Your browser is not supported. Please update it.

27 May 2021

Operators leaning towards FDD over FDX for DOCSIS 4.0

A studious survey from media access platform provider ATX Networks has uncovered how more than 40 operators of various sizes perceive DOCSIS 4.0 today and how they plan to deploy it, while exploring whether flirting with fiber is mission critical.

We love it when a vendor catches us off guard with a ballsy opening gambit. “The HFC network is destined for extinction,” was how ATX Networks kicked off the show during a webinar this week revealing results of its survey. “Fiber is destined to someday completely overtake HFC’s portion of the network,” continued the condemnation of HFC networks, in self-defamatory fashion from ATX Networks given its cableco customer base.

But unlike other suppliers of access equipment in the fixed network scene, the Canadian vendor has embraced the HFC endgame being more apparent than ever and is set on understanding the needs of operators not just of tomorrow, but of the next 20 years. This is reflected in results showing that the majority of operators surveyed have the majority of subscribers on HFC networks. However, these same operator decisionmakers expect falling number of subscribers being served by the HFC network over the next 10 years, and more still by 2040.

With this shift away from HFC, most operators (37%) expect to begin migrating to DOCSIS 4.0 in the next 3 to 4 years, while 26% plan to migrate in 5 to 9 years. A further 17% plan to be more aggressive with DOCSIS 4.0 deployments planned for 1 to 2 years from now, while only 6% answered “never” and a further 9% foresee a long-term roadmap of a decade or more. Interestingly, 5% of respondents still haven’t migrated to DOCSIS 3.1, which isn’t all that surprising given the standard is only eight years old and has been updated many times since the original.

DOCSIS 4.0 brings wider channels and a broader frequency range, as well as extended spectrum technology to pump downstream spectrum to 1.8GHz and upstream to 684MHz – creating a large spectrum for data to play with and ultimate theoretical speeds of 10Gbps up and 6Gbps down.

Some operators are more eager to harness these capabilities than others, although an overwhelming 82% of respondents taking part in the ATX Networks survey agreed to some degree that their company will begin adopting DOCSIS 4.0 as soon as supporting technology becomes available. Appetites are clearly strong for DOCSIS 4.0, even though DOCSIS 3.1 seems not that far along.

Decisionmakers agree even more overwhelmingly on the importance of any equipment used to replace components in the outside plant today being capable of supporting DOCSIS 4.0 when it becomes available, with 92% agreeing it is important. This is a slightly leaning question we feel, as few decisionmakers will dispute the importance of DOCSIS 4.0 even if they have yet to etch out anything resembling a finished roadmap for the technology yet.

This equipment really comes down to three key elements that enable the HFC network – the node amplifier, the taps and passives, and the power inserters. From a node and amplifier perspective, panelists advised that operators want to be looking at modular upgradable types of platforms. Cablecos also want to ensure these platforms can support the necessary bandwidth and thermal dissipation requirements that the next version of DOCSIS will support. ATX Networks is seeing a trend of MSOs looking for platforms that pass not only 1.8GHz but potentially up to 3GHz in anticipation of the next level of DOCSIS.

However, 53% of those surveyed are still unsure which version of DOCSIS 4.0 they are more likely to adopt – FDX (Full Duplex DOCSIS) or ESD (Extended Spectrum DOCSIS). FDX technology refers to the ability of network devices to simultaneously transmit both upstream and downstream in the same spectrum, specifically in the spectral region between 108MHz and 684MHz, while downstream-only spectrum extends from 684MHz up to 1.2GHz. ESD technology refers to the traditional frequency division duplex (FDD) mode of operation but extending the upstream-only spectrum up to 684MHz while also extending the downstream-only spectrum up to 1.8GHz.

Almost a quarter are leaning to FDD, which offers cable operators more of a traditional approach to allocating spectrum, while FDX addresses spectrum expansion by enabling traffic to flow upstream and downstream in a shared frequency range, with the latter selected by only 13%.

This question highlights a lot of uncertainties around unknown complexities in the two DOCSIS flavors, raising further questions such as which products will be easiest to migrate into and the flexibility of these technologies? Companies must develop products that are operationally friendly and can slide seamlessly into the network to avoid disruption, which is why FDD is currently preferred given its familiarity among operators in over 30 years of plant evolution. Most of the FDX uncertainty is driven towards costs, while panelists suggested that more flexibility in the FDX architecture could swing answers to this same question on the next iteration of this survey.

Looking at gigabit speeds, 15% believe the ability to offer subscribers 1Gbps symmetrical services is extremely important today, while 37% said important and 33% picked somewhat important.

But what about the long-term competitive threat from FTTH providers? ATX Networks found the greatest degree of concern from the largest operators surveyed (those with over 1 million subscribers), with 38% describing FTTH providers as a significant threat to business.

Interestingly, most operators believe that 10Gbps is the ceiling for their networks, in response to the 10G initiative from CableLabs. Panelists were a little surprised at the lack of ambition, pointing to how a few years ago people thought 750Mbps was the max, and a few years later everyone was saying cable would top out at 1Gbps, yet the industry continues to innovate and overachieve.

For some, getting to these speeds will involve migrating the HFC plant to all fiber. Of all the possible roads to fiber, 62% selected the natural evolution option, which involves node area segmentation that pushes fiber deeper into the network. Deploying fiber in greenfield/new builds was chosen by 55%, while 26% prefer proactive replacement of coax with fiber, and 22% plan to augment HFC with targeted PON overlay to top users.

We’re talking about some substantial timeframes here, mind you, with 32% of operator decisionmakers answering 21 to 30 years to the question what is the most aggressive timeframe for completely replacing the HFC network with all fiber? Another 24% said between 11 to 20 years, while 29% said beyond 30 years.