电信科学 ›› 2018, Vol. 34 ›› Issue (1): 72-79.doi: 10.11959/j.issn.1000-0801.2018005

• 研究与开发 • 上一篇    下一篇

云环境中改进FCM和规则参数优化的网络入侵检测方法

张春琴1,2,谢立春1   

  1. 1 浙江工业职业技术学院,浙江 绍兴 312000
    2 浙江工业大学,浙江 杭州 310014
  • 修回日期:2017-09-25 出版日期:2018-01-01 发布日期:2018-02-05
  • 作者简介:张春琴(1977-),女,浙江工业职业技术学院副教授,浙江工业大学访问学者,主要从事网络安全、云计算方面的研究工作。|谢立春(1974-),男,浙江工业职业技术学院副教授,入选浙江省“151人才工程”,主要从事网络安全方面的研究工作。
  • 基金资助:
    国家自然科学基金青年科学基金资助项目(61603211)

Network intrusion detection method based on improved FCM and rule parameter optimization in cloud environment

Chunqin ZANG1,2,Lichun XIE1   

  1. 1 Zhejiang Industry Polytechnic College,Shaoxing 312000,China
    2 Zhejiang University of Technology,Hangzhou 310014,China
  • Revised:2017-09-25 Online:2018-01-01 Published:2018-02-05
  • Supported by:
    The Young Science Foundation of National Natural Science Foundation of China(61603211)

摘要:

针对云环境中的网络入侵检测问题,提出一种基于模糊推理的网络入侵检测方法。首先,利用互信息特征选择对样本特征进行降维。然后,利用提出的改进模糊C均值聚类(IFCM)方法对训练样本集进行聚类,根据各样本特征与集群的对应关系获得初始模糊规则库。接着,对每个规则的前件参数和后件参数进行调优,以此获得准确的规则库。最后,基于规则库对输入连接数据进行模糊推理,对其进行分类以实现入侵检测。在云入侵检测数据集上的实验结果表明,该方法能够准确检测出网络入侵,具有可行性和有效性。

关键词: 云环境, 网络入侵检测, 互信息特征选择, 改进模糊C均值聚类, 模糊规则库优化

Abstract:

Aiming at the network intrusion detection problem in cloud environment,a method of network intrusion detection based on fuzzy inference was proposed.Firstly,it used the mutual information feature selection to reduce the feature of the sample.Then,the improved fuzzy C-means clustering method was used to cluster the training sample set,and the initial fuzzy rule base was got by the correspondence between each sample feature and cluster.After that,the refine parameter and consequent parameters of each rule were tuned to obtain an exact rule base.Finally,fuzzy inference was carried out on the input connection data based on the rule base,and it was classified to realize intrusion detection.Experimental results on the cloud intrusion detection dataset show that this method can detect the network intrusion accurately,and it is feasible and effective.

Key words: cloud environment, network intrusion detection, mutual information feature selection, improved fuzzy C-means clustering, fuzzy rule base optimization

中图分类号: 

No Suggested Reading articles found!