Operators will have options for Distributed Access Architectures

March 26, 2018
The move to Distributed Access Architectures (DAA) promises a variety of benefits. And what operator wouldn’t be interested in reduced power, cooling, and space requirements in their headend and hubs, better transmission performance (an increase in MER of 6 to 9 dB, say some industry sources), and a simplified network architecture? The question operators face is how to implement DAA. Based on a look at the DAA technology landscape, it appears cable MSOs will have plenty of options within the next few years that will enable them to customize their strategies to the requirements of individual serving areas.

The move to Distributed Access Architectures (DAA) promises a variety of benefits. And what operator wouldn’t be interested in reduced power, cooling, and space requirements in their headend and hubs, better transmission performance (an increase in MER of 6 to 9 dB, say some industry sources), and a simplified network architecture? The question operators face is how to implement DAA. Based on a look at the DAA technology landscape, it appears cable MSOs will have plenty of options within the next few years that will enable them to customize their strategies to the requirements of individual serving areas.

The benefits are remote

DAA, as the name implies, sees functions that normally reside in the headend or hub distributed into the network, closer to the user, in an intelligent node. Removing functions into the network reduces the amount of hardware the headend or hub must house, thus leading to the power, space, and cooling requirement reductions. The remaining headend or hub equipment communicates with the intelligent node via a 10 Gigabit Ethernet fiber link, using previously analog transmission spectrum converted to digital. The switch to digital transmission enables the increase in the transmission performance over the coax connection to the customer.

The question is what function or functions the operator would like to migrate to the remote node. CableLabs has begun work on a pair of DAA approaches within its Distributed CCAP Architecture initiative (a subset of its overall DAA activities). Remote PHY is the first architecture it took up, with specification work nearly complete. Here the PHY functions such as the modulation of DOCSIS and video signals move from the CCAP platform into a Remote PHY Device (RPD) in the intelligent node. The CCAP-related functions that remain in place reside in a CCAP-core, which performs Ethernet conversion, tunneling and routing for DOCSIS frame transmission as well as DOCSIS MAC functions. The CCAP-core can be paired with off-the-shelf routers and switches in a significantly simplified architecture.

With Remote PHY specifications nearly completed, systems vendors have begun to unveil Remote PHY hardware. CableLabs has conducted 11 interoperability events that have helped operators evaluate the technology’s readiness for deployment. ARRIS CTO – Network Solutions Tom Cloonan, during an SCTE∙ISBE LiveLearning Webinars™ for Professionals event in March 2018, predicted that deployments of specification-compliant RPDs should begin this year.

Meanwhile, this past December, CableLabs turned its attention to the second major DAA approach, Remote MACPHY. This strategy sees the DOCSIS MAC function also moved into the intelligent node. Removing the MAC function from the headend or hub also removes the need for a CCAP-core, leaving only routing, switching, and controller functions – with the controller functions cloud-ready.

Remote MACPHY should dovetail well with the evolution toward cloud-based network approaches.

CableLabs plans to leverage as much of its Remote PHY work as possible, reported Jon Schnoor, lead engineer at the organization, during the same webcast. The specification makers plan to create a Remote MAC Device (RMD) that would incorporate the RPD. They also envision a Remote MAC Core (RMC) that could remain structurally separated from the RPD; this would remove the MAC function from the headend/hub but still enable the MAC function to be shared among multiple RPDs. CableLabs also plans to create a virtual MAC core for SDN-enabled environments. For headend and hub requirements, specifications will cover a MAC Manager for the network’s management plane, an SDN Controller for Layer 2 and 3 control plane needs and a DOCSIS Controller for the DOCSIS control plane.

While Remote MACPHY promises the most streamlined approach to serving user needs, the necessary hardware and software isn’t likely to be ready for some time. Schnoor described Remote MACPHY specification development as "a lot of work," while Cloonan predicted RMDs won’t see deployment until around 2020, and RMCs with subtending RPDs until 2022.

Thus, operators can choose to deploy Remote PHY in the near future if their DAA requirements are acute, wait a bit to take the full plunge into Remote MACPHY or anticipate that they can start with Remote PHY and evolve to Remote MACPHY down the road.

Other considerations

However, this choice won’t be made in a vacuum. For example, while operators are using a 10 Gigabit Ethernet fiber link between the headend/hub and intelligent node, they may want to use that fiber cable to support other Ethernet-enabled services. Business services as well as 4G/5G centralized radio access network (C-RAN) and small cell fronthaul/backhaul traffic also could use the same fiber.

Optical transport systems vendors see Distributed Access Architectures as a new market opportunity.

Such converged network potential has attracted the attention of optical transport systems vendors, who believe they could have a role to play in supporting DAA alongside other forms of Ethernet traffic. Platforms that combine packet switching with optical transport would fit well within a converged network that enables DAA, they are telling cable operators.

Meanwhile, fiber to the premises (FTTP) equipment suppliers continue to push greater use of all-fiber architectures. The connection from the intelligent node to the customer doesn’t have to take place over coax, meaning there could be a demand for remote optical line terminals (R-OLTs) within the DAA framework. The FTTP platform community has already begun to tout such arrangements, pointing out that their SDN-friendly offerings would pave the way to either Remote PHY or Remote MACPHY implementations.

And, of course, there’s nothing that prevents DAA networks from supporting Full Duplex DOCSIS or, if and when it comes to pass, extended spectrum DOCSIS technology. In short, operators likely will examine DAA from a variety of angles. Asked during the webcast which of the many options would prove most popular among operators, Cloonan related that he had spoken to three operators shortly before the webcast, and all three were planning different DAA pathways, each for valid reasons.

It seems that operators will not only have several options from which to choose, but that each approach will have its adherents. And there’s nothing that says an operator must choose a single approach. The evolution of DAA deployments promises to be multifaceted, perhaps within a given operator but certainly across the community at large.

Stephen Hardy is editorial director of Broadband Technology Report.

Learn the Basics of Distributed Access Architectures

You can now view on-demand the March 2018 SCTE∙ISBE LiveLearning Webinars™ for Professionals webcast, produced by BTR, "Distributed Access Architectures." This webinar defines the various Distributed Access Architecture options, describes the work now underway on such architectures within CableLabs, offers predictions for technology evolution, and covers the ramifications for network testing. Register for the webinar and view it now!