基于机器学习的恶意双胞胎攻击检测

Evil twin attack detection based on machine learning

随着移动计算的普及,WiFi(wireless fidelity)已经成为人们上网的必备技术之一,WiFi安全也成为移动计算的主要威胁。恶意双胞胎攻击可以通过伪造与实际接入点相同的服务集标识符来窃取大量私有数据,威胁着人们的财产安全。因此,针对这种隐患,提出了一种模型,通过使用多种机器学习算法对恶意双胞胎攻击进行检测。首先介绍恶意双胞胎攻击的原理及危害,然后基于数据特征,分别使用K近邻、支持向量机、逻辑回归、朴素贝叶斯4种算法实现分类,最后,使用基于4种分类算法的检测模型对采集到的数据进行测试。结果显示,支持向量机分类算法准确率更高,达到96.4%。

With the popularity of mobile computing,WiFi(wireless fidelity)has become one of the necessary technologies for people to surf the Internet,and WiFi security has also become a major threat to mobile computing.Evil twin attack can steal a large number of private data by forging the same SSID(service set identifier)as the actual access point,threatening people's property security.Therefore,in view of this hidden danger,the proposed model uses a variety of machine learning algorithms to detect evil twin attacks.Firstly,the way of evils twin attack is introduced.Then,based on the data characteristics,K-nearest neighbor,support vector machine(SVM),logistic regression and naive Bayes,algorithms are used to realize classifi-cation.Finally,the detection model based on the four algorithms is used to test the collected data.The result shows that the accuracy rate of SVM classification algorithm is higher,reaching 96.4%.