基于遗传算法和灰色关联分析的击键特征识别算法  

User authentication algorithm with keystroke features based on genetic algorithms and grey relational analysis

作　　者：王晅[1] 陈伟伟[1] 马建峰[2] 

机构地区：[1]陕西师范大学物理学与信息技术学院,陕西西安710062 [2]西安电子科技大学计算机网络与信息安全教育部重点实验室,陕西西安710071

出　　处：《计算机应用》2007年第5期1054-1057,共4页journal of Computer Applications

基　　金：国家863计划资助项目(2002AA143021)

摘　　要：基于用户击键特征的身份认证比传统的基于口令的身份认证方法有更高的安全性,现有研究方法中基于神经网络、数据挖掘等算法计算复杂度高,而基于特征向量、贝叶斯统计模型等算法识别精度较低。为了在提高识别精度的同时有效降低计算复杂度,在研究现有算法的基础上提出了一种基于遗传算法与灰色关联分析的击键特征识别算法。该算法利用遗传算法根据用户训练样本确定表征用户击键特征的标准特征序列,通过对当前用户击键特征序列与标准特征序列进行灰色关联分析实现用户身份认证。实验结果表明,该算法识别精度达到神经网络、支持向量机等算法的较高水平,错误拒绝率与错误接受率分别为0%与1.5%。且计算复杂度低,与基于特征向量的算法相近。User authentication based on keystroke dynamics features is more secure them conventional user authentication approach only based on passwords. The neural network and data mining-based methods present high authentieation aeeuraey, but have a high eomputational eost. The statistieal and veetor-based methods have shown low eomputational eomplexity, but are less aeeurate in user authentieation. In order to improve authentieation aeeuraey and reduee eomputational eomplexity synehronously, a new user authentieation approaeh based on keystroke patterns was proposed. In the proposed approaeh, Genetie algorithm was employed to generate the eommon keystroke pattern of each user from the training set consisting of the user＇s normal keystroke samples. Then Grey Relational analysis method was applied to caleulate the degree of grey slope ineidenee between eommon keystroke pattern and eurrent keystroke pattern, the resultant value was eompared with a threshold value determined by experiment to implement user authentieation. Experimental results show this approaeh represents the same user authentieation aeeuraey as neural network and data mining-based methods in terms of False Aeeeptanee Rate （FAR） and False Rejeetion Rate （FRR）, false aeeeptanee rate and false rejeetion rate of this method are 1.5% and 0% respeetively. It is also shows that the eomputational eomplexity of the proposed method is lower than that of some other methods.

关 键 词：用户身份认证 击键特征 遗传算法 灰色关联分析 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]