Clash of the Titans: The Mobile World and Fixed Broadband – Who Will Lead in Virtualized Application Networks?
Virtualization of mobile networks must first occur in operators IT Networks and support processes before it can be made widely available to customers as either a reformulation of current service offerings or new horizontal and vertical market offerings. Gating factors include the availability of 3.5 GHz and e-Band mmWave and higher spectrum, availability and spread of 5G devices that include access to 2.5-2.6 GHz, 3.5-3.7 GHz and 5G-ready infrastructure equipment.
The push for mobile operators towards use of IT network standards including the embrace of virtualization has been driven by shared market economics as well as needs for interoperability with fixed networks. While current RAN, Radio Access Neworks, have achieved about a 25% level of virtualization according to Ericsson[i] Much of the virtualization efforts for RAN is centered around internal Telecom IT rather than customer service level offerings. That is because both the demand-side and the spectrum enabled network supply-side of fielding virtualized network offerings must coincide. The diverse operating environment MNOs deploy into and the need to evolve towards fulfillment of capacity and automation of services similar to that achieved for core IT network services and applications must evolve. MNO’s primary focus remains on fulfilling growing bandwidth needs and offering additional or enhanced services, most of which are already enabled by 4G. The RAN environment is rapidly changing but will take years before it reaches a tipping point in the way it delivers services to customers.
The profusion of white papers on the topic of virtualization of the RAN and combined wired-wireless service offerings that would be made possible tends to be oriented based on the source:
The IT/networking equipment suppliers push the adoption of their equipment and virtualization approaches including CRAN, Centralized RAN. The equipment and software are to be decoupled, shifting the RAN and network management functions to Network Functions Virtualization or NFV[ii]. The fault we see in this emphasis is that it does not address the complexity of the RAN environment, in essence is an attempt to put the horse before the cart in the pursuit of delivering marketable services to MNOs.
The mainstream mobile wireless equipment suppliers, including Ericsson, Nokia, and Huawei, show a much deeper understanding of the complexity and dualistic nature of the virtualization of the RAN. MNOs must first focus on virtualization of their own IT and automation of services before their service offerings can extend virtualization to broad markets. Along those lines, the RAN is encompassing the need to support the IoT use-case services. Even though majority of the IoT services needs run on top of the 4G-5G mobile services network rather than as unique services, they may be among the first to be provided as a virtualized service offering. Otherwise, we agree with Ericsson’s view that while a matter of necessity and expedience, this will be arrived at over time in a way that evolves with mobile broadband markets. The majority of market needs are now met by service contracts that place customers into classes of service such as ‘unlimited’ packages with most content and applications selected as OTT, over the top, by individual users. For this to evolve to a individual customer-tailored service environment, as envisioned by Ericsson and Huawei, the MNOs IT will need to implement virtualization and service automation. The impetus for MNOs to virtualize and market their network services to end markets depends on the ability to achieve market retention and containerized services. It makes little sense to virtualize the network for the sake of providing a free market for OTT vendors unless this can be tied to high customer retention and compensating package revenues.
When, how and where will Mobile Network Virtualization (MNV) take place?
The topic of Network Virtualization is broad, diverse, and complicated by the rapid evolution of spectrum and technology enabled access. It takes place on top of nationwide deployments of 3G-4G broadband networks and markets that have grown to demand high availability of video and communications services. What we see as being drivers for MNV are/will remain the end market solutions and differentiations that will drive end-user satisfaction.
In coming months, Maravedis will explore aspects of the forward outlook on 5G and MNV including the following:
- NFV, Network Function Virtualization as this pertains to
- Telecom IT
- Enablement of virtualization enabled/enhanced customer services
- Automation of the personalized service experience
- Social Network experience enhancement through 5G service-level slicing
- MNO participation in Virtualized and AI applications development and deployment
- Seizing on the new opportunities of Localized and adapative content delivery networks, LCDNs
- Working with existing dev/ops vs. in-house developments: Which is likely to further operators objectives and develop market momentum?
We titled this “Clash of the Titans: The Mobile World and Fixed Broadband – Who Will Lead in Virtualized Application Networks? Because there is convergence taking place between the two spheres of networks and because much of the writing on the topic tends to isolate virtualization as being an equipment and back-office IT software pursuit.
Mobile networks are much more intimately entwined with delivering service directly to end customers and immediate intermediaries and are much more in control of the end-to-end experience. Because of that, customers both expect ‘the complete package’ and networks are built more on a monolithic basis. This will be a core understanding for future discussions.
[i]Ericsson Telecom IT for the digital economy, October 27, 2017
[ii]Cisco, Network Functions Virtualization (NFV), by Santanu Dasgupta Sr. Consulting Engineer –Service Provider Network Architecture