Improved method of attack on an asymmetric cryptosystem based on phase-truncated Fourier transform

Area of Science:

• Cryptography and Information Security
• Image Processing and Computer Vision

Background:

• Asymmetric cryptosystems often rely on Fourier transform properties, which can be vulnerable to phase truncation attacks.
• Existing attacks may suffer from noise, require significant computational resources, or yield imprecise results.

Purpose of the Study:

• To propose an improved attack method for phase-truncated Fourier transform-based asymmetric cryptosystems.
• To develop a novel cryptosystem with enhanced security and efficiency.

Main Methods:

• A novel median-filtering phase-retrieval algorithm is employed for the attack.
• The attack targets the decryption keys and aims to recover original image information.
• A new cryptosystem is designed based on the insights from the improved attack method.

Main Results:

• The proposed attack successfully retrieves exact decryption keys and precise attack results.
• The method achieves exact information recovery in both gray-scale and binary forms.
• Demonstrated advantages include superior computing efficiency and enhanced robustness against noise and occlusions.

Conclusions:

• The novel median-filtering based attack offers significant improvements over existing methods in terms of accuracy, efficiency, and robustness.
• The newly proposed cryptosystem leverages the attack's principles to achieve enhanced security without relying on truncated phases.