logo_header
  • Topics
  • Research & Analysis
  • Features & Opinion
  • Webinars & Podcasts
  • Videos
  • Event videos

How telcos can make the edge their own

CSPs must work together and with third parties to share insights and innovate, to effectively leverage standalone 5G core networks to enable network slicing and multi-access edge computing.

15 Mar 2021
How telcos can make the edge their own

How telcos can make the edge their own

This article was contributed by Accedian, a member company of TM Forum. CSPs must work together and with third parties to share insights and innovate, to effectively leverage standalone 5G core networks to enable network slicing and multi-access edge computing. Service innovation is about to heat up as communications service providers (CSPs) leverage standalone 5G core networks to enable network slicing and multi-access edge computing (MEC). The opportunities for growth are huge. But we’ve seen this before with cloud computing. In the end, it was the cloud hyperscalers and not network operators that captured the market. To ensure a different outcome this time, CSPs will have to win customers over on security and performance, and they will need to work together and with third parties to share insights and innovate. New revenues from advanced consumer services that use augmented reality (AR) and virtual reality (VR) and from enterprise services delivered to industry verticals will depend on standardized platforms, such as TM Forum’s Open Digital Architecture (ODA) and Open APIs, so that developers know what to expect and can freely innovate. The need for guaranteed 5G performance combined with the complexities that come with multi-carrier, multi-vendor, multi-cloud and multi-region networks mean that a consistent way of understanding and assuring experience across all these infrastructures and vendors becomes an architectural imperative. Providers must ensure that critical video communications or a connected car is not going to experience a bottleneck or degradation right at the moment of truth!

Edge is already happening

MEC and 5G do not represent a far-away future. Technology deployments are already gaining pace. Gartner claims that by 2022, more than half of enterprise-generated data will be created and processed at the network edge. According to Nvidia CEO Jensen Huang, “We’re about to enter a phase where we’re going to create an internet that is thousands of times bigger than the internet that we enjoy today.”

On paper, CSPs are in a great position to help to drive this new wave of innovation. Edge data centers must be close to users and devices, which gives telcos and their cell sites an advantage over the main cloud providers, whose centralized facilities could be many hundreds of miles away. But the location of servers is not the only calculation here, and it’s still undecided who will provide what.

In one model, cloud giants like Amazon Web Services (AWS) and Microsoft Azure provide small edge nodes to enterprises to install directly on their premises, for example. In another, units like autonomous vehicles are fitted with on-board compute capabilities, enabling them to instantiate a node from the edge provider closest in proximity. The

MEC market is set to be worth

nearly $16 billion by 2025. For CSPs, losing out on this opportunity to the likes of AWS et al again is not an option. That is why operators are partnering with the big cloud players now, to ensure they get a slice of the pie. For example, Verizon has teamed up with AWS and Microsoft, while AT&T has multiple edge cloud deals in place with Amazon Web Services, Google, IBM and Microsoft. But as consultancy firm STL Partners warns, it may still be difficult for CSPs to move up the value chain. They must think strategically and act fast, or risk losing a substantial market share once again to the cloud giants.

Telcos have the edge with SLAs

Owning the edge real estate is certainly an advantage in the ongoing 5G discussions, but it is far from a trump card. There will also need to be a major focus on managing network performance. MEC applications are not tied solely to the network edge but must traverse multiple network environments. That means operators must assure performance end-to-end: from devices to the edge, to private and public clouds – including microservices-based architectures. Imagine how many different operators and service providers could be in that mix.

Finding a provider who can offer a “common trust model” that provides comprehensive visibility into infrastructure and applications will not be easy. The granularity of insight needed into latency and packet loss is particularly important for certain business-critical applications, such as robotics or augmented reality used in manufacturing.

Another consideration is how to collect this crucial network and application data: Sensors should be lightweight enough to scatter liberally at various locations around the edge, sending information back to a centralized AI-powered analytics engine for insight. The data also needs to include not only performance metrics, but also security metrics to spot anomalies that could indicate an attack. With the right data, if there are any gaps in the protection of these environments, it won’t be long before nation states and cybercrime groups find them. Finally, CSPs need to trust and cooperate with each other more. The only way to create a successful edge market globally is to standardize as much as possible, so that device makers, developers and other key stakeholders know what to expect, wherever they are in the world. That’s an approach that worked pretty well to help the biggest cloud providers grow and scale, and it must happen here too. How it all ends is anyone’s guess. But we know one thing: The battle for growth in a new world of rapid innovation will be won or lost at the network edge.