基于Merkle哈希树的异构通信网络数据异常值概率识别算法  

Probability identification algorithm of outliers in heterogeneous communication networks based on Merkle hash tree

作　　者：蒋诚智 徐浩 黄传锋 邓松[2] JIANG Chengzhi;XU Hao;HUANG Chuanfeng;DENG Song(Nanjing Institute of Technology, Nanjing 211167, China;Nanjing University of Posts and Telecommunications, Nanjing 210023, China)

机构地区：[1]南京工程学院,南京211167 [2]南京邮电大学,南京210023

出　　处：《兵器装备工程学报》2022年第6期190-195,231,共7页Journal of Ordnance Equipment Engineering

基　　金：国家自然科学基金项目(51977113);南京工程学院引进人才科研启动基金项目(YKJ201989);南京工程学院创新基金面上项目(CKJB202003);江苏省教育厅高校哲学社会科学研究重大项目(2020SJZDA069)。

摘　　要：结合Merkle哈希树算法,设计了异构通信网络数据异常值概率识别算法。该算法依据Merkle哈希树高度和数据认证恢复时间设置网络传感器,利用上升的感应数据项建立哈希树,通过网络发送器建立恢复数据包,利用网络接收端的数字编码实现网络安全认证,依据认证的网络异常数据值,设定异构通信网数据正态分布置信门限数值,通过假设检验方式建立概率分布模型,完成异构通信网络数据异常值概率识别。实验结果表明:该算法在不同门限值情况下,当网络节点攻击百分比达到0.4%,2种门限数值识别的网络节点假阳性率相差0.14%,可有效识别不同门限数值网络数据假阳性率;在数据包伪造、重放和篡改方面数据安全检测率和数据异常识别率均达到100%。Combined with Merkle hash tree algorithm,a probability identification algorithm of heterogeneous communication network data outliers was designed.The algorithm set the network sensor according to the height of Merkle hash tree and data authentication recovery time,established the hash tree by using the rising sensing data items,established the recovery data packet through the network transmitter,realized the network security authentication by using the digital coding at the network receiving end,and set the confidence threshold value of normal distribution of data in heterogeneous communication networks according to the authenticated network abnormal data value.The probability distribution model was established by hypothesis test to complete the probability identification of outliers in heterogeneous communication network data.The experimental results show that when the attack percentage of network nodes reaches 0.4%under different threshold values,the false positive rate of network nodes identified under the two threshold values is 0.14%,which can effectively identify the false positive rate of network data under different threshold values;In terms of packet forgery,replay and tampering,the data security detection rate and data anomaly recognition rate reach 100%.

关 键 词：Merkle哈希树 异构通信网络 数据异常值 概率 安全认证 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]