By: Prayson Pate

The saying, “It’s tough to make predictions, especially about the future,” has variously been attributed to Albert Einstein, Yogi Berra, Niels Bohr, Mark Twain and others. We may not know who first said it, but it is obviously true. Even so, I will stick my neck out and make a prediction regarding SDN and NFV: The biggest benefits will be a surprise to all of us. Furthermore, there are concrete steps we can take to ensure that surprising future.

Planning for history to repeat itself

The great innovation of the smart phone was that it enabled easy development of innovative third party apps, and that these apps were delivered from the cloud under the control of the user. The initial apps were great, but the coolest and most valuable apps were completely unforeseen, like flashlights and airline e-tickets. I learned about another cool app when I visited my son who is a freshman in college. I was picking him up for lunch, and he asked if he could drop off some laundry on the way out of the dorm.  I said “sure”, so he whipped out his smart phone and used an app to query the washing machines to see if one was available. I wish I had that app when I was in college!  The point is, nobody predicted these specific apps when they were building the first smart phones. However, there was an expectation that the availability of a standard development environment and an ecosystem of tools would lead to something really cool. 

Today, most people expect that the benefits of Software-Defined Networks (SDN) and Network Functions Virtualization (NFV) will include lower cost equipment, tighter integration into control systems, and the creation of Software-defined Services. These benefits are true and valuable, and they will drive the business case for early adoption. The really interesting innovation will occur where no one expects and will open a path to services that no one foresees. The real question is: how do we ensure the success of this expected but unknown innovation? The three apps mentioned above (airline tickets, flashlight and laundry) each illustrate one of the aspects of innovation that we need to support in order to achieve the maximum in innovation.

System integration

The move from paper tickets to e-tickets in the mid-90s was well under way when smart phones appeared. However, the ability to load the e-ticket on the smart phone and to also access other customer support features is what is opening the door to a truly paperless system. Reaching this point required a number of related changes in disparate systems:

• The ability of the airline to sell and read e-tickets.
• The ability of the customer to execute purchases and changes on a self-serve basis.
• The ability of a software app on a smart phone to interact with the airline’s systems.

Each of these changes required the creation of systems with open and documented interfaces, along with significant changes in how the airlines conducted their ticket sales. Similar changes will be required from Communications Service Providers (CSPs) and the equipment and software suppliers who partner with them.

For today’s CSPs, the Metro Edge is where the action is. Why? Because the Metro Edge is where the services start, and it acts as a bridge from the customer to the network and especially to the cloud.  Another reason is that the Metro Edge is a complicated place with heterogeneous equipment. Finally, you can’t virtualize pipes, so the problem is not going away.

What types of changes will be needed? First, the preponderance of Ethernet in the customer site and in the data center, along with the advances in Carrier Ethernet means that a ubiquitous deployment of Carrier Ethernet is needed to underpin services. Next, the principles of SDN and NFV need to be utilized to provide the ability to control and extend network functionality at the network edge. Finally, as with the airline e-ticket, these new systems must be effectively tied into higher-level OSS/BSS systems to facilitate creation, activation and assurance of services.