Openness in networking is a very simple concept: a specification, protocol, design, or practice is jointly developed by experts from several companies, and the outcome can be used for free. Teamwork like this can save a lot of time and money. The cooperation from stakeholders of the entire value chain ensures the value of the result.

On the other hand, why do companies share their competence and expertise with competitors? What can be so valuable that it compensates for this risk of giving away a competitive advantage?

There’s no doubt that open networking is getting a lot of interest. While openness initially addressed the standardization of interfaces, the scope is now being extended towards whole networks and open designs for any component of it. Will this trend continue, or have we overdone it? Will we move back to a closed customer engagement and vendor-specific solutions?

Let’s take a closer look at different concepts of openness, understand the motivation, and – based on this analysis – make some predictions for open engagement in 2021.

Many flavors of openness

“Open” networking is an interesting concept. It’s multi-facetted with very different motivations, value and implications for different network entities. A graphical representation of open entities in networking is shown in the figure below.

In contrast to closed approaches, an open architecture enables fast innovation. That’s because it’s not a proprietary system which can only be modified or extended by its designer and implementer. However, there are some downsides. Let’s take a closer look.

Open protocols and interfaces

There are open communities working on standards and specification for interfaces and protocols such as the Open Internetworking Forum (OIF) or IETF. Competing companies engage to develop an interface that can be used thereafter for ease of network integration and achieving a well-defined functionality. 

Participating vendors have the opportunity of early market entry. But they may steer the requirements to create reliance on their own proprietary technologies and gain a competitive advantage. We often see this situation in cellular technologies. True openness in the definition of interface and protocol specifications assures that proprietary technologies cannot limit the number of players in an unfair way.

Open network components

Several open communities such as Open Compute Project (OCP) or Linux go further and don’t just jointly develop specifications but also provide hardware designs, software code or best practices to the public. Applied by a solution-vendor, those network components and best practices reduce cost for development and simplify the design and implementation phase. 

This is why there’s a strong case to use those open entities. But why should a company invest money into something which might help a competitor to build its products and services? 

Frequently, those open components are essential parts of a solution. They are applied in different use cases in combination with other technologies and solution components. Compatibility with this enabling technology simplifies solution integration and helps a supplier to position its product. If a company provides a software application running on Linux, there’s a strong reason for making Linux widely applied, and therefore for supporting this open software community. 

Open network architectures

Combining several open designs results in open architectures or open networks. One of the most prominent candidates is Open RAN, an alliance which was kicked off in 2017. Since then it has gained the support of more than 25 major mobile network operators (MNOs) and over 200 suppliers. 

The Open RAN approach addresses complete network domains and specifies the internal interfaces, including the functionality of its building blocks. It might apply open designs, but it also creates an ecosystem of vendor-specific implementations leveraging open specifications. 

The emergence of these open architecture communities is impacting the way that networks are designed, implemented and operated. Just look at AT&T’s ECOMP approach and further engagement with established NMS solution providers as a system integrator. This might be a blueprint for transforming open architecture designs into a new network lifecycle practice.

Assessing today’s deployment of open solutions

Today, open interfaces are widely applied. Open software components are very successfully used in some application domains such as operating systems. In contrast, open hardware designs haven’t fared as well. This might be related to the principal difference between software and hardware and the nature of open design, which are a kind of smallest common denominator among the involved parties. This is okay with software, which can easily be extended with additional capabilities. Missing hardware functionality is much more difficult to add. 

Open networking creates multi-supplier solutions with a need for system integration. This becomes more relevant as principles of openness are extended from the interface to network components to end-to-end network designs. With increasing openness of a solution, the level of disaggregation also increases, resulting in more effort being needed to compose a solution. 

There are few companies which own the competence and credibility in integrating disaggregated multi-vendor networks. Mostly, those are the suppliers of legacy, monolithic single-vendor solutions. This creates the risk of networks once again being locked down by dependence on the system integrator.

The route to open networking is not an eight-lane motorway but a bumpy road that requires some care, especially if you want to drive fast. However, even when not traveling at full speed, there are advantages which shouldn’t be missed. 

2021 value proposition of openness: on the transformational nature of openness

As outlined above, openness is gaining momentum. It will increasingly impact not only network components but also the way that networks are planned, integrated, implemented and operated. Emerging concerns about substituting a dependency on single vendors with reliance on single-system integrators need to be mitigated. This is a continuous transformation, and the next major steps might happen in 2021:

• Service provider networks will continue the transformation to open, disaggregated and agile networking. We will experience a strong focus on the access network, as cost pressure is most intense here. It’ll be interesting to keep an eye on the progress of the Open RAN Alliance in mobile in combination with the work of TIP. Or have a look at the progress of CORD at the Open Networking Foundation (ONF).
• Networks will consist of vendor-specific building blocks and common platform technologies such as operating systems and general management tools. The common functions in management and control will increasingly be open source. Today’s suppliers will move into the system integrator space.
• Companies with specific technology expertise will focus on solution domains and will provide key building blocks, supporting open interfaces for ease of integration. Previous end-to-end solution suppliers will re-focus on their key technology and will replace their end-to-end business with a comprehensive service offering ranging from consultancy to operations.
• Today, there are a huge number of open communities with partially overlapping engagement such as Open Source MANO (OSM), ONAP, OpenStack, ETSI NFV, and TMF. A consolidation of the very diverse landscape of open communities is certainly on the cards for 2021.

Open networking is transformational to our industry. We at ADVA are well prepared with clearly defined key expertise in disaggregated optical transport, packet networking including NFV and NOS as well as synchronization. Each technology has its place in open networks. Our engagement in open bodies assures that solution integration can happen in an easy and seamless way.