BCE 2017: The Forecast is Still Cloudy
The technical content of this year’s BCE was, as usual, presented by some of the best and brightest minds in telecom. Unfortunately, most of those minds were in agreement that the landscape for SDN and NFV, in particular, has not gotten any clearer in the last year. In fact, in many ways the clouds have gotten denser.
The show was kicked off with a keynote by Steve Saunders (founder of LightReading) who made a splash with his arguments against open source. His message was that the telecom industry has lost its way by relying on open source rather than sticking to the principles of standards and common agreements. With open source, he argued, there are too many “good ideas” and too little results. He blamed that mindset for the mess that currently exists with SDN and NFV where there are too many people doing too many different things and there is no real, working infrastructure to build upon. If nothing else, Steve got people at the conference talking.
Others at the show disagreed with Steve’s point of view about open source. AT&T’s John Donovan in particular said that he (Donovan) expected as much as 50% of all of AT&T’s software to be open source in the next years. However, Steve was correct in stating that there is little consensus in SDN and NFV and that the problem has gotten worse over the last year, not better.
Speaker after speaker mentioned the large number of “standards” bodies working in the space and the overlapping responsibilities and holes in coverage. When a group of executives were asked what one thing they would change if given the opportunity, the answer that resonated with the group was, “a map of who is responsible for what” in the standardization process. No such map exists, and it’s causing confusion and slowing down innovation.
Why is it worse this year than last year? Probably because some people (AT&T, Verizon, Facebook, etc.) have given up on waiting for complete vendor solutions and have started filling in the gaps themselves. AT&T released ECOMP, and it’s still not clear who else will take up that ball and run with it. Verizon has their own initiatives. Various groups with “open” in the name have formed. Everyone has their own version of YANG models that must be supported. AT&T wants solutions for disaggregation while Verizon wants converged solutions. Several attendees remarked that if there are enough “open” solutions, they start to look closed again since no one company can possibly support all of them. It’s good to see companies moving forward, but this movement is resulting in confusion and not clarity.
The good news from BCE this year is that the carriers have fully embraced the idea of SDN and NFV. They are approaching it in different ways, but all agree that we will eventually get to a point where the network is more responsive, application oriented, and (perhaps) profitable. Traffic is still rising, and multiple carriers talked of becoming software-centric rather than hardware-centric companies to allow them to survive the onslaught and still provide the required services to their customers. Perhaps the most hopeful comments came on the carrier panel, when they agreed that the eventual winners in the telecom space will definitely be those who own the infrastructure. None of the fancy software works without fiber, copper, and wireless infrastructure. We’ll be using those resource more efficiently, but folks like ECI who build the endpoints will continue to be crucial to the network.