he thesis addressed multiple inputs and single outputs (Multiple Input Single Output) over a (Full-Duplex) channel, where the generated data was used in a ray-tracing manner to enhance the model’s performance. Feature engineering techniques were applied to the generated datasets to help the models distinguish patterns among the input datasets.
The thesis aimed to generate robust models by training various deep learning models, controlled by a single machine learning model, to predict the optimal Power Allocation Factor, which is assigned to users of non-orthogonal multiple access.
The discussion committee consisted of:
- Prof. Dr. Sidiq Youssef Amin, Chair.
- Assoc. Prof. Dr. Mahmoud Ahmed Mahmoud, Member.
- Assoc. Prof. Dr. Bilal Alaa Al-Din Jaber, Member.
- Assoc. Prof. Dr. Mohammed Abdul Rahman Ahmed, Member and Supervisor