SCTE/ISBE CABLE-TEC EXPO – An upgrade to a 1.8GHz spectrum ceiling represents the new frontier for cable operators that are looking at the Extended Spectrum DOCSIS (ESD) option for DOCSIS 4.0. But lifting the spectrum ceiling to that lofty level also comes with some power issues that could potentially nullify some of the capacity that operators aim to squeeze out of the network by raising the bar.

As discussed at this week's annual (and all-virtual) cable engineering event, new, compact Distributed Gain Architecture (DGA) amplifiers represent one potential fix. But it's also one that's being hotly debated.

By utilizing the ESD option for DOCSIS 4.0 and a spectrum upgrade to 1.8GHz, cable operators will be faced with needing to deploy a lot of new amplifiers.

"Every amplifier is going to need to be upgraded," John Chapman, a fellow and CTO of cable access at Cisco, explained as he kicked off a workshop this week dedicated to new amplifier considerations with Extended Spectrum DOCSIS.

But a major challenge is how to prepare the plant without the expense and labor needed to respace the amplifiers while also getting a grip on the higher distortion levels caused by upgrading the plant to 1.8GHz. As energy requirements and distortion levels rise, the network runs the risk of reducing modulation levels that impair carrier-to-noise ratios. That, ultimately, will reduce the data performance of cable modems running on the network.

Canada's Shaw Communications is taking a close look at the power challenges presented by 1.8GHz by testing the mettle of DGA. This technology involves the deployment of small booster amps that can distribute the power needs of the plant between the node and the home rather than deploying new, higher-power amps.

Shaw, which used DGA amps from Technetix for its trials, tested the technology for some longer stretches of plant that were "prime candidates for booster amplification," explained Nader Foroughi, senior network engineer at Shaw.

"The booster amplification seems to be an excellent solution for areas where we struggle ... especially in the longer plant models that we analyzed." Foroughi said.

Carrier-to-noise levels were solid even in cases where an existing node+2 architecture (two amplifiers between the node and the home) rose to N+6 or N+8 with the addition of booster amps to the equation, Shaw's study found.

Additionally the small form factors of the booster amps are "quite attractive," as they can fit into many of the pedestals on the outside plant instead of needing to upgrade a traditional amplifier, Foroughi explained.

The early test results also provide some insight into what HFC architectures might look like if the networks are someday upgraded to 3GHz, he added.

"This is fascinating type of stuff," said Chapman, who moderated the session. "It's replacing bigger, larger, fewer amplifiers with smaller, lower-power and higher number of amplifiers … but better behaved amplifiers. It's a whole new way of doing HFC architectures."

Technetix, which supplied the DGA amps for Shaw's tests, said other trials are underway with pre-production units. The vendor expects to reach full production by March 2021.

According to company CEO and founder Paul Broadhurst, Technetix originally developed the technology for Full Duplex DOCSIS (FDX), another option for DOCSIS 4.0 that can run upstream and downstream traffic in the same block of spectrum. But the vendor found that the technology could also be applied to ESD to help with the power problems at 1.8GHz. And the removal of diplex filters in this scenario can also eliminate the need for a guard band typically used to separate the spectrum dedicated to downstream and upstream traffic.

Flexible upstream

Removing the diplex filters could give cable operators the option to dynamically adjust the amount of spectrum being dedicated to the upstream and downstream as needed, Broadhurst said in a recent briefing, referring to the technique as a "flex split."

Rather than requiring operators to upgrade the upstream with a fixed "mid-split" that increases the available upstream spectrum to a range of 5MHz to 85MHz or a "high-split" that raises it up to 204MHz, operators could effectively alter the splits as desired by programming the node in new distributed access architectures, Broadhurst said.

Other power options being explored

But not everyone is sold on the use of DGA and the deployment of booster amps when raising the spectrum bar to 1.8GHz.

John Ulm, engineering fellow, broadband systems at CommScope, agreed that network performance gets worse and puts on strain as plant is moved to higher frequencies.

"We're really going to press the system" at 1.8GHz, he said during the workshop. His paper argues that the addition of mid-span amplifiers to boost the gain for those links "should not be taken lightly," as they add more active components to the network/system.

Ulm held that a concept that's been around for decades – unity gain – could also help to solve the issue that will crop up at 1.8GHz, and possibly relegate the need for booster amps to "select occasions."

As explained in the white paper supporting Ulm's presentation, unity gain is a concept of setting every amplifier output signal to the same shape and magnitude, no matter whether it's the first amp or the last one in a cascade.

"[T]rying to maintain unity gain is a key objective in any HFC design," the authors explained. Unity gain, they added, also has some "conventional wisdom" upon which to draw, as the technique has already helped to maintain a constant QAM modulation for legacy set-top boxes as well as DOCSIS 2.0 and DOCSIS 3.0 modems.

"There are a lot of tools in DOCSIS," Ulm said during this week's workshop. "Let's see how much capacity DOCSIS can get out of a system even with us falling a bit short on the unity gain … The average capacity might actually be good enough."

Related posts:

— Jeff Baumgartner, Senior Editor, Light Reading