• The cable operator landscape is shifting
• Cablecos are adopting telco-like service bundles that include broadband and mobile as well as voice and TV
• Now the cablecos have reached a broadband infrastructure crossroads
• Do they stick and upgrade with the cable sector’s DOCSIS broadband option or go down the full fibre route?

Telcos and cable network operators (cablecos) have spent the last 30 years or so probing each other's business turf and are now getting closer and closer in terms of what they can offer customers and how they operate. Those forays into each others’ service markets were often urged on by governments and regulators (as well as shareholders) in the interests of competition and, while each country did things differently and on different timescales, their convergent journeys have similar characteristics and have resulted in service portfolios that comprise voice, broadband, TV/video and mobile services.

The background

For many years, telcos and cablecos have been dipping their toes into each others’ waters. In the UK, for example, BT joined the 1980s hybrid fibre coax (HFC) multi-channel TV frenzy by acquiring two franchises to put some skin into the UK TV game. The cablecos snapped back by exploring the use of cable modems that converted one or more TV channels into a shared internet download capability, thus tentatively encroaching on what telcos had assumed was their data connectivity territory.  

At about the same time, both incumbent and competitive unbundling telcos took on the internet access challenge with broadband access services enabled by asymmetric digital subscriber line (ADSL) technology to provide an upgraded alternative to dial-up internet access. It was a similar story in the US, except it was the cablecos, not the telcos, that ended up with the majority of home broadband connections.

So, competition across established service categories was coming from multiple directions and starting points. In the UK, just to add to their challenges, cable operators found their TV services turf encroached by the introduction of commercial multi-channel satellite TV, which diminished the allure of “big bundle” cable. 

Partly in response to this, the cablecos looked to improve their revenue opportunities by striking back at the telcos with regulated access to the telephony business, offering both residential and business voice services. After all, if the telephone companies – energised by liberalisation – had the effrontery to offer ADSL-based internet access alongside voice services as a sweetener for customers, then some cable companies decided they needed to add telephony services to their cable TV packages to level the playing field.

The cablecos also needed an equivalent to ADSL, a way to standardise how they could offer internet access services over their HFC infrastructure. That came in the late 1990s when the US cable industry research consortium, CableLabs, published the first comprehensive data standard for cable service providers, the Data Over Cable Service Interface Specification (DOCSIS), which enabled cablecos to add more speed and functionality without requiring major changes to their basic physical HFC network infrastructure and which has underpinned cable broadband service delivery ever since.  

Stepping up to faster data

Fast forward through its multiple incremental improvements to 2013 and introduction of DOCSIS version 3.1, which enabled maximum download broadband speeds of 10 Gbit/s, channel bonding and, with a sub-10 millisecond low latency option added in 2019, kept the cable broadband sector competitive with the broadband capabilities of the telco community, which was exploring faster versions of the DSL specifications, such as VDSL2, and investing (in pockets) in fibre-to-the-home (FTTH) access networks. 

It’s worth noting at this point that while DOCSIS 3.1 has a theoretical top-end speed of 10 Gbit/s, that’s the maximum shared capacity of the network and not the downstream bandwidth the end-user experiences. That’s not really a big deal because, as any cableco executive will tell you, residential subscribers don’t need anything more than 1 Gbit/s and, indeed, most of the time need much less than that, even in this day and age. 

In addition, just because the DOCSIS specification can enable 10 Gbit/s, that doesn’t mean the HFC infrastructure is capable of managing that level of service without a costly upgrade to various components, cables and customer premises equipment. Even if this was not the case, broadband business (and marketing) models suggest it’s prudent to develop and build up tiers of service offerings that provide scope for incremental increases in speeds (and prices) as and when customers might need the speeds and are prepared to pay for the upgrades.  

So, anxious to avoid the trap of over-investment, cablecos have historically road-mapped software enhancements and kept costly hardware upgrades at a minimum while constraining speeds, hence the persistence of HFC – until, that is, the tick-tock, back-and-forth volleying between software and hardware means a major capital investment or upfront cost per individual subscriber is required to keep up with the latest high-end applications or to satisfy leading-edge users. 

