Implication Authentication Algorithm Utilizing User Keystroke Features Research

In recent years,people pay more attention to the problem of network information security with the rapid development of internet technology.In order to further improve the network system security,biometric authentication technologies having higher security are gradually replacing traditional authentication methods.As a new technology of biometrics,keystroke authentication has its predominance among other biometric technologies,which has made it become a research hotspot at home and abroad.Keystroke authentication owns broad development prospects and commercial value.In this paper,for the problems of authentication for fixed text and free text,the keystroke features extraction and keystroke authentication algorithm for fixed text and free text are researched,respectively,to further improve the performance of authentication system.Firstly,by combining the traditional one-factor authentication method utilizing username and password with the keystroke authentication technology for fixed text,a new two-factor authentication system is designed and a novel authentication algorithm based on the keystroke feature curve diversity degree is also proposed.The keystroke feature curve diversity degree is based on the traditional keystroke feature vector used in traditional methods and introduces the changing rate of any two adjacent keystroke features.Therefore,the algorithm based on the keystroke feature curve diversity degree can increase the distinguish degree of keystroke features and improve the accuracy of authentication.Secondly,the keystroke features extraction and a novel authentication algorithm based on the weighted keystroke feature curve diversity degree are researched in this paper,for the problem of authentication for free text.The keystroke interval time of some double keys frequently used in free text and some important semi-timing features are extracted as the free text keystroke features.Considering the difference between free text and fixed text,based on the keystroke feature curve diversity degree for fixed text,the concept of weighted keystroke feature curve diversity degree for free text is proposed.The weighted keystroke feature curve diversity degree introduces the each double key frequency of use,thus,the authentication algorithm based on weighted keystroke feature curve diversity degree can effectively solve the problem of authentication for free text.Finally,some experiments are designed to demonstrate the effectiveness of the proposed authentication algorithms for fixed text and free text,respectively.And the experiment results are compared with that of some traditional authentication algorithms at present.According to the experiment results,it has been proved that the proposed authentication algorithms in this paper have better performance than traditional authentication algorithms.