MSA Repository "MSAR"

MSAR University's Digital Repository is a documentation and digitization of all university outcomes that are of effective value in the scientific and academic community and reflects the university's image, work, and effective contribution to society Through MSAR Digital Repository, the university managed to collect, store, archive and publish digital content - including documents, audio files, images and data sets - all in a safe place. MSAR is one of the strongest University Digital Repositories in Egypt and documented in the DSPACE community with its latest versions.

Communities in DSpace

Select a community to browse its collections.

Now showing 1 - 5 of 6

Recent Submissions

  • Thumbnail Image
    Item type: Item ,
    A Novel Explainable AI Framework for Real-Time Cybersecurity Threat Detection and Mitigation
    (Association of Scientific Research Technology and the Arts, 2025-01) Ghada Abdelhady
    Cybersecurity remains a critical challenge as cyberattacks grow increasingly sophisticated and diverse. This paper presents a novel Explainable AI (XAI) framework for real-time detection and mitigation of cyber threats, including Distributed Denial of Service (DDoS) attacks, Shellcode exploitation, Reconnaissance, and Worm propagation. The framework employs advanced feature engineering and class-specific techniques to enhance detection accuracy, particularly for overlapping categories like DoS and Exploits. It integrates visual explainability tools, automates incident response processes, and seamlessly connects with Security Information and Event Management (SIEM) systems to support operational decision-making. Using eXtreme Gradient Boost (XGBoost) combined with SHapley Additive exPlanations (SHAP) for explainability, the system achieves both high detection accuracy and transparency. Additionally, a comparative analysis with Random Forest (RF) and Support Vector Machine (SVM) highlights the proposed framework's superior performance. Experimental results demonstrate an accuracy of 89% and an F1-score of 0.88, with strong detection capabilities for high-priority threats like Generic and Shellcode while maintaining high precision across all classes. This research underscores the potential of the framework to transform real-time cybersecurity by ensuring precise, transparent, and actionable threat detection.
  • Thumbnail Image
    Item type: Item ,
    Enhancing Neuroprosthetic Control Using CNN-LSTM Models: A Simulation Study with EEG-Based Motor Imagery
    (Egyptian Association for Technological Development, 2025-02) Ghada Abdelhady
    The development of intuitive and responsive neuroprosthetic systems remains a critical challenge in assistive technologies, particularly in decoding neural signals to enable precise and adaptive motor control. This study addresses the problem of translating EEG-based motor imagery into effective neuroprosthetic control, overcoming challenges such as limited data, overfitting in predictive models, and practical constraints in robotic actuation. A CNN-LSTM hybrid model was developed to classify motor imagery tasks using EEG signals. The application of data augmentation and regularization techniques improved the model’s performance, achieving a test accuracy of 93.5% and balanced precision and recall across motor imagery tasks. To validate its practical application, a PyBullet-based simulation demonstrated the successful control of a robotic gripper, where the model’s predictions were translated into accurate "open" and "close" actions. The gripper joints performed these actions with high precision, showcasing the system's potential for real-time neuroprosthetic applications. However, constraints such as dataset limitations and simulation-specific constraints underscore the need for further optimization. This study provides a robust proof-of-concept for integrating deep learning with brain-computer interfaces to achieve adaptive, reliable, and real-time neuroprosthetic control. By addressing key challenges, the proposed framework bridges the gap between neural signal decoding and physical actuation, offering a pathway toward advanced and responsive neuroprosthetic systems
  • Thumbnail Image
    Item type: Item ,
    Dynamic Machine Learning Framework for Secure, Ultra-Reliable, and Energy-Efficient Open RAN Systems
    (Institute of Electrical and Electronics Engineers Inc, 2025-09-09) ABDULRAHMAN GHANDOURA; ABDULLAH ALAJMI; MAHMOUD EL-SAKHAWY; ABDELWAHED MOTWAKEL; FADL DAHAN; GHADA ABDELHADY
