels through the connection mechanism of hash tree nodes,and realizes the automatic identification of abnormal data flow in the communication network.The experimental results show that,compared with the traditional methods,the Merkle hash tree based method for automatic identification of abnormal data flow in communication network performs better on each data set,especially in data set 4,which achieves an F 1 value as high as 0.93,which proves the accuracy and reliability of this method for identification of abnormal data flow in communication network,with almost no omission.

关 键 词：Merkle哈希树 通信网络 哈希树节点 异常数据流 自动识别 

分 类 号：TN913[电子电信—通信与信息系统]