What is the market for network slicing?

Atila Horvat, Head of Network Slicing at Rakuten Symphony, examines how network slicing can be a viable market offering for mobile operators and offers a pathway to success.

As one of the original architects of network slicing for mobile networks, I’d like to offer perspective on the real market opportunity for slicing without the hype, false hope and delusion that sometimes confuses conversations.

Since I started working with network slicing in 2015, it has increasingly become a hot topic of discussion. Some believe it will solve telecom’s revenue growth problems. Others fear it is analogous to the ‘The Emperor’s New Clothes’ folktale, with everyone nodding along instead of speaking up about the problems they see right before their eyes.

My years of experience in this market tells me there is value in network slicing if correctly approached—‘correctly’ being the operative word.

In this article, I’ll dive into network slicing market understanding and readiness, implementation challenges and economics.

Why cover network slicing in an automation and AI-focused newsletter?

Network slicing can only be successful if it is zero touch and on-demand, from service ordering to total operational and lifecycle management. We cannot make the mistake of creating a market with potential but no way to economically scale or quickly meet demand. It is important to remain realistic, grounded and focused on real opportunities that solve business problems.

What is network slicing?

Network slicing is a software-defined logical network overlaid on the physical, virtual or hybrid physical and virtual network. It is already a familiar concept in the broader networking and technology market. VPNs create, for example, a secure logical tunnel for enterprise security, and virtual machines and containers create logical versions of servers on top of physical servers.

The basic premise is that one-size-fits-all networks (i.e., 4G) are not capable of addressing the unique needs of certain business segments or advanced applications in an economically viable way. But if we carve up the network into slices with unique attributes like high security or low latency, mobile operators can appeal to specific needs of specific customers willing to pay more for the privilege.

Network slicing gives operators a way to more precisely tailor connectivity to specific business cases, applications, services or users (e.g., VIP users), and thus monetize the network to a higher value.

Think of network slicing like selling cakes.

If the different network elements, network functions, network function services, transport resources and end points are the cake ingredients, network slices are the different cakes that can be made from a combination of ingredients.

Network slicing limitations from a network slicing realist

Network slicing only works on 5G networks and is not federated across mobile operators by default. It is an excellent tool for managing experience, traffic and security, specifically on individual mobile networks.

5G Standalone (5G SA) fully enables the capabilities of network slicing, providing a framework that allows this service to be explored. To actually deliver network slicing, the framework needs to be further instrumented, operationalized and enhanced for a specific business case, application or service.

Developer APIs should be designed and aligned with industry efforts to simplify and unify external exposure. Network slicing is not a mass market API market and is not focused on B2C markets. It was always designed to enable B2B and B2B2C markets and the more specific developers therein.

Stakeholders should not worry about the mass adoption of smartphone app developers. Instead, efforts should be spent making adoption as easy as possible for internal, B2B and B2B2C markets developers and select external communities.

What types of network slices meet a real market need?

Every market will be different and every operator will need to consider different factors when deciding which markets to pursue.

Some of the most viable use cases Rakuten Symphony has identified revolve around meeting specific requirements for highly viable needs. When we advise operator clients, the use cases we have been evaluating most closely include:

• Emergency service slices. By creating a dedicated slice for emergency communications, operators can reliably prioritize critical communications services. This could potentially replace older, dedicated emergency networks, making more efficient use of resources and better integrating with other network services.
• Assisted living slices. Network slicing could be leveraged to provide enhanced services for people with disabilities, such as the visually impaired. Network slices could support applications (e.g., AR/VR) that help these individuals more effectively navigate outdoor environments for improved quality of life and independence.
• Remote broadcasting slices. Network slicing could transform remote broadcasting by reducing dependence on traditional broadcasting trucks, equipment and satellites. Instead, broadcasting services could utilize dedicated network slices for transmission of high-quality video and audio with lower latency and higher reliability.
• Private network slices. Network slices could competently serve enterprises requiring secure, customized network environments. Operators can offer enhanced control, security and efficiency tailored to specific operational needs, supporting dynamic scaling and adjustments without extensive physical infrastructure changes. This can be accomplished with minimized TCO integration between private and public networks to open the door to hybrid operations, expanding potential applications and reach of private network solutions.
• MVNO slices. For example, an MVNO can purchase a slice that maximizes the performance of video download or another that maximizes an important capability like advanced security or privacy. These new capabilities can enrich the MVNO market, for example, by opening the possibility for highly specialized MVNOs.

Once you’ve zeroed in on viable use cases, it is best to take a phased approach to rollout to really understand the market, experiment and learn what works.

Start slowly and become your own first customer

Success in slicing is not guaranteed. But watching this early market develop, we understand that there are steps every operator can take toward determining if a viable opportunity exists:

• Make sure networks are “slicing ready.” The right infrastructure, management tools and support systems for managing the complexity of multiple, simultaneous network slices needs to be in place.
• Start with legacy services and go slow. Rather than try to sell a custom slice to early customers, operators should become their own first customer. Create a slice to carry legacy services like voice, video and MMS to gain experience in a controlled environment before expanding into more complex use cases. This crucial step will help develop expertise and understanding about how to effectively manage and optimize network slices.

• Educate the market. Operators should actively engage prospective customers with education and listening campaigns to fully understand how network slicing can meet their specific needs. No two markets will have the same requirements.
• Focus on B2B. Telecom understands from 5G that upselling consumers is hard. On the other hand, enterprise users have a clearer and more immediate need for specialized network services. But not all B2B use cases will be created equal so time spent proving a market exists becomes invaluable.
• Prove there is a business case. Slicing is not a 'build it and they will come' proposition. Just because you offer a slice that theoretically has major market addressability doesn’t mean customers want to buy it. Be certain that a market exists before carving up your network to chase it.
• Prove that telecom is up to the task. Finally, operators need to build trust and confidence, proving that 5G network slices can solve a customer need better and cheaper than legacy or new competing solutions. Enterprise means mission critical and moving customers away from what works to something enticing but unproven is an uphill battle.

There are of course other considerations. For instance, optimizing slice location (e.g., edge, regional data center, centralized data center, etc.) based on the customer, application, and required cost and performance will ultimately be an important component of the delivery calculus.

But as I said above, nailing the basics will be the single most deciding factor in whether an operator can successfully pursue this market.

A slice of reality as we look to the future

We do not know what the future holds for network slicing, but telecom can become better positioned to test the market and determine the best path to success.

Network slicing could be the key to unlocking real-time, efficient customization of network resources tailored to specific user needs and applications. It could be the 5G development that finally creates a paradigm where end users–not just operators–can influence or even control network characteristics.

Lest we stray too far from the stated focus of this newsletter, I should also reemphasize that management through AI and automation is a prerequisite for realistically introducing any kind of network slicing at scale. AI and automation will be required to manage the allocation and optimization of slices, and guarantee SLAs by enhancing the network’s responsiveness and efficiency.

Finally, as sustainability becomes a priority, network slicing could potentially optimize network resource usage and become a key contributor to more energy efficient operations.

But first thing’s first: let’s all do the hard work to give slicing its best chance at winning the 5G future.

‍