|
The use of mobile communication devices with advance sensors is growing rapidly. These sensors are enabling functions
such as Image capture, Location applications, and Biometric authentication such as Fingerprint verification and Face &
Handwritten signature recognition. Such ubiquitous devices are essential tools in today's global economic activities
enabling anywhere-anytime financial and business transactions. Cryptographic functions and biometric-based
authentication can enhance the security and confidentiality of mobile transactions.
Using Biometric template security techniques in real-time biometric-based authentication are key factors for successful
identity verification solutions, but are venerable to determined attacks by both fraudulent software and hardware. The
EU-funded SecurePhone project has designed and implemented a multimodal biometric user authentication system on a
prototype mobile communication device. However, various implementations of this project have resulted in long
verification times or reduced accuracy and/or security.
This paper proposes to use built-in-self-test techniques to ensure no tampering has taken place on the verification process
prior to performing the actual biometric authentication. These techniques utilises the user personal identification number
as a seed to generate a unique signature. This signature is then used to test the integrity of the verification process. Also,
this study proposes the use of a combination of biometric modalities to provide application specific authentication in a
secure environment, thus achieving optimum security level with effective processing time. I.e. to ensure that the
necessary authentication steps and algorithms running on the mobile device application processor can not be undermined
or modified by an imposter to get unauthorized access to the secure system.
|
