Document Type : Research Paper
Authors
1
Msc. Student of Electrical Engineering, Amirkabir University of Technology, Tehran, Iran
2
Department of Electrical Engineering, Amirkabir University of Technology (Tehran polytechnic), Tehran, Iran
3
Department of Electrical Engineering,,Ayatollah Boroujerdi University, Boroujerd, Iran
Abstract
In this paper, we aim to use biometric fingerprint features to generate random cryptographic keys. The complexity of the key generation algorithm, the high number of bits, and the randomness are three important factors for strong cryptographic keys. In the proposed method, firstly, we extract the biometric features, i.e Minutiae points after an appropriate image processing step. Then, to increase the complexity of the key generation method as well as its security, each image is divided into 44 segments. Thus, we could increase random data by calculating the Euclidean distance and the angle between the central pixels of each 44 segmentswith all Minutiae points.Finally, we have proposed a three-step algorithm to generate the final key. It includes placing the related valuesof the angle and distance in the even and odd positions of a single matrix, two methods of bit replacement and permutation, and applying a uniform distribution on data. We could extract subkeys of 128, 256, and 512 bits due to the high number of bits of the final key which is useful in cryptography applications. The performed statistical analysis tests such as NIST prove the randomness and high security of the final 63751-bit key. According to these tests, the proposed method outperforms the previous ones in the literature which only use the distance/angle among minutiae points for generating cryptographic keys with much lower lengths. Considering the 15 tests of NIST for evaluating the key randomness, the proposed algorithm outperforms the previous methods up to 20%.
Keywords
)