Making Energy Efficiency Simple

The conservation of energy, efficiency and sustainability has become a paramount concern for service providers across the globe. In order to adhere to national climate action plans’ long-term net neutrality objectives or combat escalating energy costs, service providers are exploring alternatives to reduce their energy consumption and establish a sustainable, eco-friendly operation.

It makes sense for them to concentrate on their network to lower energy usage since a significant amount of carbon emissions stem from the electricity consumed by mobile network base stations, communication station premises, and data centers. Moreover, as the rollout of 5G persists and communication traffic volumes increase over time, additional energy consumption is expected.

Energy Efficiency: High energy-consuming network architectures

Traditional methods for reducing energy consumption in data centers involve adjusting workload resources based on their demand through auto-scaling and dynamic scheduling. However, this approach is challenging to apply in service provider networks where cloud-native network functions (CNFs) are typically assigned specific CPU cores and granted specialized access to system resources to optimize their performance. The 5G core UPF is one example of such CNF.

In the 5G UPF, polling and the implementation of the data plane development kit (DPDK) serve as crucial methods to ensure deterministic performance. However, polling is not affected by the load level of the 5G UPF. As a result, the associated CPU cores are continuously operating at their maximum capacity, even when the UPF is not in use or receiving minimal traffic. This means that 5G UPFs always consume the highest amount of power possible.

Energy Efficiency: Innovative reduction of energy consumption

Intracom Telecom has integrated a new functionality called the frequency feedback loop (FFL) into its NFV-RI™ solution to address the energy impact of polling in DPDK workloads. FFL assesses current demand and selects the most efficient CPU frequency accordingly. It also accurately predicts the short-term fluctuations in the traffic load of a DPDK-based CNF. By adjusting the CPU core frequency, FFL can reduce energy consumption without compromising packet delivery. This innovation opens up possibilities for reducing power consumption and thereby minimizing carbon emissions.

Unlike traditional energy-saving techniques, FFL can be implemented seamlessly in a 5G UPF without any modifications to CNFs or the need to overhaul deployment and traffic management. FFL operates without compromising performance, as it maintains low latency and prevents packet loss. Additionally, FFL has minimal resource overhead and only utilizes less than 50% of a single CPU core.

How Red Hat and Intracom Telecom can help

Collaborating together, Red Hat and Intracom Telecom have developed a completely integrated and certified solution that combines NFV-RI and Red Hat OpenShift. This integration streamlines deployment and ensures a stable implementation in service provider environments. By utilizing NFV-RI on OpenShift, 5G UPF workloads can efficiently utilize power management features to optimize overall energy consumption.

Red Hat and Intracom Telecom have released a reference architecture to expedite the assessment and implementation of their joint solution. The reference architecture outlines a simple and replicable process for enhancing the energy footprint of the 5G UPF. It includes instructions for installation and deployment, configuration options, and Helm charts for the open-source components utilized as reference workloads (such as 5G UPFs and traffic generators).

The reference architecture also encompasses two prevalent 5G data center scenarios:

• Centralized 5G core deployments with fully-loaded UPF nodes
• 5G edge deployments with mixed-workload nodes

Both of these use cases rely on the distribution of the open mobile evolved core (OMEC) 5G UPF. In each of them, the overall savings in server power are reported for a real-world 24-hour traffic pattern. The fully-loaded use case delivers more than 25% savings, while the mixed-workload use case achieves more than 15%.

Here at CourseMonster, we know how hard it may be to find the right time and funds for training. We provide effective training programs that enable you to select the training option that best meets the demands of your company.

For more information, please get in touch with one of our course advisers today or contact us at training@coursemonster.com