On Tuesday, March, 3rd ,2026 the department of Communication and Information engineering at Al-Khwarizmi College of engineering held a master thesis defense titled “Machine Learning Defragmentation Techniques for Elastic Optical Networks”, for the master candidate (Muhammad Baqir Sattar) at the thesis defense hall, where a degree of passing with distinction was awarded for his research findings.
The committee is composed of the following members:
- Professor Dr. Ahmed Hamid Wajih, Chair
- Assistant Professor Dr. Liwa Fasel Abdul Ameer, Examiner
- Assistant Professor Dr. Sarah Kadhim Muhsin, Examiner
- Lecturer Dr. Omar Yusif Shaban, Supervisor
This thesis demonstrated a significant challenge faced in the elastic optical networks, that considered a building block to meet growing bandwidth demand in modern communication systems, but suffering from spectrum fragmentation resulting from dynamic traffic and allocation constraints, causing inefficient spectrum to use and higher blocking.
This study also highlighted the limited conventional approaches, traditional heuristic or ILP-based methods, as it is either inflexible or unscalable for such environments. To address this, a scalable DRL framework using A2C, PPO, and DQN is proposed for traffic-aware RMSA and proactive defragmentation.
Results showed up to 61.63% SBR improvement over SP-FF, with 95% confidence intervals confirming improved scalability, spectral efficiency, and overall performance compared with conventional methods.
