Research On Device Fingerprinting Generation Algorithm Based On Implicit Information

Traditional methods of device identification are mostly based on explicit identifiers,such as Cookies,IP address,MAC address and so on.However,with the improvement of the security awareness of the Internet users,acquisition of these explicit identifiers has been limited.Therefore,it is necessary to study new device identification methods to solve the limitations of traditional identification methods that the inconvenience brought.A reliable solution is provided by extracting the device fingerprinting generated by combination of implicit information with identification capability in the device.Device fingerprinting has attracted much attention because of its high accuracy,and it has become an effective means of device identification.Due to the high application value of device fingerprinting in the field of fraud prevention,user authentication and information security,the device fingerprinting has become a new research focus in the academic field.However,due to the instability of the fingerprinting,the existing device fingerprinting generation algorithm still needs to be improved.The implicit information in the fingerprinting is easy to change,which will change the original device fingerprinting.Therefore,the device fingerprinting generation algorithm should focus on the timely updating of device fingerprinting.Aiming at the problem of device fingerprinting update,this paper has done the following research work:1.Based on the stability of the device fingerprinting,this paper analyzes the shortcomings of the existing fingerprint generation algorithm,and proposes a new updating algorithm of device fingerprinting based on association matching.The algorithm uses the computation similarity of edit distance to complete the matching and correlation of fingerprints,and then completes the update operation of the fingerprinting.The experimental results show that the algorithm can improve the accuracy of fingerprinting recognition and can be applied to the change of single implicit information in the fingerprinting.2.In this paper,a new updating algorithm of device fingerprinting based on difference degree is proposed to solve the problem of insufficient fingerprinting update accuracy.This algorithm quantifies the difference of the implicit information in the fingerprinting,and calculates the difference value between two fingerprints through the global difference function.And then the fingerprinting is updated by the replacement operation.Finally,the fingerprinting data are tested under different threshold of difference degree.The experimental results show that the algorithm performs well in the case of single or multiple implicit information changes.3.In this paper,the device fingerprinting change problem is transformed into the fingerprinting classification problem,and a fingerprinting update algorithm based on naive Bias classifier is proposed.The algorithm classifies the fingerprinting by the naive Bias classifier,and the fingerprints are correlated to the same category before and after the change,and then implements the update of the device fingerprinting,thereby improving the original algorithm of device fingerprinting and solving the instability problem.In this paper,the method of fingerprinting generation is deeply studied,and an innovative work is done on the device fingerprinting generation algorithm.Finally,the online fingerprinting data sets verify that the algorithm in this paper has achieved a higher evaluation index,and can effectively solve the fingerprinting change problem.