A hybrid blowfish-based cryptography and chaotic quantization steganography framework with genetic key generation

Area of Science:

• Computer Science
• Information Security
• Cryptography

Background:

• Information exchange security is critical, with threats including data theft, loss, and manipulation.
• Cryptography and steganography are key data hiding techniques, offering distinct security approaches.
• Existing methods face challenges in balancing encryption strength with data imperceptibility.

Purpose of the Study:

• To propose a novel security mechanism, Advanced Blowfish Encryption with Chaotic Image Quantization (ABECIQ), integrating cryptography and steganography.
• To enhance the Blowfish algorithm with a modified round function and chaotic image quantization for secure key generation and data embedding.
• To evaluate ABECIQ's performance against existing algorithms using various security and efficiency metrics.

Main Methods:

• Developed ABECIQ by combining a modified Blowfish encryption algorithm with chaotic image quantization for steganography.
• Utilized a genetic algorithm and real-time clock values for generating secure secret keys.
• Embedded encrypted data into images using the chaotic-quant technique.

Main Results:

• ABECIQ achieved high imperceptibility with Peak Signal-to-Noise Ratio (PSNR) values from 65-74 dB and Structural Similarity Index (SSIM) values above 0.999.
• Generated keys exhibited high entropy, close to the theoretical maximum of 8 bits per character.
• Demonstrated superior performance, including higher throughput and better security metrics (PSNR, MSE, SSIM, Correlation Coefficient) compared to Chaotic, Blowfish Encryption, and AES-RDH algorithms.

Conclusions:

• ABECIQ offers a robust and efficient solution for secure data exchange by effectively combining cryptography and steganography.
• The proposed method provides enhanced security features and improved computational efficiency over existing techniques.
• ABECIQ is a promising approach for applications requiring high levels of data confidentiality and integrity.