The Importance of Multi-access Edge Computing in a 5G World

By Iain Gillott

November 9, 2017

Multi-access edge computing (MEC) servers runs virtualized software on general-purpose computing hardware in edge nodes that can emulate parts of the core network, serve as reliable caching units, run virtualized applications from any number of an operator’s developer partners or a combination of these steps.

Mobile operators are preparing their networks for 5G. Part of that process includes moving to software-defined networking and network functions virtualization. Multi-access edge computing (MEC) and Central Office Re-architected as a Data Center (CORD) are two subsets of the overall shift away from the traditional network architecture to one that looks more like a data center.

Multi-access edge computing emerged on the wireless industry stage several years ago. It has the potential to be as disruptive a technology as anything that is being discussed today. In fact, MEC is quite likely to help realize the promise of 5G.

Simply put, MEC marries a radio with a data center-like hardware architecture. Today, that radio is LTE, but it could also be Wi-Fi, 5G New Radio or some combination of them all. The hardware component is a secure, virtualized platform that network owners can open up to third parties — content providers and application developers, for example. In so doing, the network owner allows content to be placed at the edge; that is, close to the end consumer of that content. That content can be anything — streaming video, augmented reality, location-based services, connected vehicle and internet of things applications.

By putting content and applications at the edge, the network owner can achieve operational and cost efficiencies while introducing new services, reducing network latency and, ultimately, improving the end consumer’s quality of experience.

iGR, my market strategy consultancy focused on the wireless and mobile communications industry, believes that operators will primarily deploy MEC to improve the function and efficiency of their networks — caching and local breakout, for example. Those same MEC units can be used as platforms on which other businesses could host and deliver their content and thus generate new revenue for the operators.

Defining 5G

Although MEC does not require a 5G network to be deployed, 5G is likely to need MEC, along with other virtualization solutions such as network functions virtualization and software defined networking.

That begs the question: What is 5G? According to the Next Generation Mobile Networks Alliance, “5G is an end-to-end ecosystem to enable a fully mobile and connected society. It empowers value creation toward customers and partners, through existing and emerging use cases delivered with consistent experience and enabled by sustainable business models.”

More specifically, the following points appear to represent the consensus view of what a 5G network should eventually be able to achieve:

· 1–10 Gbps connections to end points in the field (i.e., not theoretical maximum per-cell throughput, but actual end user speeds).

· 1 millisecond end-to-end round trip delay (latency).

With respect to mobile communications, cellular networks’ MEC has more to add in the realm of latency — reducing it — than it does in the case of raw throughput.

What is network latency?

Improving the latency of a network decreases the amount of time an end user has to wait for a response. Put another way, latency is the time the network takes to actually process the request. Each component in the network adds a little latency — the more hops there are in a network, the greater the latency.

Figure 1. How latency adds up. Source: iGR, 2017

Figure 1 illustrates how latency adds up per element introduced into the mobile operator’s network. Note that the total one-way latency caused by the network is approximately 35 milliseconds — a round trip is 70 milliseconds. This does not include time for any application processing that may be required at the data center. If the consumer is accessing a piece of data in particular that is in higher demand, the total latency may be longer.

Clearly, reducing LTE network latency from an approximate average of 70 milliseconds to 1 millisecond would require a massive amount of expense and effort — especially considering that the average fiber network’s latency is just shy of 20 milliseconds.

iGR believes that the 1-millisecond latency target for 5G only applies to the latency on Layer 1 and Layer 2 of the given mobile operator’s network. Once the subscriber steps a virtual foot off the mobile operator’s network into the public internet, latency could spike dramatically and for reasons completely outside of the mobile operator’s control.

This may argue, then, for applications and services that demand extremely low latencies to be tightly integrated into the mobile operator’s network. It may also argue for new architectures such as MEC, which can put both processing power and content closer to the end user.

What Is Multi-access Edge Computing?

MEC is led by an Industry Specification Group (ISG) within the European Telecommunications Standards Institute that has been set up by Huawei, IBM, Intel, Nokia Networks, NTT Docomo and Vodafone. Since late 2014 when the formation of the MEC ISG was announced, more than 50 companies have joined the effort to create an open standard for mobile edge computing. The MEC ISG has been generating considerable interest in the market, and participation in the effort continues to grow.

In short, MEC provides cloud-computing capabilities and an IT service environment at the edge of the mobile network — or any network. Each of these MEC servers can be placed where small cells would be likely to be placed or at other locations — central offices and head-ends, for example.

Each of the MEC servers runs virtualized software on general-purpose computing hardware that is contained in a secure form factor. These edge nodes can emulate parts of the core network, serve as reliable caching units, run virtualized applications from any number of an operator’s developer partners or a combination of these steps. Each MEC also provides low latency, high bandwidth and real-time access to radio network information.

Figure 2: Reduced latency with apps/content server at the network edge.

All of these capabilities are either built into the hardware of the base station itself or into a box that is collocated with, and connected to, the base station. As Figure 2 shows, by moving the content and apps server closer to the edge of the network, latency is reduced considerably. Applying other techniques, together with MEC, will allow mobile operators to further reduce latency and get much closer to the goals of 5G.

Iain Gillott is the founder and president of iGR, a market strategy consultancy focused on the wireless and mobile communications industry. The company researches and analyzes the effect new wireless and mobile technologies will have on the industry, on vendors’ competitive positioning and on its clients’ strategic business plans. Visit www.igr-inc.com.