Energy-aware approaches to resource allocation in open radio access networks

Abstract

The evolution of mobile communication networks toward 5G and beyond seeks to meet the demands for high throughput, low latency, and energy efficiency. Open Radio Access Networks (O-RAN), with their disaggregated and virtualized architecture, provide flexibility and interoperability but also pose significant challenges in terms of energy consumption. This work addresses energy-aware resource allocation in O-RAN, focusing on Virtualized Network Function (VNF) placement and user equipment association. We propose a Mixed Integer Linear Programming model that jointly considers radio equipment, transport network, and VNF migration energy costs while supporting flexible functional split options and routing decisions, complemented by a heuristic for scalability. Using synthetic load and topology generators, we evaluate diverse scenarios and show that hierarchical topologies achieve up to 15% more centralization and reduce energy consumption by about 28% compared to current widely adopted topologies. Additionally, joint optimization of VNF placement and user equipment association enables significant energy savings by disabling entire base stations during moments of low load. A disjoint approach to the problem is able to solve larger instances of the problem and achieves solutions close to optimal while surpassing a maximum-Signal to Noise Ratio (SNR) solution. Despite the optimal solutions being unable to meet the stringent response time required in practical deployments, they present a robust baseline for the evaluation of both the network and non-optimal approaches to the problem.