    This paper proposes a machine learning (ML)-driven framework for Open Radio Access Networks (ORAN) to address security, Ultra-Reliable Low-Latency Communication (URLLC), and energy efficiency challenges. By integrating Isolation Forest for security, One-Class SVM for URLLC, and Generative Adversarial Networks (GANs) for cost optimization, the framework dynamically adapts to network traffic shifts while adhering to ORAN’s open architecture principles. Evaluations on a dataset of 72,000 instances with 21 features demonstrate 92% attack detection, 15% latency reduction, and 20% energy savings, validated through rigorous metrics such as precision-recall the area under the curve (AUC) and synthetic cost modeling. The proposed solution represents a significant advancement in the field of intelligent network management, offering a comprehensive approach to the most pressing challenges in modern telecommunications infrastructure. Through extensive simulations and real-world testing, we validate that our framework outperforms existing approaches in terms of attack detection accuracy, latency optimization, and energy efficiency without compromising the essential interoperability principles of the ORAN architecture.
  • Thumbnail Image
    Item type: Item ,
    Hybrid Machine Learning Anomaly Detection and Lightweight Zero Trust Authentication for LoRaWAN Networks
    (Institute of Electrical and Electronics Engineers Inc, 2026-02-06) GHADA ABDELHADY; ABDULRAHMAN GHANDOURA; ABDELWAHED MOTWAKEL; ABDULLAH ALAJMI
    LoRaWAN has become one of the most widely adopted LPWAN technologies, but its large-scale use has also exposed several persistent security weaknesses. Recent studies show that unprotected LoRaWAN links are still vulnerable to basic attacks such as eavesdropping (67% success rate) and replay attempts (about 43%), which highlights the need for more practical and adaptive security solutions suitable for low-power devices. In this work, we develop a security framework that combines a hybrid machine-learning model for anomaly detection with a lightweight Zero Trust authentication mechanism. The anomaly detection module merges a tuned LightGBM classifier with an autoencoder-based unsupervised detector. Across multiple attack categories, the combined model achieved an average detection accuracy of 95% and an AUC-ROC of 0.973 while keeping the memory footprint small enough for LoRa-class devices. We also design a Zero Trust authentication scheme based on an optimized Proof-of-Authority blockchain model and a low-cost mutual authentication protocol. The evaluation shows an average authentication delay of about 187 ms and system availability above 99.9%. Although the full framework introduces an energy overhead of approximately 18%, the projected device lifetime remains more than seven years, which is significantly longer than what existing blockchain-based IoT authentication systems provide.
  • Thumbnail Image
    Item type: Item ,
    Energy-Efficient Load Balancing in Fog Computing Using a Hybrid Dijkstra Greedy Approach
    (International Association of Engineers, 2026-08) Mahmoud A. El-Sakhawy; Ghali Abdullah; Imane Aly Saroit Ismail; Shaimaa M. Mohamed; Ghada Abdelhady
    Fog computing has become a key paradigm for latency-sensitive Internet of Things (IoT) applications by bringing computation closer to data sources. It extends processing capabilities to the network edge, thereby reducing reliance on centralized cloud infrastructures and enabling realtime data handling. However, the distributed and heterogeneous characteristics of fog environments create significant challenges in resource management and energy efficiency, especially when distributing workloads across dynamically changing nodes. To overcome these challenges, this paper aims to distribute workloads efficiently among fog nodes. It introduces a Hybrid Energy-Aware Load Balancing (EALB) algorithm that integrates Dijkstra’s shortest path algorithm for routing with a greedy resource allocation strategy. To evaluate the proposed algorithm, simulations were conducted across multiple deployment scales, from small IoT setups to large smart city environments. Results demonstrate that, unlike existing approaches, the proposed method simultaneously accounts for network topology and resource availability when making load balancing decisions. The results indicate notable improvements in average latency (5.36% in medium-scale and 4.55% in large-scale deployments), together with a manageable trade-off in power consumption. By combining shortest-path routing with greedy allocation, the hybrid model improves latency performance as the system scales, besides a slight increase in energy usage. In small-scale scenarios, the Greedy algorithm achieves superior performance in both latency (13.4% reduction in total latency) and energy consumption (5.57% lower power use). However, in large-scale environments, the Hybrid EALB algorithm reaches substantial latency gains with only a marginal increase in power consumption (0.36%). Overall, the EALB approach enables more intelligent routing and task allocation through its integrated strategy, effectively balancing latency and energy efficiency. These results show how useful hybrid optimization approaches can be when building sustainable fog computing architectures. They also offer practical guidance on choosing the right load balancing strategy, depending on the system’s size, network structure, and specific performance needs.