Energy-optimized secure remote fingerprint authentication in a mobile environment

To optimize the energy consumption during transmission in a secure remote fingerprint authentication system, most approaches have considered a fixed-rate compression of WSQ(Wavelet Scalar Quantization). To balance the energy consumption and the recognition accuracy, however, we need to consider the total energy consumption during compression/encryption/transmission and the quality of the given image. In this paper, we first develop an energy-accuracy prediction model. Then, from the given image, we can evaluate the quality of it. Finally, from the user´s minimum requirement of the recognition accuracy, the energy-optimized compression rate under the requirement can be derived adaptively. Based on the experimental result, we show that total energy consumption of the fixed approach can be reduced by 29%.