AT&T remains intently focused on its network, which over the decade has dramatically evolved using a software-defined architecture that the operator can now basically operate as an application. However, ongoing updates to that network will be tied to monetization and spectrum needs.

Chris Sambar, EVP for technology at AT&T, during a keynote speech at the recent Brooklyn 6G Summit noted that the carrier’s decade-long “Domain 2.0” program has resulted in a powerful platform.

“I would actually argue that the next killer app is the network itself,” Sambar said, touting the carrier’s push to disaggregate its network resources and drive monetization of network investments.

“To date in the wireless industry, and surely in the telecom industry, the connectivity networks have really been closed boxes where partners would provide to us a system that worked great. It was all integrated hardware, software and applications in that system and if you needed help with it you generally had to call the partner: ‘hey, Cisco; hey, Nokia, Ericsson, can you help me with this?’ We're really working on breaking that up into open disaggregated networks.”

AT&T hit a long-standing goal in 2020 to virtualize 75% of its network functions with SDN, and more recently hit another disaggregated goal for its core network backbone.

“Now, 65% of our backbone traffic in the U.S. and actually around the world is now sitting on open disaggregated platforms,” Sambar said. “We're now moving that out into the metros. This is primarily on our fiber networks, on our wireless networks. We're moving that out into the metros now with our routing infrastructure in the metros, moving from closed boxes to open disaggregated.”

This has positioned AT&T’s network as the “killer app” that Sambar said can now be accessed via APIs that can be the basis for innovative new applications. He specifically cited the ability for customers to now be able to turn up security firewalls running on its network in an instant, which can help drive much-needed monetization efforts.

Telecom networks are expensive

That monetization is a growing area of focus for telecommunication operators that have spent small fortunes to deploy their 5G networks.

Prefacing his comments by stating that he did not “want to come across as negative or cynical,” Sambar laid out AT&T’s investment hurdle. As an example, he noted that AT&T spent $40 billion over the past two years on new mid-band spectrum. (You may need to reload your coffee for another spit-take as Verizon spent closer to $50 billion.)

That’s just the beginning.

“If you buy a band of spectrum at auction and you deploy it in the network, that is mid-single-digit billions, $6 [billion] to $8 billion for a single spectrum band and putting that on most of your towers nationwide, not even every single tower,” Sambar said. “Because it’s a new spectrum band, especially a big one – 40, 50, 60 megahertz of spectrum – not only do you have to put up a new radio, a new antenna, you're going to have to do cabling work, you're probably going to have to augment the baseband at the bottom of the tower. You're going to have to augment the transport link.”

AT&T is currently updating those transport links from 1 Gb/s speeds to 10 Gb/s, which is an expensive proposition.

“As that goes back into the network and waterfalls into the network, all of those transport links need to be augmented … so now you're up in the $7 [billion], $8 [billon] just for a single band of spectrum,” Sambar said. “And if now I need to go back and retrofit all of the radios on the tower and take them from LTE radios to [5G] radios, you can imagine the cost of doing that and how long it takes.”

That financial burden has seen operators like AT&T and Verizon grow their capex spend to more than $20 billion per year over the past several years to support their 5G network builds. However, many operators that were early in 5G builds to begin this decade are now looking to slash capex and work on monetizing those network investments.

“When you see in the wireless industry that it seems like investment is in a trough right now, that has a lot to do with it. We're getting a little bit worn out with the economics of the industry,” Sambar said. “It's a very healthy and vibrant industry. We're seeing that in the earnings releases from the three mobile operators just recently, but at the same time the capital investments, they have to be logical, we have to have clear line of sight to what the consumer use cases are for those capital investments.”

You need look no further than nascent domestic 5G operator Dish Network to see this delicate financial balance. That carrier is working through similar financial deployment models without the benefit of more than 100 million wireless customers paying around $50 per month for service.

But, we need more licensed spectrum!

While spectrum made up a majority of the costs that go into constructing a wireless telecom network, Sambar was adamant that operators need and want more dedicated, licensed spectrum.

Sambar pointed to the current quasi-licensing model being used to manage the Citizen Broadband Radio Spectrum (CBRS) band, which entails around 150 megahertz of spectrum between the 3.55 GHz and 3.7 GHz bands. The CBRS licenses garnered more than $4.5 billion in total winning bids but have been tricky to deploy due to power limitations and sharing requirements.

Those license requirements have been a boon for new entrants targeting the private network space. Others have said the convoluted model has so far wasted the use of precious mid-band spectrum resources that sit alongside the deeply deployed C-band spectrum ones that are powering current 5G networks.

Count Sambar in the latter group.

“We have thousands of small cells in the AT&T network that are using licensed assist access with CBRS and if I could snap my fingers today and pull all of them out and put mid-band spectrum in place of that CBRS, I would do it because that experiment has not worked out well,” Sambar said.

The executive went on to warn that cable companies looking to use CBRS spectrum to power their mobile 5G ambitions will run into hurdles.

“They're going to find out the same thing we did that it’s not a very good idea,” Sambar said. “I don't care because it's cable, but at the same time I'm a little sad that this is where our nation is, and I think we can do better than this.”

Michael Thelander, president of SRG Research and Consulting Services, in a recent LinkedIn post somewhat countered this stated need for immediate licensed spectrum by stating operators have not done a good job in architecting their networks to take full advantage of the spectrum they already have.

“Operators complain about not having enough spectrum,” Thelander wrote. “While I don’t completely disagree with them, and I recognize one of the biggest inhibitors to the continued growth of FWA [fixed-wireless access] is insufficient capacity, I’d argue that 90-95% of all poor user experiences on today’s cellular networks are caused by poor coverage and not by network loading.”

Regardless of need, the Federal Communications Commission (FCC) is in the process of looking at freeing up additional spectrum in the 3.1-3.45 GHz band that is currently used by the Department of Defense (DoD) to see if it can be used for shared access by commercial entities.

“Any help advocating for more spectrum … and finding more spectrum … especially sub-7 GHz spectrum, would really be helpful,” Sambar said.