Research And Implementation Of Identity Authentication Method Based On Finger Operation Behavior Of Mobile Terminal

With the rapid development of mobile Internet and mobile payment in recent years,the value of mobile terminals is increasing.Driven by the huge black gray interests,malicious attackers began to attempt to invade mobile terminals and steal them.In the face of endless malicious intrusion methods,how to authenticate users' identity has become an important research content to protect users' property.At present,most of the researches on the identity authentication of finger operation behavior are focused on the process of user touching the screen,extracting the characteristics of the behavior data generated when the user touches the screen,and using various detection algorithms for comparison or classification.Because the behavior data generated by touching the screen is easy to obtain,this method can quickly and intuitively carry out user identity authentication.However,the number of user features extracted from touch screen behavior is small,which leads to low detection accuracy and unsatisfactory authentication effect.Therefore,how to increase the user behavior characteristics and improve the accuracy of user recognition has become an important research direction.In this paper,we focus on the research of user's mobile phone shaking and the realization of identity authentication with the help of gyroscope sensor.The research results of this paper are mainly composed of the following four aspects:Firstly,process of user touching the screen is studied in detail.The regularity of single user touching the screen and the difference of multiple users touching the screen are studied respectively.The "gyroscope slope"which is physically defined as the shaking acceleration of mobile phone is proposed as a new user-behavior feature.Secondly,with the help of gyroscope sensor and sliding window algorithm,the specific time when the finger starts to click and the end of lifting is predicted,which can expand the research scope from touch screen to the whole click process.The whole click process includes"finger down click stage","touch screen stage" and "finger out of screen stage".Morover,the user's whole click process is studied deeply,and the user behavior characteristics of press and lift which can't be acquired in the past are acquired.Thirdly,the specific experiment of finger operation authentication algorithm is designed,and the overall architecture and the specific implementation of each functional module are described.Finally,the experiment is carried out and the experimental results are analyzed in detail.In the "Click To study" scenario,the method of grouping user data by region is proposed,and the low shaking region is filtered by using the gyroscope to judge the left and right hands.Experimental results show that the proposed method can effectively improve the detection rate and recognition rate of users.