The move to distributed access architectures is a transition that many in the cable industry have called a once-in-a-lifetime network transformation event. And if last fall's SCTE Cable Tec-Expo was the coming-out party for new DAA models, then March 2017 may be considered the commercial launching point when DAA products start to roll out in earnest.
At Light Reading's Cable Next-Gen Technologies & Strategies conference in Denver last week, Cisco Systems Inc. (Nasdaq: CSCO) Fellow and CTO of Cable Access John Chapman revealed that as of this month, his company is shipping its new Remote PHY device in volume. The product literally moves the physical layer of a Cable Modem Termination System (CMTS) or Converged Cable Access Platform (CCAP) down to the network node. The goal, as with any DAA solution, is to alleviate space and power constraints at the headend or hub site, and to push the IP-to-RF conversion point deeper into the network.
The evidence for DAA's momentum extends far beyond Cisco product shipments too. According to a survey of 35 cable operators conducted by SNL Kagan, 17% say they plan to begin deploying distributed access architectures this year, with another 34% committing to deployment plans in 2018. Yet another 26% say they will move forward with DAA in 2019 or later, and only 20% say they have no plans to implement DAA at all.
What does that commercialization of DAA mean? Unfortunately, the answer is far from simple. There's still a debate in the industry about whether cable companies should remote only the physical layer of a CMTS/CCAP, or whether it makes more sense to shift both the MAC and PHY layers down to the node. This debate has raged for some time, but there doesn't appear to be any definitive answer in the offing. (See Cable's Great Debate: How to Split Functions.)
Jeff Heynen, director of sectors and technology at SNL Kagan , for example, says he sees a slight preference for Remote PHY today, but that's likely because there's less standardization in Remote MAC/PHY technology at the moment. That will change. (See Cable's Upgrade Moment – Part IV.)
Comcast Corp. (Nasdaq: CMCSA, CMCSK) Vice President of Network Architecture Robert Howald echoes Heynen's statement. "A set of vendors have lined up behind a standard that is Remote PHY that makes it easier to grab on to, and these solutions that are Remote CCAP or Remote MAC/PHY are not as standardized so I think that makes it easier to gravitate if you're unsure to a multivendor solution in the beginning," Howald said in a fireside chat at the conference. He added, though, "I think those things get worked out in time ... [And] we have looked at both. We are looking at both right now."
Meanwhile, DAA technology -- whatever version operators choose to deploy -- is only one part of a much bigger evolutionary shift in the cable access network. Distributed access architectures have to be viewed in the context of other developments, including the move to drive fiber deeper into cable networks and the emergence of Full Duplex DOCSIS.
As Chapman explains, Fiber Deep is an enabler for cable operators looking to extend the capacity and lifespan of their HFC networks. But by pushing fiber deeper in their networks, cable companies also saddle themselves with a huge number of new nodes, which in turn create a need for more ports and CMTS/CCAP devices -- something operators can't afford.
"Fundamentally the deep fiber market broke the CMTS market," says Chapman. "They're just completely incompatible. What the distributed architectures do is give you ten times the number of ports on the same CMTS. It saves the economics. It's that simple."
"You can't really do deep fiber without a distributed architecture," Chapman adds.
Full Duplex DOCSIS (FDX) also benefits from a distributed architecture. As Howald puts it, DAA isn't a requirement for deploying FDX –- a technology that will significantly increase cable's upstream bandwidth capacity. But it does simplify the process. (See How Cable Plans Symmetrical Gigabit via FDX.)
"Full Duplex could work without DAA," notes Howald, "but you'd have to essentially invent new upstream to do that ... [It] doesn't seem like the best investment right now."
The various dependencies in the access network make it clear why cable operators are moving quickly toward distributed architectures. Howald confirms that Comcast has multiple DAA trials underway now, and says the company is working with all vendors to determine the most effective DAA deployment models.
Cisco's Chapman characterizes the shift in the access network a bit differently. With Cisco's preference for a Remote PHY approach, Chapman calls R-PHY not just another feature for cable networks, but a tipping point.
SNL Kagan's Heynen agrees that DAA is at a tipping point. "By this point [next year], we'll already have 35% of those operators that will be in the midst of [DAA] deployment, possibly more," says Heynen. "I 100% agree. The technologies and the benefits that they enable, they're just too good to pass up. So many operators are already down the road of pushing fiber deeper that [DAA is] just the logical next step."
— Mari Silbey, Senior Editor, Cable/Video, Light Reading