节点文献

一种基于EA-BinGRU算法的网络入侵检测算法

AN INTRUSION DETECTION ALGORITHM FOR NETWORKS BASED ON EA-BINGRU ALGORITHM

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 周明月; 常明航; 付殿臣; 程超;

【Author】 Zhou Mingyue;Chang Minghang;Fu Dianchen;Cheng Chao;School of Computer Science and Engineering, Changchun University of Technology;

【机构】 长春工业大学计算机科学与工程学院;

【摘要】 根据大多数通过手动预处理网络数据包无法正确表征网络流量特性的缺点,提出自动编码变体门控循环单元,通过自动预处理数据包进行有效负载感知入侵检测。自动编码变体门控循环单元指自动编码的二进制门控循环单元。此外,自动编码的二进制门控循环单元减少了所需的内存大小,并用二进制加法取代了大多数算术运算。基于ISCX2012公共入侵检测数据集的实验结果表明,自动编码的二进制门控循环单元的准确率为99.8%,检测率为99.8%。门控循环单元的内存使用量约为自动编码的二进制门控循环单元的32.8倍。

【Abstract】 According to the shortcomings of most network packets preprocessed manually that cannot accurately characterize network traffic characteristics, an automatic coding variant gated recurrent unit is proposed. Payload aware intrusion detection was carried out by automatically preprocessing packets. The automatic coding variant gated recurrent unit referred to the binarized gated recurrent unit of automatic coding. In addition, the automatically encoded binarized gated recurrent unit reduced the required memory size and replaced most arithmetic operations with binary addition. The experimental results based on the ISCX2012 public intrusion detection data set show that the accuracy of the automatically encoded binarized gated recurrent unit is 99.8%, and the detection rate is 99.8%. The memory usage of the gated recurrent unit is approximately 32.8 times that of the binarized gated recurrent unit with automatic encoding.

【基金】 吉林省发改委基金项目(2019C040-3)
  • 【文献出处】 计算机应用与软件 ,Computer Applications and Software , 编辑部邮箱 ,2023年11期
  • 【分类号】TP393.08
  • 【下载频次】47
节点文献中: 

本文链接的文献网络图示:

本文的引文网络