AI-Driven Optimization and Control for NextG Cellular Networks
Event details
| Date | 17.06.2025 |
| Hour | 14:00 › 16:00 |
| Speaker | Aoyu Gong |
| Location | |
| Category | Conferences - Seminars |
EDIC candidacy exam jury
Exam president: Prof. Karl Aberer
Thesis advisor: Prof. Haitham Al Hassanieh
Co-examiner: Prof. Paolo Ienne
Abstract
Reliably achieving low latency in Next-Generation (NextG) cellular networks is critical for emerging applications like autonomous vehicles and virtual reality, where minimizing uplink latency is vital. However, current uplink scheduling relies on grant-based access, a reactive and multi-round procedure that introduces substantial delays and undermines low-latency targets. Motivated by recent advances in AI-driven optimization and control, along with the software-defined Open Radio Access Network (O-RAN) architecture, we explore previous research on 5G resource allocation across non-real-time, near-real-time, and real-time timescales to analyze their strengths and identify key limitations. Building on these insights, we aim to propose a real-time system for uplink traffic characterization, prediction, and scheduling to enable proactive, low-latency resource allocation in NextG cellular networks.
Selected papers
Exam president: Prof. Karl Aberer
Thesis advisor: Prof. Haitham Al Hassanieh
Co-examiner: Prof. Paolo Ienne
Abstract
Reliably achieving low latency in Next-Generation (NextG) cellular networks is critical for emerging applications like autonomous vehicles and virtual reality, where minimizing uplink latency is vital. However, current uplink scheduling relies on grant-based access, a reactive and multi-round procedure that introduces substantial delays and undermines low-latency targets. Motivated by recent advances in AI-driven optimization and control, along with the software-defined Open Radio Access Network (O-RAN) architecture, we explore previous research on 5G resource allocation across non-real-time, near-real-time, and real-time timescales to analyze their strengths and identify key limitations. Building on these insights, we aim to propose a real-time system for uplink traffic characterization, prediction, and scheduling to enable proactive, low-latency resource allocation in NextG cellular networks.
Selected papers
Practical information
- General public
- Free
Contact
- edic@epfl.ch