Ministry of the Internal Administration

ABSTRACT The contemporary challenges of peace and security in the African continent are most prominently pinned to regional organizations. African regional organizations have a significant opportunity to consolidate and develop their vision of peace and stability on the continent, but in general they still lack the depth of resources and experience to act alone. The EU is a decisive partner in reinforcing their capacity to respond to the peacebuilding challenges of the continent. There is nothing preordained about this process; political leadership, as always, will be decisive.

This study introduces a novel and secure multi-factor authentication framework designed to enhance the protection of biometric data through the integration of cryptographic and steganographic techniques. The primary objective is to safeguard sensitive biometric identifiers, such as fingerprints, by employing a hybrid authentication mechanism that combines traditional password-based verification with biometric authentication. The proposed model utilizes the SHA-256 hash function to transform user passwords and biometric templates into fixed-size encrypted representations. Furthermore, the Least Significant Bit (LSB) steganographic method is applied to conceal the hash value of biometric data within the user's profile image, ensuring covert storage. To enhance security, a pseudo-random number generator, incorporating password and biometric hashes, is employed to introduce randomness into the embedding process, significantly strengthening resistance against steganalysis attacks. Experimental evaluations were conducted using various biometric samples and profile images to assess metrics such as Normalized Mean Absolute Error (NMAE), Peak Signal-to-Noise Ratio (PSNR), payload capacity, sensitivity to tampering, and processing time. The results show that the system achieves high PSNR (up to 80.56 dB), low NMAE (as low as 0.00048), and acceptable processing times, confirming its robustness, efficiency, and suitability for secure authentication in real-world applications.

Black box testing plays a crucial role in software development, ensuring system reliability and functionality. However, its effectiveness is often hindered by the sheer volume and complexity of big data, making it difficult to prioritize critical test cases efficiently. Traditional testing methods struggle with scalability, leading to excessive resource consumption and prolonged testing cycles. This study presents an AI-driven test case prioritization (TCP) approach, integrating decision trees and genetic algorithms (GA) to optimize selection, eliminate redundancy, and enhance computational efficiency. Experimental results demonstrate a 96% accuracy rate and a 90% success rate in identifying relevant test cases, significantly improving testing efficiency. These findings contribute to advancing automated software testing methodologies, offering a scalable and efficient solution for handling large-scale, data-intensive testing environments.