Is There a Hidden Cost to Network Openness?

Ulrich Kohn
Person looking out onto city through keyhole shape

Openness Creates Competitiveness

What’s the main thing that prevents communication service providers (CSPs) from responding to customer needs in a fast and flexible way? Single-vendor network solutions based on proprietary technology too often create barriers that hinder CSPs and stop them tapping into new revenue streams. Convincing their single supplier to develop new features is usually a CSP’s first step in a time- and resource-intensive service innovation effort. In a highly competitive market with margins under continuous pressure, significant upfront investment and slow innovation cycles are not an attractive option.

Open interfaces and open protocols, in combination with open source solutions, promise salvation by replacing vendor lock-in by agility and lower cost to innovate. A transition from single-vendor solutions based on proprietary technologies towards open systems will change the way networks are designed, implemented and operated. This change will transform our industry and bring with it new opportunities for any stakeholder. But, as I’ll come to later, it also carries a risk.

Innovation Builds on Open Technologies

Open standards combine the views and interest of a large community of experts. This results in specifications which address a wide range of options and alternatives. Open communities create tool boxes rather than address a narrowly defined set of functions. Hence, open solutions are highly flexible and easy to adapt towards a wide range of applications.

In consequence, leading CSPs as well as innovative vendors engage with industry alliances, standards bodies and open source communities. Today, organizations such as OpenStack, ONF, OPNFV, OIF, ETSI NFV ISG and OpenConfig experience unprecedented support.

Open interfaces lead to a logical partitioning of networks into technology domains, which allows specialized companies to focus on providing best-in-breed solutions for those domains. This approach replaces vendor-specific, monolithic networks with fast and agile service production factories built on best-in-breed technologies.

Open source software is mainly applied with common functions, which are the basis of a wide range of specialized services and applications.  Application developers can build on a common software base, allowing them to address a huge market without facing market-entry barriers of vendor-controlled interfaces. Open software in combination with a fast and agile software industry results in fast service innovation at competitive cost.

That’s why open standards and open source software allow CSPs to respond to new customer requirements and business opportunities much faster than with networks built on proprietary technologies supplied by a single vendor. Open technologies allow CSPs to benefit from solutions from many specialized players, all of which work together through open interfaces and open-source software.

Many Specialist Companies Versus a Few Big Players

The transition towards open networks is being catalyzed by the many young and highly dynamic companies producing interesting new solutions. These companies offer virtual network functions (VNFs) utilizing open VNF infrastructures or provide open switching technology for SDN-controlled networks. In this way openness creates an environment where specialized companies address a specific technology domain.

But the acquisition of Alcatel Lucent by Nokia and the cooperation between Ericsson and Cisco seem to contradict the trend towards confined technology domains being addressed by those expert companies. And a recent Ovum report suggests that big players actually own more than 80% of the $80bn equipment market.

At a time when many CSPs get sick of being choked by single-vendor solutions, it seems strange to see co-operation creating even bigger consortia.

So is something wrong with above assessment of where openness is taking us?

Don`t Confuse “Open” With “Free of Cost”

Open does not mean being simple and providing plug-and-play solutions. There is a price tag on openness. As explained above, open technologies frequently come with a wide range of options. Building a network solution on open technologies requires some level of adaptation and parameterization based on an in-depth understanding of the involved technologies. So an open solution that might be efficiently applied by an expert team could create risk and frustration when used by those without specialist knowledge.

In consequence, open technologies will need to be supported by experts, either from specialized companies or major operators who know how to efficiently apply open source software and open interfaces with specific use cases and applications. The system integration effort of putting together components from a significant number of suppliers is the entry cost for utilizing the huge saving and time-to-market advantages of open technologies.

Take a look the open source operating system, Linux. Today it’s used with a myriad of IT and network solutions, but always where it’s been integrated by technology experts and optimized for a specific application. You rarely find Linux being applied with enterprises as a general purpose operating system for PCs. Managing and maintaining a large scale Linux deployment would easily overload a typical, resource-limited IT team, as open source is optimized for functionality rather than lowest operational cost.

It’s the same with networks. Switching to an open-source system will demand a lot of engagement with standards bodies and a lot know-how. Unfortunately, open-source communities don’t provide hotline services. So, while it’s true that openness provides agility and speed of innovation, enabling systems to be built on technologies provided by fast moving, highly innovative companies and communities, the price tag of openness is effort. A lot of hard work is required to integrate, test and operate networks built on solutions provided by a several specialized suppliers.

Resolving the Paradox

The integration of fixed and mobile networks based on open technologies requires the expertise of a system integrator with end-to-end solution competence. A service provider with this competence can take responsibility for integrating components from various suppliers, performing the integration test, maintaining the network and in some cases even operating it.

Tier 1 operators might possess such expertise while smaller operators will tap into the competence pool of major system integrators. These frequently run networking partner programs for the pre-integration of multi-technology solutions and for competence sharing.

In an effort to create this competence, even major companies need to join forces – mobile network specialists with fixed network experts, public network specialists with IT experts and system integrators with vendors who have products as well as operational competence.

In future we should expect to see those giants putting more focus on services rather than products, a journey that leading vendors such as Ericsson started years ago. The increasing shift from products to services with any of those major players will create competitiveness between system integrators who come from the enterprise IT market and those who provide major system solutions for public networks.

Will this lead to single vendors being replaced by end-to-end system integrators who will dominate and control a network? CSPs will be very keen to not allow any partner to take a role which can’t be substituted by a different player. The recent decision by Telefonica to re-tender MANO of their NFV solution is a case in point. After all, our industry knows very well that the “too big to fail” paradigm has long since ceased to exist.

Related articles