Improve Wireless Application System Authentication Security by Using ICA and Factorial Face Code

In this paper we use the independent component analysis (ICA) and Factorial Face Code as an additional user identity checking method to strengthen the solidity of the wireless application system authentication security. The basic theory of ICA and factorial face code are presented firstly and then the main works of this method are detailed. After that an experiment based on this method is given. We use the face database down-loaded from the website of http://www.uk.research.att.com/ to buildup the basis images, get the user's factorial face code to encode the user's face and then check the user's identity to test the affectivity of this method by computing the user's factorial face code with different expressions. According to this experiment, we know that this method can confirm user's identity and can be an additional method to check wireless application system user identity. In order to improve the affectivity of this method, we should increase the amount of image data in the face database, getting more ICs and enlarging the size of the basic images set. All this will increase the length of the factorial face code, and subsequently will differentiate one person from another more easily.