A Computationally Efficient Fingerprint Matching Algorithm for Implementation on Smartcards

Biometric methods have established as the pinnacle of reliable authentication, but while able to overcome the limitations of traditional methods, many social issues (e.g., the methods of preserving privacy and maintaining security) require resolution. With the ability to securely store and execute data, smartcard technologies are increasingly playing an important role in establishing highly usable, secure and trusted solutions, but due to technological constraints computationally intensive applications are presently restricted or prohibited. The implementation of biometric methods is a challenging proposition: to date, existing solutions address this problem by decoupling the matching algorithm, but this only serves to reintroduce significant security vulnerabilities. With the primary aim of reducing the complexity of computation and so enabling the development of a solution which enhances data security, this paper presents a fingerprint representation/encoding scheme, which through an adaptation to the generic architecture can establish a disturbance rejection methodology capable of differentiating between equivalent and extraneous data. Initial experimental results demonstrate this novel concept addresses practical system limitations by reducing the affects of measurable disturbances, yielding a significant reduction in the magnitude of computation, so aiding the development of truly secure solution based within resource-constrained environments.