SCTE CABLE-TEC EXPO – To hear fiber players tell it, cable is facing a hopeless disadvantage in the years to come. But experts from CommScope, Vecima and CableLabs contended there’s still plenty of juice left to squeeze out of DOCSIS. In fact, they said, cable’s ability to roll out 10-gig speeds – downstream, at least – doesn’t rely on the rollout of DOCSIS 4.0 but boils down to the need for new modems to fully capitalize on DOCSIS 3.1’s capabilities.
DOCSIS 3.1 offers operators two primary upgrade paths on the road to 4.0: mid splits and high splits, which each allocate additional upstream spectrum. Speaking during a conference session, CommScope VP of Product and Strategy Craig Coogan noted mid-splits offer up to a 561% increase in upstream throughput capacity and a 45% increase in downstream capacity, while high splits offer a 1397% improvement in the upstream and a 28% increase in the downstream.
Coogan pointed to high splits as the “sweet spot for operators to move to today” to better compete with 10G PON offers.
Based on CommScope’s projections, high split configurations “can buy operators up to 10 years of service with 150 subs [within a service group] with 5-gig down and 1-gig up speeds,” he said.
In addition to moving to high splits, Vecima CTO Colin Howlett said during the session there are a few other tricks operators can implement to further “turbocharge” DOCSIS 3.1. These include leveraging Profile Management Application (PMA), low latency DOCSIS tools and modem swaps.
Curtis Knittle, VP of wired technologies at CableLabs, told Fierce that in OFDM networks, PMA allows operators to improve channel modulation by catering to different profiles. Previously, in single carrier QAM networks, modulation had to be tuned to the lowest common denominator of performance. But now, outliers can be modulated using different profiles, allowing better performance for the remainder of subscribers. While downstream PMA is more commonly used today than upstream PMA, Knittle said the latter is equally important since it can help maximize upstream performance.
Back in the session, Howlett said low latency DOCSIS tools such as non-queue building traffic (which aims to get traffic through the network quickly rather than efficiently) and proactive grant services (PGS) for bandwidth can help address another pressure point from fiber. Using PGS, operators can deliver sub-1 millisecond latency for 99th percentile traffic, he said.
Howlett also asserted that operators can actually deliver 10-gig services downstream without moving the entire network to DOCSIS 4.0 if they pair next-gen modems with DOCSIS 3.1 plant.
In a nutshell, the performance of DOCSIS 3.1 is currently limited by DOCSIS 3.1 modems. While the spec supports the use of five OFDM spectrum channels, today’s modems were only built for two. If operators can get their hands on next-generation modems which support all five channels, Howlett said, then they can deliver speeds of 8 to 10 Gbps downstream on DOCSIS 3.1 plant. This can be done on a case-by-case basis to avoid the costs of a widespread upgrade while enabling faster service tiers for certain subscribers.
“I think turbocharging DOCSIS 3.1 is really a first step for us to get to 4.0,” Howlett concluded. “Even high split covers us into 2030, so aside from the marketing wars we’re going to see around multi-gigabit upstream service, we are going to have a long life in DOCSIS 4.0 and a long life in that HFC.”
This story has been updated with Craig Coogan's correct title.