BROWSE
Detail
Conference Paper
Green Intelligence: Traffic Forecasting using O-RAN AI/ML Framework
Cited 0 time in 
Share 
Share
- Authors
- Geon Kim, Sung-Jin Lee, Hyuk-Sun Kwon, Ho-Seong Choi, Hyun-Min Yoo, Kyung-Sook Kim, Jee-Hyeon Na, Een-Kee Hong
- Issue Date
- 2025-07
- Citation
- International Conference on Ubiquitous and Future Networks (ICUFN) 2025, pp.380-385
- Language
- English
- Type
- Conference Paper
- Abstract
- As mobile networks rapidly evolve with the proliferation of 5G technologies, the imperative to reduce carbon emissions and enhance energy efficiency in network operations has become central to developing sustainable communication systems. This paper presents implementation of a machine learning pipeline for base station traffic forecasting, utilizing open-source tools from the O-RAN Software Community (O-RAN SC) within a real Kubernetes environment. Our approach not only leveraged the capabilities of O-RAN SC's offerings but also contributed significantly to the community by implementing key functionalities aligned with O-RAN Alliance standards, ensuring full interoperability with existing O-RAN architectures. By tailoring our pipeline to address the specific challenges of mobile traffic prediction, including handling dynamic data patterns and realtime processing, we achieved high-performance outcomes while rigorously measuring carbon emissions and energy consumption. This dual focus on technical excellence and sustainability underscores our commitment to advancing eco-friendly mobile network operations and demonstrates the potential of collaborative, opensource development to shape the future of green communications.