摘要
针对当前通信系统入侵行为自动识别技术存在入侵信号样本识别成功率较低、误识别率和漏识别率较高的问题,提出基于GASVM算法的通信网络入侵信号自动识别技术。利用混沌原理提取通信网络入侵的非平稳信号时域特征,并凭借自回归模型提取对应频域特征,捕捉邻域入侵信号间的非线性时空动作频率,评价相邻行为间的状态关联性,预测入侵信号后续行为,完成入侵信号的识别。实验表明,所提方法识别精度高、误识别率较低,漏识别率非常低,具有可应用于实际的理论价值。
For problems of low success rate of intrusion signal sample identification,high false identification rate and high miss recognition rate in the current automatic identification technology of intrusion behavior in communication systems,a communication network intrusion signal automatic identification technology based on GASVM algorithm is proposed.The time domain characteristics of nonstationary signals in communication network intrusion are caught by applying the chaotic principle,and the autoregressive model is applied to extract the corresponding frequency domain characteristics,capture the nonlinear spatiotemporal action frequency between neighboring intrusion signals,evaluate the state correlation between adjacent behaviors,predict the followup plans of the intrusion signal,and complete the identification of the intrusion signal.Experiments show that the proposed method manifests itself with high recognition accuracy,low false recognition rate,fairly low missed recognition rate,and theoretical valuewhich can be applied to practice.
作者
杨本胜
YANG Bensheng(Guangzhou Songtian Polytechnic College,Guangzhou 511300,China)
出处
《机械与电子》
2021年第12期25-29,共5页
Machinery & Electronics
关键词
GASVM算法
通信网络
入侵信号识别
时域以及频域特征
长短期记忆模型
混沌原理
GASVM algorithm
communication network
identification of intrusion signal
time domain and frequency domain characteristics
the long and short term memory model
chaos principle