Cable operators around the globe are getting serious about upstream spectrum upgrades on their widely deployed hybrid fiber/coax (HFC) networks. But many are likely to start with a "mid-split" upgrade as some of the issues and concerns associated with a bigger "high-splits" continue to get ironed out.

That's the latest reckoning of John Caezza, SVP of product management and technical support for CommScope, which intends to play a major role in these future upgrades. The supplier, which faces off with vendors such as Casa Systems, Harmonic, ATX Networks, Teleste, Vecima Networks, Cisco and Technetix in various facets of the cable access network game, last year introduced a new product line that can support 204MHz high-split upgrades. Then this week it expanded on its thoughts on the direction that cable operators are heading with the upstream.

A mid-split would expand the amount of spectrum dedicated to the upstream in North American DOCSIS networks from the current range of 5MHz-42MHz (5MHz-65MHz on EuroDOCSIS networks), to a new, extended range of 5MHz-85MHz. A high-split would raise the upstream ceiling to 204MHz, and put operators in position to deliver speeds up to 1 Gbit/s in the upstream via HFC.

"Our sense is that … the biggest drive looks like it will come from mid-splits," said Caezza, an industry vet who joined CommScope via its merger with Arris in 2019, and is also late of C-COR, ADC and Philips Broadband Networks. CommScope, he said, is gearing up for "big investments" from operators into mid-split enhancements.

Some of that accelerated action comes way of upstream usage that rocketed during the pandemic. "COVID[-19] kind of synchronized the upgrade activities and the thought process of everyone," Caezza said.

CommScope recently released a bandwidth consumption study based on data from a cross section of Tier 1 and Tier 2 operators. It found that demand for upstream capacity far outran increases seen in the downstream during the pandemic.

Although node splits have helped to relieve contention on the HFC network, Caezza says node splits face diminishing returns due to the higher expenses (partly in the form of more cable modem termination system ports along with the need for more space and power) associated with each node split. Expanding the spectrum used for the upstream, he adds, provides the additional bandwidth required to deliver much faster upstream speeds.

But cable operators have some technology and operational challenges to overcome if they try to make a move to 204MHz. Operators, particularly in North America, have to contend with potential interference, as their legacy set-top boxes leverage out-of-band signaling (used for set-top authorization and authentication) in a part of the spectrum on the network that would be allocated to the upstream in the 204MHz high-split scenario.

Cable operators will also have to work around a "noisy neighbors" issue that can crop up as signals move around the tap at these higher frequencies, Caezza explains.

"There is a lot of work being done" to resolve these issues, he adds, noting that CommScope has invested much effort over the last year or so trying to understand what's involved with a true high-split implementation.

Need for high-splits will vary by market

Despite those challenges and concerns, Caezza stresses that some operators are pursuing high-split upgrades. They include operators in the US and Europe that are already running "limited" field trials. Notably, cable operators in Europe have an advantage because they use different set-top designs that obviate the out-of-band problem that many US operators face.

Caezza estimates that a bigger push for high-split upgrade could be another three to five years out on the horizon, with mid-splits playing a major role in the here and now. But the market is far from uniform, as competitive demands vary by global region and can also differ within some regions.

Caezza expects to see more initial pickup for high-splits in Europe due to the aforementioned set-top situation paired with competitive pressures from fiber providers. He also expects parts of Asia to move more rapidly to high-splits as HFC networks there pivot to stay within shouting distance of PON network performance and because of the need to offer competitive 1-Gig upstream speeds.

In the US, meanwhile, operators could take a "targeted market approach" with high-splits where they face FTTP competition from the likes of AT&T and Verizon. In Latin America, he expects a greater focus on mid-split upgrades.

Long live the amplifier

The drive towards upstream upgrades and future overall plant capacity upgrades is also expected to push operators to upgrade millions of cable amplifiers.

That's a big deal considering the cable industry all but declared the coming death of the amplifier some two decades ago, when many believed that the need for them would end as fiber was pushed deeper and deeper into the network, Caezza recalls.

But reports of the death of the cable amp appear to be greatly exaggerated.

"I think the next few years are going to be the biggest years for amplifiers that we've seen in the last 20," he said.

Caezza estimates that, in the US alone, there are 12 million to 15 million cable amplifiers deployed, with most in need of upgrades. The new amps won't be free, of course, but the big bucks are in the labor to get them deployed and installed, he points out. Part of the payoff for operators will be a network outfitted with newer amps endowed with much lower failure rates than their predecessors.

"As the amplifiers get to be 15 to 20 years old, it's a lot like your car being 15 or 20 years old. You worry about whether or not it's safe to go on a long trip when your car gets that old," Caezza says. "Are you willing to take a long trip with that amplifier? Going from those old amps to new ones really enhances the reliability of the network and the overall subscriber satisfaction when it's all done."

In Caezza's view, 1.2GHz is the current state-of-the-art for cable amplifiers, with 1.8GHz, a potential spectrum target for future DOCSIS 4.0 networks, firmly on the company's roadmap. The cable industry has already started some work around future 3GHz technology.

He said CommScope is working with suppliers on 1.8GHz technology, but stresses that power requirements in that end of the spectrum remain a concern and a source of industry debate about how the issues can or should be resolved.

"We don't view it as having a viable solution there yet," Caezza said of 1.8GHz amplifier technology. "We're working towards it, but we're not quite there yet. We'll get there."

On that point, he said CommScope has some intellectual property pending that would enable operators to use existing gallium nitride (GaN)-based transistor technology. That proposed approach would likewise maintain amplifier spacing for an HFC network built to 1.8GHz, which would vastly reduce construction costs, Caezza said.

Related posts:

— Jeff Baumgartner, Senior Editor, Light Reading