Telcos’ energy usage is under the microscope like never before.
A convergence of geopolitical tensions and climate change concerns are ramping up the pressure for operators to become more efficient in everything they do and wherever they do it—from network cores to data centers and beyond.
According to the 2022 Climate Action Survey Report by Telecoms.com, the reduction of energy consumption was recently flagged as telcos’ most urgent operational challenge (64% of respondents).
Across the world, thoughts of service consolidation and alternative, modern deployment models are clearly gaining traction throughout the industry as decision-makers search for ways to improve efficiency with minimal disruption and incremental cost.
Indeed, in Heavy Reading’s 2023 5G global survey of network strategists, respondents’ top approach to reducing power consumption focused on moving as many functions as possible to a common infrastructure platform (52%). This was followed by reducing infrastructure footprints and increasing power efficiency with edge computing (45%), and consolidating functions and vendors for tighter energy management and cost efficiency (40%).
Commenting on the findings, Gabriel Brown, a principal analyst at Heavy Reading, notes that “disaggregation and multi-vendor by nature introduces some inefficiencies, so thinking about how to address this is clearly important.” He goes on to add that consolidating core workloads onto a common cloud platform appears to be “the most consequential move operators can make to reduce energy consumption in the 5G mobile core.” However, the overall spread in responses indicates that operators “will combine multiple approaches in their power reduction strategy.”
In other recent study, Capgemini’s Research Institute Report—Networks on Cloud: A Clear Advantage—claims that almost half of telecom networks' capacity will be totally cloud-native in the next three to five years.
The report also indicates that operators will spend $206 million annually on that cloud transformation over the next five years. Organisations getting in early on a shift to cloud-native are likely to realize the most value in terms of economics and environmental sustainability. For the latter, the research suggests that those embracing telco cloud are expected to reduce greenhouse gas emissions by 5% in the next three to five years. Telco cloud is also set to yield sustainability benefits from lower facility emissions (e.g. reduced physical hardware footprints, less power usage, auto-scaling of network on demand, and managing mobile towers' power consumption using AI and machine learning).
This is why Cloud-Native Functions (CNF) will increasingly come into play. Or at least they should!
CNFs are software-implementations of a function, or application, traditionally performed on a physical device.
Purpose-built for moving workloads to cloud-native architectures, the technology can eliminate telcos’ “heavy” legacy virtualization software layers, as well as automate and orchestrate operations for maximum efficiency. All while scaling their networks.
It also means they can combine multiple functions on a single platform, boosting performance while using less server capacity, less CPU cycles and, therefore, less energy. For example, those running 5G networks can migrate existing network functions to cloud-native versions of the same.
It is important to note that, with consolidated CNFs, a single command (or API call) can activate multiple network functions, such as domain name server (DNS), gateway-Internet firewall (GiFW) and carrier-grade network address translation (CGNAT). This unlocks a whole new world of flexibility, including scaling workloads up and down ‘on the fly’ whenever needed. An operator could, for example, use CNFs to quickly provide compute and network capacity for a major sports event and then take it down at the end of the tournament. A traditional approach of running functions on dedicated hardware can take several months to set up.
By contrast, a consolidated CNF platform can be quickly spun up and down as required and its power consumption is kept to a minimum. The principle is the same as when your TV or smartphone switches to energy saving mode when not in use.
At F5, we’ve completely deconstructed our control and data planes to build true cloud-native functions from the ground up. We can now scale the data plane and spin up and down critical apps/functions (such as CGNAT) in a few seconds, compared to around 15 minutes for competitive CNF-adjacent solutions. A virtualized ‘wrapper’ is less effective than a true CNF and will require different functions sitting in different servers.
F5 is also introducing refactored versions of the software security functions used for BIG-IP hardware and virtual network functions in a CNF form factor, including Edge Firewall, CGNAT, DNS, and Policy Enforcer. This compatibility allows telcos to seamlessly migrate to a cloud-native architecture when it makes sense to do so, while protecting their initial investment.
What’s more, additional flexibility is available to run our suite of CNFs on F5 CNF-optimized hardware (rather than commercial off-the-shelf hardware). Offloading selected functionalities to hardware means a telco can significantly amplify performance. It also gives them the best of both worlds: cloud-native characteristics with the higher performance power of an F5 hardware appliance. As ever, choice is important—there's no one-size-fits-all approach to running more sustainably.
Even if energy prices fall in future, consolidation will likely pay off: a streamlined solution can not only perform better, and with more reliability, but also do so at a lower cost. By curbing complexity, consolidation can enable telcos to take full advantage of the flexibility, scalability and portability of 5G CNFs—the key advantages over running the same functions on dedicated hardware or virtual network functions (VNFs).
Encouragingly, all of this isn’t just theoretical tech talk, and we’re starting to see some influential moves from telcos on the consolidation front.
For example, by consolidating CNFs in its 5G data plane, Rakuten Mobile was able to cut central processor unit (CPU) usage by more than 60% while also reducing the number of CPU hops and lowering latency. The innovative telco's architecture now optimizes resource consumption, while its streamlined network structure makes it easier to deploy new services and software updates. It’s a vivid case in point of what is possible, and many more telcos are predicted to follow suit.
At F5, we also estimate that consolidating network functions in the N6-LAN can reduce CapEx by 60%, while lowering operating costs. In addition, we’ve run tests combining F5 CNFs with higher density 64-core high-performance chips, which means performance can actually be increased while the data center consumes less power and requires less cooling.
It is still early days but we expect to see telco consolidation plays and related F5 innovation—aligned to specific use cases—continue to gain momentum. Watch this space!
Keen to learn more about cloud-native consolidation? Visit F5 at Mobile World Congress in Barcelona from 27 February - 2 March (Hall 5, C60)