Skip to main content




Here's how a 5G Distributed Cloud Network will help you shift markets and scale

By Sue Marek, Special Technology Contributor

One of the big benefits of deploying a 5G network based on a cloud-native model and incorporating a distributed cloud network is that it can be manipulated on the fly to accommodate a number of different scenarios or use cases. This is also called network slicing.

Consider the following scenarios. A CSP competes with a fixed broadband provider by delivering fiber-like broadband speeds and reliability to homes and businesses via a 5G fixed-wireless network. Or a highly specialized doctor, wearing a virtual reality headset connected to a 5G network, performs an in-depth diagnostic evaluation on a patient in a remote telehealth setting. Or a local Port Authority uses 5G to connect 4k video surveillance cameras to the network, making it possible to provide real-time monitoring of the area and increase safety at the port.

These are just a few use-cases that are possible thanks to the transformative nature of 5G, and state-of-the-art technologies, like distributed cloud. According to a study conducted by Omdia and commissioned by Nokia, these scenarios — fixed wireless access, eHealth services, video surveillance and analytics — are considered by CSPs to be the top global use cases for 5G.

An open, distributed cloud architecture is effective for reducing operating expenses and capital expenditures, while at the same time allowing CSPs to provide better connectivity to their customers and enable new use cases, like those mentioned above. As COVID-19 forces shifts in the markets and the economy at large, it’s more important than ever for businesses to be agile and take advantage of these technologies.

IT industry roots

Mobile operators believe that open networking technologies and distributed cloud architectures will help them develop and deploy telecom networks more quickly. When we consider the explosion in mobile data traffic and growth in wireless connections, an efficient response cannot be understated.

It’s important to note that these components are also key to deploying 5G. In fact, many CSPs view 5G as a catalyst for making these changes in the network —the confluence of 5G and the cloud-native world provides the opportunity for CSPs to completely transform their networks.

In November 2019, a GSMA Intelligence survey of mobile operators from around the world asked participants about their priorities when it comes to their network transformation strategy. Incorporating IT industry strategies such as virtualization of the network and cloud architectures was ranked as a top priority for operators and was only surpassed by network security. In fact, cloudification of the network was considered a higher priority than energy efficiency, vendor diversity, automation or even regulatory compliance.

But the survey also found that the reality didn’t really match the enthusiasm operators had for the cloud. In other words, while the operators surveyed said they considered the cloudification of the network to be a top priority, in reality only 12% of those surveyed had virtualized and cloudified their networks at scale commercially, while 17% had started to commercially deploy the technology. Only 18% of operators were testing the technology and 35% were in the planning stages. And 18% of the operators surveyed said they currently had no plans to deploy the technology.

While this might underline a hesitancy that exists amongst CSPs when it comes to actually making the changes to optimize their networks, it also shows how unprepared many are to deal with massive surges in traffic, or to use cloud’s scalability instead of overbuilding to handle surges such as that brought on by the current pandemic. By migrating to a distributed cloud, CSPs can become more efficient, cost effective, and agile.

What is distributed cloud?

The distributed cloud is a significant change in how telecom networks are designed and orchestrated. In the simplest form, distributed cloud refers to the distribution of cloud services to different locations. This is different from the centralized cloud model in which services are centralized at one location, such as a data center.

Today many cloud-based networks reside in central private clouds. However, in the distributed cloud model, workloads are moved between the central cloud, the regional cloud and the edge cloud and also distributed between both private and public clouds, such as Microsoft’s Azure or Amazon Web Services (AWS).

The benefits of the distributed cloud model are that it can reduce latency, network congestion and data loss for operators. It also makes it possible for CSPs to deploy and deliver new services more quickly because they can run workloads anywhere, making it possible to meet the demands of emerging 5G use cases.

But there are certain requirements that are needed to effectively deploy a distributed cloud architecture. For one, CSPs need to have a clear understanding of how and where an application should reside in the network and how and where it should be processed. This will help determine the network resources (such as compute and storage) that are needed.

CSPs also will need to have some automated systems in place to reduce the complexity of the distributed cloud, such as a centralized management system that will provide a view of all the servers and devices in the network and an automated provisioning system that can manage the workloads.

Edge, regional or public cloud

Edge clouds are an integral part of edge computing, which is computing that is done at or near the source of the data. One example of an edge cloud is one that is located at or near the base of a cell tower enabling the CSP to process data near the radio instead of sending that data back to a regional cloud or to a central cloud. By processing the data at the edge, a CSP will be able to reduce the latency (which means less delay) making real-time applications such as multiplayer gaming or health monitoring possible.

Another advantage of an edge cloud is that when data is stored locally, or near the source, it has less likelihood of falling into the wrong hands or getting lost, so it is more secure. And finally, if data is stored locally it doesn’t take up a lot of bandwidth on the network being transferred back and forth to the central cloud. That makes the network more efficient and keeps costs down because CSPs are able to reduce their transport expenses.

Regional clouds are another tool in the distributed cloud architecture - this is where CSPs can handle non-real time functions such as vehicle diagnostics that do not require the low latency of the edge cloud applications. By moving some applications to regional clouds, operators are able to efficiently handle bandwidth and costs.

