摘要
为解决以LTE-5G(Long Term Evolution-5G)为代表的新一代移动通信技术存在的网络恶意节点检出效率较低、流量查证识别能力较弱等不足,提出了一种基于多维参数采样判决机制的LTE-5G网络恶意节点检出算法.首先,针对网络恶意节点行为所产生的外溢流量特点,根据节点时域抽样特性及指纹特点,设计了多维参数采样判决机制,该机制通过采样鉴权序列定向匹配节点时域特征并筛除不符合特征的疑似恶意节点,从而达到较高的检测效果.其次,为进一步清除潜伏状态的恶意节点,结合微分机制,构建黑洞数据鉴权阈值,通过该阈值控制网络清洗流量,仅将高于该阈值的疑似节点予以清洗处理,从而降低误判情形,改善因节点离线而导致网络传输性能出现下降的现象.仿真实验表明,与当前常用的基于神经网络的加权投票鉴权机制的数据清洗方案和基于鲸鱼-狮子联合优化机制的数据清洗方案相比,本文算法具有更高的网络传输带宽和恶意节点检出率,以及更低的网络节点离线频次,在实践中具有较高的推广价值.
In order to address the shortcomings of low network malicious node detection efficiency and weak traffic verification and identification ability in the new generation mobile communication technology represented by LTE-5G,a LTE-5G network malicious node detection algorithm based on multi-dimensional parameter sampling and decision-making mechanism is proposed.Firstly,in response to the characteristics of the overflow traffic generated by malicious node behavior in the network,a multidimensional parameter sampling decision mechanism is involved based on the time-domain sampling characteristics and fingerprint characteristics of nodes.This mechanism matches the time-domain features of nodes through sampling authentication sequences and filters out suspected malicious nodes that do not match the features,thereby achieving a high detection effect.Subsequently,in order to further eliminate malicious nodes in a latent state,a black hole data authentication threshold was constructed using a differential mechanism.The network cleaning traffic was controlled through this threshold,and only suspected nodes above the threshold were cleaned,thereby reducing false positives and improving the phenomenon of network transmission performance degradation caused by offline nodes.Simulation experiments show that compared with the commonly used data cleaning schemes based on neural network-based weighted voting authentication mechanism and whale lion joint optimization mechanism,the algorithm proposed in this paper has higher network transmission bandwidth and malicious node detection rate,as well as lower offline frequency of network nodes.It has obvious promotion value in practice.
作者
王来兵
Wang Laibing(Chuzhou Vocational and Technical College,Chuzhou,Anhui 239000,China)
出处
《伊犁师范大学学报(自然科学版)》
2024年第4期69-76,共8页
Journal of Yili Normal University(Natural Science Edition)
基金
安徽省重点科研项目(2023AH053091)
安徽省重大科研项目(2023AH040388)
安徽省职成教重点项目(Azcj2022043)
安徽省职成教一般项目(AZCJ2024219)
安徽省一般教学研究项目(2022jyxm1141)
滁州职业技术学院校级科研项目(YJY-2021-12).
关键词
LTE-5G网络
多维特征
采样鉴权
黑洞清洗
流量挖掘
LTE-5G network
multidimensional features
sampling authentication
black hole cleaning
traffic mining
