Last week, AT&T announced that they had selected Ericsson as the sole provider of their Radio Access Network (RAN) equipment. This was a highly unusual announcement in many ways, chief among them was the announcement that AT&T would replace all of the existing Nokia gear in their network and only use Ericsson. Since then, the industry has been abuzz with speculation as to what caused this dramatic move.
For background, the RAN is the part of mobile operators’ network that is the first point of contact between with a user’s phone, the proverbial ‘last mile’ of a wireless connection. This is critical stuff. For the operators, the RAN determines the overall profitability of their network by sharing spectrum among all their customers. The importance of this layer merits having two vendors, just in case, and it is rare for an operator to willingly rely on a sole provider. For the equipment makers, the RAN is the last bastion of their ability to differentiate. The RAN feeds into the operators’ core networks which increasingly look like all the other networks out there, and is a realm where Ericsson and Nokia have proven they are decreasingly competitive. And for what it’s worth Nokia acquired Lucent and Bell Labs a decade ago, and since those entities were once part of AT&T, it is theoretically possible that there is “Nokia” gear deep down in the network.
When asked why they took such an unorthodox move, AT&T has given answers that are less than clear. One executive cited the excessive number of fans required for Nokia gear. Another mentioned something about Intel chips. Neither answer is anywhere close to sufficient. The plot thickened further when Light Reading reported that Ericsson is footing the bill for replacing all that relatively new Nokia gear.
We can think of two plausible reasons for the move. The first is purely economics. Many people believe that Ericsson paid a heavy price to win its position, as evidenced by the cost of ripping out the Nokia gear. Theoretically, Ericsson felt that AT&T was important enough as a long-term customer that it was worth paying a lot up front and somehow make up the cost down the road. The problem with this theory is that it is hard to imagine a scenario in which the math for that pencils out. This is a highly competitive market, and AT&T has been negotiating these contracts for over 100 years. So the operator was unlikely to sign a contract which locked them into paying huge margins to Ericsson years down the road. It is possible, but seems like the kind of contract that CFOs eventually get fired over.
That being said, there is a clue in the companies’ deal press release. This makes repeated mentions of Ericsson’s new factory in Texas. So our guess is that US taxpayers may be footing part of this deal somehow. The US government has made no secret of its goal of onshoring manufacturing, so there are some forms of subsidies involved with that plant, and maybe they are sufficient enough to make this deal work out for Ericsson.
Another possible reason for the deal was AT&T’s stated goal of opening up their RAN. The goal here is to decouple the hardware and the software that run the RAN. Of course, this is the heart of Ericsson’s business, both they and Nokia, have profited over the years by keeping those two things tightly bound to each other. Neither of the vendors wants to see the two split and opened up to more competition. Nonetheless, this is stated as an explicit goal by AT&T in the press release. So maybe the appeal of moving to all-Ericsson was that vendor was more willing to open up their network protocols in some way than Nokia was.
Which brings us to Open RAN (O-RAN), this is an industry-wide initiative to standardize this decoupling. We have written a lot about O-RAN over the years, most recently concluding that growing friction within the group likely means it will not be able to fulfill its objective. And to be clear, a lot of that friction is coming from Ericsson.
In the end, if AT&T’s goal is to open up its network, moving to a single vendor is a very roundabout way to get there. Instead, what we could be seeing is Ericsson making a bet that by agreeing to AT&T’s ‘openness’ goals it could still find ways to profit and at the same time take the reins in defining what “open” really means in the RAN. This is some complex strategy, but Ericsson is perfectly capable of that level of 4D Chess. And if that is true, then it likely spells the end of O-RAN. If one of the world’s largest mobile operators is willing to settle for Ericsson’s version of an open RAN, rather than O-RAN, smaller operators are likely to take note. Building an O-RAN network is proving challenging, so maybe it is easier to let Ericsson build an open-ish RAN instead.
Leave a Reply to SGCancel reply