Public cloud is also a key component to the distributed cloud network. CSPs may want to use the public cloud to off-load data during certain peak times in the network or on a case-by-case basis. For example, during a popular event or a disaster, like the current pandemic, the network may become strained from all the usage. In this case, a CSP may want to offload some data to the public cloud to ease congestion.

Another example of an effective use of the public cloud might be when launching a new service. In this scenario, an operator may want to use the public cloud to temporarily store an application until demand for it grows instead of building out its cloud infrastructure in advance. Savvy businesses will respond and shift gears by shuttering some services in favor of new ones, making public cloud a very attractive option for transitioning.

It’s important to remember that in a 5G network, there will be a lot more devices, bandwidth and applications to manage. That’s why being able to move different functions to the cloud while steering traffic and sharing resources will make the network more efficient and cost effective.

Transitioning the network

To move away from the traditional telecom network architecture and towards a cloud-native architecture, CSPs will need to first transition to a virtualized network by incorporating tools such as software-defined networking (SDN) and network functions virtualization (NFV). This will allow the operator to move traffic to different parts of the network as demand fluctuates and will provide the stepping stones to new capabilities such as network slicing.

Some wireless operators have already decided to transition to this new software-oriented network architecture even though it promises to be a technological and operational challenge.

AT&T is one global operator that has publicly committed to virtualizing 75% of its network by year-end 2020. “We are on track to control 75% of our core network functions with software by the end of 2020 and we are nearly there,” said Scott Mair, Senior Vice President of Technology Planning and Engineering at AT&T, in a January 2020 blog post. “Today, 100% of the data traffic that runs through the infrastructure connecting the elements of our core network together is backed by SDN.”

And while SDN and NFV make existing 4G networks more efficient and competitive, these technologies are an integral part of 5G. In fact, the standards group, the Third Generation Partnership Programme (3GPP), has included a service-based architecture that allows for flexible deployments of session management control function and user plane function in its Release 16 of the 5G standard.

Adopting a cloud-native model

Cloud-native is the concept of building and running applications that take advantage of the cloud computing delivery model. Cloud-native applications that are designed for the cloud can be deployed more quickly and also have a more fluid architecture, which means they can be placed and moved into different environments easily.

Adopting a cloud-native model is a key part of moving to a distributed cloud model. CSPs will have to consider whether they want to adapt their legacy applications that were built before the cloud to have cloud-native characteristics. In addition, they will need to determine if all future apps should be cloud-native or whether some will not adopt cloud-native characteristics.

So far, cloud-native has primarily been used for applications such as voice-over-LTE and in cloud radio access network (C-RAN) to decouple the hardware and software in the radio access network.

But many legacy applications are too manual. They can’t be reconfigured into cloud-native applications because they will be too unwieldy, and will use too many cloud resources. By re-architecting the network -- and using techniques that Webscale firms such as Google, Amazon and others have used to create their networks -- CSPs can take advantage of this more efficient architecture and pursue new business models.

Netflix, another massive web scale company, relies heavily on this kind of efficient cloud-based network architecture. The fact that the video streaming service has such a massive global membership, with regional populations varying in use, means they have to juggle different sets of needs for different groups of people.

For example, Netflix was recently asked by the government of France to reduce its streaming quality in the country. The French government did so because more people were staying home due to the pandemic, which means more people using at-home entertainment (like Netflix). By reducing the quality of video streaming, Netflix reduced the overall bandwidth being used, alleviating some stress on European networks.

Having a distributed cloud network enabled Netflix to be fluid and allocate bandwidth where it was needed (and not needed) quickly for the benefit of many.

Incorporating network slicing

One of the big benefits of deploying a 5G network based on a cloud-native model and incorporating an end-to-end distributed cloud network is that it can be manipulated on the fly to accommodate a number of different scenarios or use cases. This is also called network slicing.

Network slicing allows operators to design, deploy and customize different “slices” of the network that are sharing on a common network infrastructure. Some even call it “dynamic” network slicing because these different customized slices will be able to be designated very quickly.

With social distancing measures in place, telemedicine appointments may skyrocket, putting stress on networks. They can better manage this traffic and avoid appointment interruptions by creating a new network slice with specific parameters. This way, the new slice can allocate bandwidth to only telemedicine traffic, ensuring doctors and patients secure and steady communication.

By using slicing in a similar fashion, CSPs can quickly prioritize the network to whatever concerns are most pressing and eliminate congestion, making the user experience better.

Making the 5G promise a reality

CSPs that want to take advantage of all the benefits and power of 5G such as being able to offer ultra-reliable low latency services and network slicing will need to embrace the distributed cloud model. Virtualizing the network will play a pivotal role but that alone isn’t enough.

Being able to offer a diverse mix of services —to both consumers and enterprises — will require a new network architecture.

Many wireless operators around the globe have started on this path. It isn’t a quick or easy transformation. But the transition will be necessary if they want to stay competitive and thrive in a post-pandemic world.

About the author

Sue Marek


Sue Marek is a Denver-based freelance writer that has been reporting on the telecom and tech industries for more than 25 years.

You can follow her at