Abdullah Hasan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Abdullah Hasan has a total of 1 co-authors. Here are the top co-authors by number of shared publications:
- Farhan Khalid
National University of Computer and Emerging Sciences, Islamabad, Pakistan
1 shared publication
Latest Publications
- Open Access
- 26 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 47952-47967
AI-Driven Self Optimization of 5G Network Coverage and Capacity Using Multi-Agent Deep Reinforcement Learning
Coverage and capacity optimization (CCO) is critical for satisfying the high data rate and quality of service (QoS) demands in dense fifth-generation (5G) wireless networks. In this study, we propose a novel multi-agent deep reinforcement learning architecture based on proximal policy optimization (MADRL-PPO) to adjust the electronic antenna downtilt for CCO in 5G urban macro (UMa) deployments operating in the sub-6 GHz band. Based on the observations of spectral efficiency and interference, each macro base station (MBS) in the UMa scenario operates as an independent learning agent, dynamically adjusting its electronic antenna downtilt. To ensure fairness among users, a coverage-capacity fairness metric (CCFM) was introduced, which combines the average and edge spectral efficiencies. The reward function has an adaptive fairness coefficient that is tuned in real time based on the evolving state of the network. The proposed algorithm was implemented and evaluated in a 3GPP-compliant 5G UMa simulation environment developed using MATLAB. The simulation results demonstrate significant improvements in spectral efficiency and throughput, and that the proposed MADRL-PPO improves the average spectral efficiency, edge spectral efficiency, and throughput by up to 31%, 41%, and 30%, respectively, compared to fixed downtilt scenarios. The proposed MADRL-PPO is effective in creating scalable, distributed, self-organizing networks.
- 26 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 47952-47967
Open AccessAI-Driven Self Optimization of 5G Network Coverage and Capacity Using Multi-Agent Deep Reinforcement Learning
Coverage and capacity optimization (CCO) is critical for satisfying the high data rate and quality of service (QoS) demands in dense fifth-generation (5G) wireless networks. In this study, we propose a novel multi-agent deep reinforcement learning architecture based on proximal policy optimization (MADRL-PPO) to adjust the electronic antenna downtilt for CCO in 5G urban macro (UMa) deployments operating in the sub-6 GHz band. Based on the observations of spectral efficiency and interference, each macro base station (MBS) in the UMa scenario operates as an independent learning agent, dynamically adjusting its electronic antenna downtilt. To ensure fairness among users, a coverage-capacity fairness metric (CCFM) was introduced, which combines the average and edge spectral efficiencies. The reward function has an adaptive fairness coefficient that is tuned in real time based on the evolving state of the network. The proposed algorithm was implemented and evaluated in a 3GPP-compliant 5G UMa simulation environment developed using MATLAB. The simulation results demonstrate significant improvements in spectral efficiency and throughput, and that the proposed MADRL-PPO improves the average spectral efficiency, edge spectral efficiency, and throughput by up to 31%, 41%, and 30%, respectively, compared to fixed downtilt scenarios. The proposed MADRL-PPO is effective in creating scalable, distributed, self-organizing networks.