Font Size: a A A

Research On Security Authentication Methods Based On Users Behavior Characteristics

Posted on:2021-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:Z C ChengFull Text:PDF
GTID:2428330605981161Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the development of technology,mobile phones have become an indispensable tool in daily activities such as mobile payment and social interaction.Identity authentication is an effective measure to protect the security of people's privacy and property,and is also a focus field of current security research.Among the current mainstream security authentication methods,the digital password requires users to keep in mind a set of numbers,and needs to input the set of numbers actively in each authentication,which is not friendly in terms of user interaction.Authentication methods,such as fingerprint and face recognition,may cause privacy disclosure.These authentication methods belong to passive authentication,which requires the user to initiate authentication operations.Moreover,these authentication methods can only authenticate the user once,and cannot continuously verify whether the user's identity is valid.Aiming at the problems of existing security authentication methods,based on the analysis of user behavior characteristics,this paper proposes a security authentication framework named AABC(Authentication Architecture based on Behavior Characteristics)based on user behavior characteristics.The framework includes two identity authentication methods,i.e.,the authentication method based on user gait behavior and the authentication method based on user trajectory behavior.These two identity authentication methods are transparent to the users and can continuously authenticate the current user.The main research contents of this paper are as follows.(1)This thesis proposes a security authentication method based on the user's gait behavior characteristics.Based on the different gait of people,this paper proposes a safety authentication method based on the user's gait behavior characteristics.Firstly,this method uses the built-in triaxial acceleration sensor to collect the user's gait data.After that,by performing preprocessing operations such as noise reduction filtering and data standardization on the original gait data,the spectral features of the original gait data in the frequency domain are extracted to form the feature vectors.Finally,this method uses the SVM(Support Vector Machine)classifier to periodically classify the feature vectors,and transfers the classification results to the dynamic reward and penalty model(DRPM)to judge whether the user is valid.According to the low frequency characteristics of different user gaits,the authentication method can effectively authenticate the validity of a user.(2)This thesis proposes a security authentication method based on user trajectory behavior characteristics.According to the personalized characteristics of user trajectory data,this paper proposes an identity authentication system called TIAM(Trajectory-based Identity Authentication Method)based on trajectory behavior characteristics.TIAM includes two stages,i.e.,the registration stage and the identity authentication stage.In the registration phase,users set up multiple travel routes and predefined point coordinates according to their travel habits to initialize the template library,thereby constructing template trajectory features and predefined point features.In the identity authentication stage,firstly,TIAM uses the PMP(Proportion of Matching-similarity)method proposed in this paper to calculate the maximum similarity between the sample trajectory and the template trajectories,and uses the Stay-points Searching Algorithm(SSA)to calculate the hits between the stay points within the sample trajectory and the predefined points.Secondly,the TATU(sample Trajectory Authentication and template Trajectory Update)algorithm judges whether the current user is legal based on these two identity features.The TATU algorithm can verify the validity of sample trajectories,and use the verified sample trajectories to expand the template library,so that the template library can dynamically match the diversity of user's travel trajectory.According to the travel trajectory characteristics and stay-point characteristics of users,the authentication method can effectively authenticate the validity of a use.The above two authentication methods proposed in this paper have the advantages of high accuracy and low costs,and can continuously authenticate users'identities.They can be used for security authentication of mobile devices and safety behavior monitoring of the elderly and children.
Keywords/Search Tags:Identity Authentication, Behavior Characteristics, Gait Recognition, Trajectory-based Authentication
PDF Full Text Request
Related items