Energy Profiling and Analysis of 5G Private Networks: Evaluating Energy Consumption Patterns
Johirul Islam, Ijaz Ahmad, Shakthi Gimhana, Juho Markkula, Erkki Harjula · 2024
Read Energy Profiling and Analysis of 5G Private Networks: Evaluating Energy Consumption Patterns by Johirul Islam, Ijaz Ahmad, Shakthi Gimhana, Juho Markkula, Erkki Harjula (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-01634.txt.
Introduction
Johirul Islam, Ijaz Ahmad, Shakthi Gimhana, Juho Markkula, and Erkki Harjula measured energy use in private 5G networks in 2024. Private networks give owners full control, more security, and services tuned to a site, which suits places like smart hospitals. Their long-term cost and footprint depend on how much energy they use, a topic past work had largely skipped.
The team built a lab setup that mimics a real private network and profiled each part under varied conditions. Their data show how energy use shifts with load and settings. The aim is to guide greener designs for private networks now and for the next generation of mobile systems later.
Catalog record
| CREATOR | Johirul Islam, Ijaz Ahmad, Shakthi Gimhana, Juho Markkula, Erkki Harjula |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 9 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-01634
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Energy Profiling and Analysis of 5G Private Networks: Evaluating Energy Consumption Patterns Johirul Islam, Ijaz Ahmad, Shakthi Gimhana, Juho Markkula, Erkki Harjula Abstract . Private 5G networks provide enhanced security, a wide range of optimized services through network slicing, reduced latency, and support for many IoT devices in a specific area, all under the owner’s full control. Higher security and privacy to protect sensitive data is the most significant advantage of private networks, in e.g., smart hospitals. For long-term sustainability and cost-effectiveness of private 5G networks, analyzing and understanding the energy consumption variation holds a greater significance in reaching toward green private network architecture for 6G. This paper addresses this research gap by providing energy profiling of network components using an experimental laboratory setup that mimics real private 5G networks under various network conditions, which is a missing aspect in the existing literature. 1. Introduction Private Standalone 5G networks are getting more popular as they offer unique advantages such as greater control, enhanced security, and reduced latency. These networks are largely deployed and controlled by organizations, such as healthcare and Industry 4.0 initiatives. In academia, private 5G networks are popular, considering their unique features of ultra-low latency, unparalleled high speed, and are largely deployed as sandbox for experiments pertaining to new communication protocols, emerging authentication and encryption techniques, and thus paving their way towards 6G. A few examples of 5G in smart industries are BMW using 5G for deploying blockchain pilot and Bosch 1 1 1 https://www.bosch.com/stories/5g-industry-4-0/ deploying private 5G in manufacturing plants to improve automation and increase productivity. The University of Oulu is leading the EU-sponsored project of Future Hospital - Hola 5G 2 2 2 https://ouluhealth.fi/hola-5g-oulu-brings-to-healthcare-europes-first-ever-private-5g-sa-network/ aiming to test the applicability, feasibility and efficiency of the private 5G network in The Oulu University hospital. The project equips the hospital with private 5G test network and see its feasibility in future healthcare scenarios such as those involving extended reality (XR) requiring ultra-low latency for real-time sugeries and surgical simulations. Sustainability has become a key aspect of ICT, as there is a global shift toward greener computation and reduced carbon footprint in response to increasing environmental concerns and regulatory requirements. Rising energy prices are forcing companies to look for energy-efficient and cost-effective solutions. Private 5G networks provide an opportunity for energy optimization due to its granulated controls. Considering the data-driven applications of 5G such as Internet of Things, extensive machine-to-machine (M2M), Edge Computing and integration of Artificial Intelligence makes energy optimization a critical factor for success of private 5G networks. 