A measurement-based approach to analyze the power consumption of the softwarized 5G core

TL;DR

This work addresses the challenge of quantifying power consumption in softwarized 5G core networks deployed at the network edge. It proposes a measurement-based methodology that combines hardware meters with software-based metrics to attribute power to VNFs across bare metal, VM, and containerized deployments, using Open5GS and Free5GC on a COTS-driven testbed. The study demonstrates substantial energy differences among virtualization options and 5GC implementations, highlighting up to ~80% higher energy for VM and around 25% more for container-based setups versus bare metal, with Free5GC offering notable power savings due to kernel-space processing. The findings enable real-time energy awareness and pave the way for green orchestration and ML-driven energy optimization in edge-to-cloud 5G ecosystems, with data openly available for further research.

Abstract

In light of the ever growing energy needs of the ICT sector, a value that is becoming increasingly important for a mobile network is its power consumption. However, the transition away from legacy network deployments tightly coupled with the underlying hardware and the adoption of the Network Function Virtualization (NFV) paradigm has made more difficult to accurately evaluate their energy and carbon footprint. In this paper, we propose and validate a measurement-based approach to analyze the power consumption of a virtualized 5G core network (5GC) deployment. We design an experimental testbed using commercial off-the-shelf (COTS) hardware and open-source software as a sample architecture simulating an edge computing node and supporting three different virtualization options. We make use of both hardware-based and software-based power meters to investigate the power consumption trends associated with increasing levels of traffic and multiple 5GC deployment types. The results show the feasibility of a real-time power monitoring system and highlight how deployment choices, such as virtualization framework and 5GC software, can significantly impact on the power consumption of the network.

Figures (10)

• Figure 1: Diagram showing a power consumption monitoring system integrated with a virtualized 5GC deployment. The parts we are concerned about in this work are highlighted in blue.
• Figure 2: Experimental testbed architecture.
• Figure 3: Power consumption (in mW) against CPU load (in percentage with respect to the maximum achievable load) measured on the testbed hardware. We interpolated the data with a fourth degree polynomial function showing the nonlinearity
• Figure 4: Power consumption (in mW) for the three different virtualization technologies at different uplink traffic levels. Notice that the VM deployment cannot sustain a throughput higher than 700 Mbps, hence we do not show it past that point.
• Figure 5: Power consumption trends for the 3 different virtualization technologies and 2 power meters.