It’s been over 50 years since John F. Kennedy famously declared “we choose to go to the moon.” It’s interesting to try and imagine how the last 50+ years would have unfolded if instead he stated “let’s dabble in space, see if we can develop a business case, and see where it leads us”. As a society, would we have the innovations and prosperity that we enjoy today?
A question that we must challenge ourselves to ask is: Where are the NFV moon shots?
In June we participated at both TMFLive and LightReadings Big Telecom Event where we announced our new network security solution. At both events there was a lot of discussion about NFV and SDN are key to enable new revenues, and not about cost savings alone. However, from many aspects, these aspirations were not supported with a lot of details. In fact, at BTE there was a lot of discussion about incremental evolution and making small deliberate investments to prove-in business cases in order to executives and boards of directors. That is, a slow incremental approach.
Virtual CPE was cited most commonly as one of the business drivers. But is this really a new business case, or about delivering existing services cheaper, or about re-capturing business that is being lost today, or all of the above? We are the first the acknowledge that transformation isn’t going to be easy. Service provider networks are inherently complex, they have evolved over generations, incorporate a range of technologies and involve an interwoven fabric of people, systems, and processes. It is this complexity that partially explains why new revenue creating services may be delayed until some of the operational realities are resolved (this is something about we’ve previously written: Until MANO is Solved, Creating New Revenues from NFV will be Delayed).
During BTE we also participated in a panel discussion entitled “Improving Customer Experience: The Critical Role of Service Management in a Virtualized World” moderated by Michael Howard, Senior Research Director at Infonetics/IHS Technology with colleagues from Procera, Intel and Colt.
Michael shared his view of the Carrier Network Architecture 2020. While a simplified representation, it still illustrates the striking complexity that service providers face. For instance, the combination of highly distributed, multi-technology networks with multiple domain-specific orchestration and control points explicitly highlights some of challenge service providers will face: multiple systems and people each “touching” network, and more importantly affecting the services delivered.
Unlike today’s networks, virtualization blurs the natural delineations that existed when functions were implemented in physical devices, and instead creates more inter-dependencies from a performance and security perspective. On the panel we spoke that a service-oriented view towards management and operations must be adopted. And it is important to recognize that service chains span physical and virtual networks, all of which must be correctly tuned in order to deliver the best possible customer experience. As an industry, these are some of the issues which we must figure out, “not because they are easy, but because they are hard”.
Unquestionably, there has been considerable, and accelerating, progress made over the last few years. But where are the moon shots? The closest is perhaps AT&T’s stated ambition to virtualize 75% of their network functions by 2020.
Do they have all the details in place? Do they understand all the risks? Do they understand all the drivers? Probably not. But they do understand they urgency in that something must be done, because if they wait to understand all the aspects of commercialization first, that perhaps we will never get there. After all, others are working on their own moon shots.