Home / Current Issue / Paper 1723105
Intelligent Predictive Energy Management Of 5G Base Stations Using Reinforcement Learning Abstract
Subject area: Science,Engineering and Technology · Area of research: Energy Manreinforcement Learning
Abstract
The aim of the paper is design of AI-based predictive energy management model for efficient wireless base stations using Reinforcement Learning (RL). The methodology employs a multi-phase methodology to develop and validate a reinforcement learning-based predictive energy management model for 5G base stations. Energy consumption data were collected and used to model the base station’s operation and formulate an optimization problem. A reinforcement learning algorithm was then implemented in Python to learn real-time energy-saving strategies based on traffic load and environmental factors. The model was validated through simulations and real-time tests, evaluating improvements in energy efficiency, service quality, and cost reduction. The testbed is MTN 5G site at Polo Park, Enugu. Findings from investigation revealed the cost of daily energy used by the site is ₦31377.7787 and the monthly running cost is ₦972711.1397. To reduce the economic burden, RL-based predictive energy management model was proposed as predictive and adaptive learning techniques which utilized Q- learning algorithm as the learning agent to adjust the action of the site based on dynamic state space of the network condition. Python programming language was used to train the model and also integration on the 5G network. Results obtained revealed that with RL, the daily energy saved when compared with characterized is ₦4625.0896. The monthly energy saved amount to ₦170130.4667 with RL-based predictive energy management model, while every year, ₦2041565.6004 is the amount saved. The reason was because of the ability of the RL to adjust the action of RL based on the different state space of the site to save energy. The percentage reduction in daily energy consumed with RL based station is 33.90%. The percentage reduction in monthly energy consumed is 14.74%. In conclusion the study has demonstrated the RL can help manage energy consumption in the 5G network while maintaining quality of service.
Keywords
energy management; base stations; reinforcement learning (rl); energy consumption; polo park, enugu
References
[1] Ezzeddine, Z., Khalil, A., Zeddini, B., & Ouslimani, H. H. (2024). A Survey on Green Enablers: A Study on the Energy Efficiency of AI-Based 5G Networks. Sensors, 24(14), 4609. MDPI
[2] Ichimescu, A., Popescu, N., Popovici, E. C., & Toma, A. (2024). Energy Efficiency for 5G and Beyond 5G: Potential, Limitations, and Future Directions. Sensors, 24(22), 7402. MDPI
[3] Ichimescu, A., Popescu, N., Popovici, E. C., & Toma, A. (2024). Energy Efficiency for 5G and Beyond 5G: Potential, Limitations, and Future Directions. Sensors, 24(22), 7402. MDPI
[4] Kaur, P., Garg, R. & Kukreja, V. Energy-efficiency schemes for base stations in 5G heterogeneous networks: a systematic literature review. Telecommun Syst 84, 115–151 (2023). Springer
[5] Ma, X., Zhu, Q., Duan, Y., et al. (2022). Optimal configuration of 5G base station energy storage considering sleep mechanism. Global Energy Interconnection, 5(1), 66–76.
[6] Premalatha, J., & SahayaAnselin Nisha, A. (2023). Base station energy management in 5G networks using wide range control optimization. Intelligent Automation & Soft Computing. Tech Science Press
[7] Wang, Y. (2020). Predictive traffic load forecasting for base station energy management in cellular networks. IEEE Access, 8, 74508–74517.
[8] Zhu, Y., Li, K. and Zhang, L. (2025). Base station microgrid energy management in 5G networks - a brief review. In: Smart Grid and Cyber Security Technologies. The 2024 Intelligent Computing for Sustainable Energy and Environment (ICSEE2024), 13-15 Sep 2024, Suzhou, China. Communications in Computer and Information Science. Springer, Singapore ISBN 978-981-96-0224-7. Springer
How to cite this paper
@article{1723105,
author = {Onwuha, U. H.},
title = {Intelligent Predictive Energy Management Of 5G Base Stations Using Reinforcement Learning Abstract},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {10},
number = {3},
pages = {1982-1992},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1723105.pdf},
abstract = {The aim of the paper is design of AI-based predictive energy management model for efficient wireless base stations using Reinforcement Learning (RL). The methodology employs a multi-phase methodology to develop and validate a reinforcement learning-based predictive energy management model for 5G base stations. Energy consumption data were collected and used to model the base station’s operation and formulate an optimization problem. A reinforcement learning algorithm was then implemented in Python to learn real-time energy-saving strategies based on traffic load and environmental factors. The model was validated through simulations and real-time tests, evaluating improvements in energy efficiency, service quality, and cost reduction. The testbed is MTN 5G site at Polo Park, Enugu. Findings from investigation revealed the cost of daily energy used by the site is ₦31377.7787 and the monthly running cost is ₦972711.1397. To reduce the economic burden, RL-based predictive energy management model was proposed as predictive and adaptive learning techniques which utilized Q- learning algorithm as the learning agent to adjust the action of the site based on dynamic state space of the network condition. Python programming language was used to train the model and also integration on the 5G network. Results obtained revealed that with RL, the daily energy saved when compared with characterized is ₦4625.0896. The monthly energy saved amount to ₦170130.4667 with RL-based predictive energy management model, while every year, ₦2041565.6004 is the amount saved. The reason was because of the ability of the RL to adjust the action of RL based on the different state space of the site to save energy. The percentage reduction in daily energy consumed with RL based station is 33.90%. The percentage reduction in monthly energy consumed is 14.74%. In conclusion the study has demonstrated the RL can help manage energy consumption in the 5G network while maintaining quality of service.},
keywords = {energy management; base stations; reinforcement learning (rl); energy consumption; polo park, enugu},
month = {September},
}