Authenticating mobile phone users using keystroke analysis

Mobile handsets have found an important place in modern society, with hundreds of millions currently in use. The majority of these devices use inherently weak authentication mechanisms, based upon passwords and PINs. This paper presents a feasibility study into a biometric-based technique, known as keystroke analysis - which authenticates the user based upon their typing characteristic. In particular, this paper identifies two typical handset interactions, entering telephone numbers and typing text messages, and seeks to authenticate the user during their normal handset interaction. It was found that neural network classifiers were able to perform classification with average equal error rates of 12.8%. Based upon these results, the paper concludes by proposing a flexible and robust framework to permit the continuous and transparent authentication of the user, thereby maximising security and minimising user inconvenience, to service the needs of the insecure and evermore functional mobile handset.