The announcement that Telefonica has turned to Huawei for its virtual EPC is certainly of major significance. Huawei is the only major telecom vendor who’s managed to gain significantly in revenue over the last several years, largely due to its price-leader status. Competitors like Cisco, Ericsson, and Nokia have hoped to wrestle away some market share through product innovation, and aspiring NFV vendors like HPE saw virtualization as an entry point into the networking space. Now, Huawei seems to be taking a product innovation lead too. Is there no stopping them?
Not if you don’t address the reason that they got started. I had a memorable strategy session with a major network vendor about eight years ago, and during the meeting I was asked to list out their competitors and their strengths and weaknesses. When I did, one of the strategists called me out triumphantly—“You forgot the most important competitor—Huawei!” My response was “Huawei is not a competitor, they’re a winner if you decide not to compete.” That was a more critical point than I think that vendor recognized. Eight years ago, there was plenty of time to take an aggressive position on feature-driven network evolution. Absent such a position, you have price-driven evolution, and that’s what happened.
But it’s also important to recognize that Huawei has been active in the feature side of the game, almost from the first. Perhaps their most important contribution is something most have never heard of—the “Next-Generation Service Overlay Network” or NGSON. This was something pursued about a decade ago, through the IEEE rather than through one of the more prominent network standards bodies. It presaged the MEF’s Third Network, SD-WAN, and many of the concepts now emerging in 5G. It’s also, in my view at least, an element in the formulation of the CloudEPC solution that Telefonica bought into.
NGSON’s failing isn’t technology, it’s publicity. The IEEE, of course, isn’t as vocal and effective a body for promoting new technology as a nice fresh forum like the MEF, ONF, etc. Huawei, no slacker in sales or technology, is surely weak in marketing and positioning, and when you combine the two it’s no wonder that NGSON didn’t shake the earth.
Service overlay network technology is the foundation of the Nuage SDN strategy, bought by Alcatel-Lucent who then merged with Nokia, and it’s the thing I’ve already called the primary asset Nokia has in the SDN/NFV space. However, it’s obvious that Nuage was a bit of a Cinderella in ALU, and it seems to be the same in Nokia. Certainly neither parent has done a lot in positioning Nuage as a strategic asset, so Huawei’s failure to promote NGSON hasn’t hurt as much as it might have.
Which, I think, is the key point here. Certainly Huawei or any other telco equipment vendor doesn’t need the media to engage with customers, but the lack of aggressive positioning of key feature assets, for any vendor, lets competitors dodge embarrassing questions from the media. Those questions might impact sales success were they to be asked. Could that happen even with the Telefonica CloudEPC success?
Virtual overlays could be a key to making things like EPC work, in no small part because the whole notion of EPC (which relies on tunnel networks) is a form of an overlay network. It’s also very likely that virtual overlays would make sense in a multi-tenant network-slicing solution, because overlay networks (Nicira, now NSX) was the first big step in cloud multi-tenancy and it’s an overlay solution.
One critical result of all the positioning mismanagement going on is that there’s no clear vision of just what a next-generation EPC would look like. Most vendors call their next-gen EPC an NFV element, but there’s a lot more in common with traditional cloud computing than with NFV, since the elements of EPC are not single-tenant and that’s been the focus of NFV work so far. And nobody is really dealing with the fact that EPC is about accommodating mobility, and the EPC mobility accommodation is supposed to be obsoleted by 5G.
If a key goal of 5G is to eliminate EPC altogether, then the smart money would bet on a mobile core architecture that did both 5G and EPC seamlessly. Something like that would also introduce the notion of a “selective multi-tenant” approach to NFV, and it could exercise the flexibility in packet forwarding of OpenFlow and the tenancy management capabilities inherent in overlay networks. Why isn’t that the right answer? Could it be that vendors have just not bothered to position their stuff, or that they don’t really have a solution.
We can’t tell from Huawei. They have documentation on CloudEPC on its website, but there’s little there beyond a vague diagram. There is customer product data available, but that’s not generally available to the media/analyst community and it’s not really a tutorial on the conceptual framework of CloudEPC. We can’t tell from Ericsson or Nokia either. Might a competitor with little more in the way of a product than Huawei has be able to jump off with a strong campaign and take thought leadership? Yes, but they’ve not done that so far.
Huawei’s win with Telefonica could be huge for both companies; as far as I can tell it would be the largest-scale “NFV” strategy yet deployed and it certainly hurts all the other vendors who have been hoping to win big with Telefonica as the leading EU operator in the NFV world. That’s especially true for HPE, who was the first vendor to appear to have an inside track with Telefonica, and Ericsson who seemed to have the recent lead. Is it really seminal for “NFV” or “EPC” or “5G?” I don’t think so, and it’s not just because there’s no substance to the positioning of the vendors.
5G looms over everything, and yet we really don’t know what it will do or how it will do it. We’re only a couple years from the supposed deployments, and the specifications are still swimming in uncertainty. The biggest question for Huawei and Telefonica is how far their deployment will get without 5G capability, or perhaps how Huawei could introduce 5G evolution into their offering when there aren’t firm standards to work from.
In any event, mobile evolution is the last gasp of opportunity for Huawei’s telco equipment competitors. As I’ve pointed out in the past, there are no other service transformations that could credibly position a critical mass of NFV assets in the near term. Even IoT, which has great NFV potential, is likely to be subsumed into the 5G transformation. Given all of that, Huawei needs to leverage its mobile NFV position to the utmost, and its competitors need to knock them down before it’s too late.