After more than 40 years of operation, DTVE is closing its doors and our website will no longer be updated daily. Thank you for all of your support.
Teleste: cable upgrade paths diverging in North America and Europe
As cable MSOs look to upgrade their networks to deliver 10G services on both sides of the Atlantic, the technologies deployed in practice are continuing to diverge, according to cable access technology specialist Teleste.
Speaking to press at ANGA COM in Germany, Hanno Narjus, SVP at Teleste Broadband Networks, said that there have been significant shifts in strategy among cable operators over the last year.
He said that last year, DOCSIS 3.1 was widely available and cable operators were beginning to deploy Distributed Access Architecture (DAA) technologies in earnest to prepare for higher downstream and upstream broadband. The general view, he said, was that there would be two big changes – a shift to virtualised DOCSIS core and a move to Remote MACPHY-based DAA.
However, Charter Communications questioned whether to move ahead with two big shifts in one go. “It is clear that the industry is eliminating one technology shift by sticking to Remote PHY architecture while going for virtualised CCAP [Converged Cable Access Platform] Core,” he said.
Narjus said this would speed up the journey to DOCSIS 4.0. “It is excellent news and speeding up moves to 4.0,” he said.
He said there was now consensus-driven momentum to this shift.
Upgrade paths diverging
Upgrades are not proceeding in tandem in North America and Europe however, said Narjus.
In North America Remote PHY has already been deployed by two large MSOs – Comcast and Cox – with around 100,000 nodes installed.
In Europe, however, there are only limited DAA deployment – in Spain and Denmark – where there is a significant Remote MACPHY footprint. The latter is however based on Huawei’s now terminated line of products, and operators will need to make hard choices to move forwards.
While DAA in Europe is still at a nascent stage, European operators are well ahead in delivering higher spectrum. In the US, available spectrum in the network has lagged behind and upstream capacity remains a bottleneck, said Narjus. “Operators have been struggling to keep customers happy,” he said.
In Europe, by contrast, there have been large-scale upgrades to 1.2GHz downstream across the continent, with 65/85/200MHz upstream.
Narjus said that the difference is largely down to historic patterns of network deployment. In Europe, with its higher density of population, going from mid to high-split is relatively straightforward for cable operators.
Cable also has a smaller share of the overall broadband pie. In North Amer4ica DOCSIS has a 70% share of the market compared with 20% in Europe. “That makes a big difference in how investments are planned,” said Narjus.
With a larger cable footprint, in North America there is now a sense of urgency behind moves to DOCSIS 4.0, with mid split and high split rollouts using 1.8GHz technology. Operators want to build systems with 1.8GHz products as soon as amplifiers are available, skipping 1.2GHz upgrades, said Narjus.
The plan is to so a spectrum upgrade to 1.8GHz when DOCSIS 4.0 Remote PHY nodes are available, most likely from 2026.
Narjus said that this would benefit Teleste, because it is a leader in developing working 1.8GHz products.
All about fibre
In the US DOCSIS reigns, and there is little or no overbuilding of HFC networks by fibre, except in rural areas. Narjus said he believed it is likely that under 20% of US cable plant will be overbuilt by fibre by 2030.
In Europe, by way of contrast, “it is all about fibre”, he said. Operators are working out strategies to move towards fibre, because of competition from telcos and other fibre players.
The result is greater fragmentation and variety in upgrade strategies.
“In Europe, there is no one size fits all,” said Narjus. “There will be a combination of fibre and DOCSIS 4.0, even within markets.”
Some operators will invest fully in DOCSIS 4.0 to give new life to their HFC plant over the next 10-15 years, while at the other extreme, operators want to move to fibre straightaway.
In the middle there are operators who will extend the life of cable for the next five to 10 years while focusing on fibre in regions outside the existing coax network, he said.
Operators in Europe who commit fully to DOCSIS 4.0 will need to deploy DAA and then to move to DOCSIS 4.0 Remote PHY as soon as the technology is available. They will start rolling out 1.8GHz technologies, replacing amplifiers and passives before deploying home gateways.
Narjus said he believed cable-intensive markets such as the Benelux and Switzerland are likely to adopt this strategy.
Narjus said he also believed that some operators will try to squeeze more out of DOCSIS 3.1 instead of migrating to DOCSIS 4.0 everywhere, while selectively deploying fibre. They will introduce DAA as an intermediate step, as this also paves the way for Remote -OLT and 5G.
Some operators will consider deploying DOCSIS 4.0 modems with Remote PHY nodes, replacing amplifiers with 1.8GHz equipment and an ultra high split to offer symmetrical multi-Gigabit services, said Narjus.
Narjus said that the German market itself was lagging behind the rest of Europe because of legacy network issues. He said he expected Germany operators to squeeze out DOCSIS 3.1 and 1.2GHz plant, complemented by fibre overbuilds. He said that the legacy issues associated with German networks meant that swapping out passives is difficult making an upgrade to DOCSIS 4.0 a difficult proposition.
For some European operators, meanwhile, deploying fibre is the answer. Narjus said that these players may nevertheless continue to leverage DOCSIS to provide multi Gigabit access in MDUs, where taking fibre all the way to the individual home is difficult.
Narjus said that the group of operators focusing on fibre also need to maintain the HFC footprint for now, but will not introduce DAA and will not upgrade spectrum from the current 1.2GHz plant. Instead, they will continue building capacity as needed, including node splits.