The College of Information Technology Witnesses the Defense of a PhD Dissertation on Enhancing 5G Resource Allocation Using Deep Learning
Duhaa Fadill Abbas
In line with the strategic directions of the Information Networks Department at the College of Information Technology, University of Babylon, aimed at advancing cutting-edge research and keeping pace with developments in modern communication technologies, a PhD dissertation was successfully defended by Mr. Ammar Abdulhadi Abdullah, under the supervision of Professor Dr. Mahdi Abadi Manea, on Tuesday, March 17, 2026, at the college’s conference hall, in the presence of faculty members, researchers, and postgraduate students.
The dissertation, entitled “Enhancing 5G Network Resource Allocation Using Deep Learning,” addressed the critical challenges associated with the transition to fifth-generation (5G) networks, particularly the diverse service requirements ranging from ultra-reliable low-latency communications (URLLC) to massive machine-type communications (mMTC). These dynamic and complex environments render traditional static resource allocation mechanisms insufficient and inefficient.
The research aimed to develop a comprehensive framework for cross-layer resource management by integrating hybrid artificial intelligence techniques. The proposed approach combines deep reinforcement learning with transfer learning, alongside advanced optimization algorithms, to overcome the limitations of conventional methods, including the cold-start problem and slow convergence, thereby enabling faster and more efficient decision-making in resource allocation.
The methodology introduced a two-stage model focusing on optimizing scheduling in the Radio Access Network (RAN), in addition to implementing an intelligent hierarchical orchestrator for managing Mobile Edge Computing (MEC) resources. This integrated approach ensures a balance between high performance and energy efficiency.
Experimental results demonstrated that the proposed framework outperforms traditional approaches in reducing latency, maintaining network stability under peak loads, and significantly improving energy efficiency while preserving Quality of Service (QoS) requirements. These findings highlight the effectiveness of the proposed solution in addressing the demands of next-generation network environments.
The dissertation concludes that hybrid artificial intelligence represents a promising and scalable solution for 5G networks, offering a robust framework that bridges the gap between performance optimization and resource efficiency, and contributes to the development of intelligent, adaptive, and energy-efficient communication systems.