02-Doctorat

Permanent URI for this communityhttps://dspace.univ-tebessa.dz/handle/123456789/38

دكتوراه

Browse

Filters

20262

Settings

Subject: search.filters.subject.Blockchain

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Biometrics-Based Information Security
    (University of echahid cheikh larbi tebessi - tebessa, 2026-07-08) Yacine Belhocine
    The accelerating digitization of human activities has redefined how information is cre- ated, transmitted, and protected. However, this interconnected digital ecosystem has also become a prime target for sophisticated cyberattacks, data breaches, and iden- tity fraud. Traditional security mechanisms, such as passwords, PINs, and centralized databases, are increasingly inadequate to safeguard sensitive assets and personal iden- tities. In this context, biometric-based authentication, supported by emerging technolo- gies, offers a path toward more resilient and user-centered security. This thesis proposes an integrated, multi-layered security framework that unites biometrics, blockchain, ex- plainable artificial intelligence (XAI), and quantum machine learning (QML) to achieve secure, transparent, and future-ready identity verification. The research first introduces a revocable biometric template protection scheme that ensures both privacy and renewa- bility, allowing biometric credentials to be securely reissued if compromised. Building upon this, the MedBioCh system is developed, an innovative blockchain-enabled archi- tecture for digital healthcare that decentralizes trust, prevents tampering, and enforces privacy-aware access control through smart contracts. To strengthen accountability and interpretability, an explainable phase-aware biometric framework is proposed, enabling transparent authentication decisions and traceable reasoning paths suitable for sensitive sectors such as healthcare and justice. Furthermore, a training-free quantum classi- fier based on fidelity-based similarity is designed to deliver efficient, interpretable, and noise-resilient classification, offering a realistic step toward post-quantum security. Ex- perimental evaluations across multiple biometric datasets confirm the frameworks su- perior resilience against spoofing, template compromise, and adversarial interference while maintaining high recognition accuracy and system scalability. By harmonizing these emerging paradigms, this work establishes a unified vision of trustworthy digital identity, combining the precision of biometrics, the decentralization of blockchain, the clarity of XAI, and the resilience of quantum computation. It lays a scientific and technological foundation for the next generation of secure, ethical, and transparent au- thentication systems.
  • No Thumbnail Available
    Item
    Analysis and improvement of QoS and security of Mobile Edge Computing based networks
    (Echahid Cheikh Larbi Tebessi University - Tebessa, 2026-05-11) Amira Cheriet
    The rapid evolution of mobile technologies and connected objects (IoT) has led to a massive increase in the volume of data to be processed and secured. Traditional cloud architectures, while efficient for large-scale storage and processing, no longer meet current requirements in terms of latency, bandwidth, and security. To address these challenges, the Mobile Edge Computing (MEC) paradigm is emerging as an innovative solution that brings computing and storage resources closer to end users while improving the performance of distributed services. This thesis offers a series of contributions aimed at optimizing the processing, security, and storage of sensitive data in MEC environments through the integration of complementary technologies such as Software-Defined Networking (SDN), Blockchain, Network Functions Virtualization (NFV), and IPFS. In the first part, we design a secure mobile data offloading architecture in small cell networks (SCN), combining SDN, public/private blockchain, and Wi-Fi and D2D communications. This solution aims to guarantee ultra-low latency while ensuring the integrity and confidentiality of data transferred from mobile devices to the cloud. Secondly, we address the issue of secure edge video streaming management. We propose a distributed model based on dynamic orchestration of security tasks via NFV, allowing for optimized resource allocation while ensuring consistent quality of service, even under performance constraints. Finally, we present a secure storage solution for video surveillance systems, based on blockchain and IPFS, with a MEC layer enabling local preprocessing. The use of mechanisms such as Segregated Witness and smart contracts ensures the integrity, traceability, and efficient storage of critical video data. All the proposals are validated by simulations and experimental evaluations that demonstrate their effectiveness in terms of latency reduction, security improvement, data resilience, and resource optimization. This thesis thus paves the way towards intelligent, secure, and efficient edge environments, adapted to the growing needs of modern mobile applications, particularly in the fields of IoT, multimedia services, and urban security.