2. Background & Related Work With the energy harvesting methods gaining attention and emerging research on wireless power transfer to network elements, energy profiling of the network components under different traffic patterns requires a more in-depth research focus ( Mughees et al. 2021 ) . Moreover, the Open radio access network (O-RAN) emerging as the entry point for 6G, addressing the security of the open interfaces of O-RAN is leading to increased energy consumption ( Porambage et al. 2023 ) . Element-wise energy profiling under varying conditions is key for designing green security policies for O-RAN networks. The research on energy efficiency in adaptive massive MIMO networks ( Salah et al. 2022 ) ( Khwandah et al. 2021 ) discuss overall energy savings and optimization strategies, but do not highlight the breakdown of energy consumption in individual network components such as core, radio and edge separately. The existing studies on adaptive massive MIMO systems, explore energy optimization methods using genetic algorithms, but mainly in the context of public 5G networks. There is an opportunity to adapt and test these techniques specifically for private 5G networks, which often have different traffic patterns and requirements than public networks. This research focuses on addressing the research gap of energy profiling under varying network conditions with an experimental network setup to encourage the research community to conduct more network-level energy profiling research which can be highly beneficial in designing energy-efficient 6G network architecture. 3. Methodology The energy profiling measurements are done with Netio PowerBox 4KF. In the experiment, the MQTT protocol is used to obtain the associated data from Netio. The general data collection, storage, and visualization process is shown in Figure 1 with various steps. At first, Netio acquires the the voltage (V), current (I), and true power factor (pf) for a target device that is connected to the Netio’s certain socket (step 1) and then publishes (step 2) these data to a Ubuntu VM (step 3-4), which runs in a laptop. Later, the Ubuntu VM extracts and manually calculates power with the equation 1 (step 5) and stores these in InfluxDB (step 6). (1) P=V*I*pf Finally, Grafana retrieves the data from InfluxDB (step 7) and visualizes the data through a browser (step 8). All these 1-8 are a continuous process of our energy measurement framework. Figure 1. Experiment Setup. Energy measurement framework. In the lab, OAIBox Max and Universal Software Radio Peripheral (USRP) B210 are used to analyze the energy consumption of the private 5G networks whereas the OAIBox Max includes the core network 5G (CN5G) and gNB (gNode B) while the USRP B210 acts as a radio unit of gNB. The overall hardware and software components that are used in the lab are discussed below. Hardware components: (1) OAIBox Max is running with Ubuntu 22.04 having a 5.7 GHz AMD Ryzen CPU with 32 GB RAM and 1 TB storage. (2) USRP B210 is a dual-channel transceiver that covers continuous frequency from 70 MHz – 6 GHz. (3) Raspberry Pi 5 (RPi 5) is running with Debian GNU/Linux 12 having a 2.4 GHz ARM Cortex A76 CPU with 8 GB RAM and 32 GB storage. (4) Quectel Modem (RM500-GL) with 5G enabled SIM (from Allbesmart) that acts as a UE that is getting radio signal from USRP B210. (5) Ubuntu 24.04 VM has an Intel Core i7 2.40 GHz CPU with 8 GB RAM and 25 GB storage. (6) Netio PowerBox 4KF that has 4 sockets and includes both LAN and Wi-Fi. Software components: (1) Both the mosquitto and mosquitto-clients of version 2.0.18 are installed for MQTT broker and client at the Ubuntu VM (2) InfluxDB 2.7.10 server and CLI are installed in Ubuntu VM for storing the Netio measured data (3) Grafana 11.2.2 is installed in Ubuntu VM as a visualizer to present the energy data in graph format (4) iPerf 3.16 is installed in both the OAIBox Max and RPi 5 to evaluate the performance of the private 5G network whereas OAIBox Max is considered as a client and the RPi 5 is considered as a server. 4. Energy profiles We observe the energy consumption of OAIBox (5G core) and USRP B210 (5G radio) at different stages for the 20 MHz as shown in Figure 2 . It is observed that in the idle mode (step 1), OAIBox and USRP consume roughly 6.91 Wh and 0.23 Wh respectively. For CN5G (step 2), the energy increases by roughly 1.93 Wh in the OAIBox once it starts. Additionally, 0.04 Wh and 0.03 Wh are increased both in OAIBox and USRP respectively for gNB while it starts (step 3). In the current setup, Quectel UE gets power from the OAIBox. It increases