Continuous Identity Authentication System Based On Keystroke Behavior And Mouse Behavior

As computers have become more and more widely used in people's work and life,the frequency of information security incidents has also increased.As the first barrier to the security of a computer system,the identity authentication system has a very important position.Traditional static identity authentication technology cannot adapt to the increasingly severe information security environment.Continuous identity authentication technology refers to a technology that continuously verifies a user's identity during the active session of a user.It can detect abnormal or malicious behaviors,complements the traditional static authentication technology,and jointly guarantees the security of the computer system.Human biological characteristics have the advantages of uniqueness,stability,universality,and non-replicability,so they can be used as credentials for identity authentication.Human biological characteristics can be divided into two categories: 1)Physiological characteristics,such as fingerprints,iris,etc.;2)Behavior characteristics,such as mouse behavior,keystroke behavior,etc.Physiological characteristics can be used for static identity authentication,which has the advantages of high accuracy and fast speed,but it is not suitable for continuous identity authentication.The identity authentication technology based on behavior characteristics can authenticate users' identities according to their behaviors,and is a feasible solution for realizing continuous identity authentication.This paper proposes a continuous identity authentication system based on the user's keystroke behavior and mouse behavior,which can realize the identification and authentication of the user's identity according to the behavior mode of the user using the keyboard and mouse.The main work and innovations of this paper are as follows: 1)In order to ensure the capture of the real behavioral patterns of users,this paper builds a completely free data collection environment,the collection time is up to 5 weeks,during which users can use the computer exactly as they wish.2)Exploring the optimal keystroke sequence length through experimental analysis.While ensuring that the accuracy rate is not greatly reduced,select a shorter sequence length to enhance the real-time nature of the model.3)Through experiments,a variety of model combination methods are compared and analyzed,and the combination method with the best effect is selected.4)The decision-making fusion scoring system proposed in this paper can display the credibility of user behavior in real time,can effectively detect the intrusion behavior of illegal users,and can tolerate normal behavior fluctuations of legitimate users.