For some cablecos that time may be approaching – others are adamant that it’s not. As a result, the next iteration of the DOCSIS standard has options for different approaches to network infrastructure deployments: Cablecos are now facing decisions around the if, when and how to deploy the next step – DOCSIS 4.0.

How soon is now?

DOCSIS 4.0 was first announced in 2017 and technology products and systems based on the specifications are being made ready for commercial availability this year. It will sport a full-duplex option (which enables data to be transmitted in both directions over one signal simultaneously) to enable operators to offer symmetric services at relatively high speed (the underlying network cables still have an upstream limit, so for symmetric services that upstream limit becomes the downstream limit too) and it pencils in a range of other options to enable cablecos to choose what best suits their broadband network and services strategy. 

So how do the cablecos appear to be approaching the DOCSIS 4.0 opportunity?

Cablecos are no doubt taking account of what appears to be a general levelling off of data demand (even allowing for a post-pandemic rebound in consumption). While there has very recently been much excitement generated by the impact of the generative AI (GenAI) applications, the fact is these large language model (LLM) AI applications drive data capacity demand mostly in datacentre-to-datacentre networks, not in broadband access networks. In fact, as more applications (even gaming) become ‘cloudified’, it’s likely the shift will play its part in keeping access network data demand growth relatively subdued for the foreseeable future: Even the much heralded arrival of virtual or augmented reality applications look unlikely to move the needle all that much. That’s not to say that data demand for broadband access has stalled, just that growth in capacity demand will no longer be exponential (if it ever was).

At the same time, of course, mobile broadband services have become a critical component to any type of communications service provider offering and, in the spirit of convergence, cablecos all over the world have, in one way or another, bundled mobile into their service packages, especially in markets where cable TV services (for which demand is rapidly on the wane) were for a long time at the core of service bundles. In the US, for example, the major cablecos have been adding mobile services in recent years, with Cox Communications joining the party just this year, as this article from Axios explains. In fact, Cox has been stressing the increasingly convergent nature of its services this year as it tries to grab a slice of the US mobile services pie.  

Caution: Works in progress

Understandably, cablecos have historically remained cautious when it comes to upgrading their networks in expectation of promised (or feared) runaway data growth that never happens, for the very good reason that tinkering with fixed network infrastructure can be eye-wateringly expensive when it involves millions of subscribers. Go too early and a cableco may have committed a huge dollop of capital without a commensurate revenue uplift in what is now a highly price-competitive market; move too late and it may start to bleed subscribers. Caution coupled with innovation are the watchwords.

Cableco strategies, therefore, obviously depend on the variables hitting each provider’s business model: The age of its infrastructure; the demography of its subscribers (and how likely they are to be interested in, or afford, the latest shiny applications and services); the level of historic disgruntlement (built up over decades of customer experience failures); and so on.

For many, the next step up is to deploy DOCSIS 4.0 for high-speed services on the existing cable infrastructure, perhaps with modifications to elements such as filters on the coax cable infrastructure. Some are expecting to roll out DOCSIS 4.0-enabled systems over the next few years, at least in the form of large-scale trials. Many others are more cautious, and may be pencilling in their upgrades for up to 10 years from now, content to sit on earlier versions in the meantime, perhaps with minor upgrades, especially if their main telco rivals are still in the early stages of their FTTH rollouts.

There’s little doubt that DOCSIS 4.0 will enable cablecos to deliver data at sub-10 millisecond latencies, potentially enabling them (or whoever ends up using their infrastructure) to keep up with the capabilities of most other network types, depending of course on customer demands and how much competitive pressure is being applied in specific markets.

Some cablecos, however, are aiming higher, investing in optical fibre connections to some premises or aggregation points on at least some of their infrastructure, thus enabling them to compete on a broader range of services to specific customers or areas. Others are aiming to effectively morph into telcos by abandoning their HFC heritage and deploying passive optical networking (PON)-based full fibre to the premises architectures to match their telco peers. 

This is the path that Virgin Media O2 appears to be taking in the UK, as it plans to replace its cable access infrastructure with FTTH technology by 2028. Once completed, it will have effectively shifted from being a pure cableco to one that also owns and runs a mobile network (following its 2021 merger with O2 UK) and, ultimately, to running an extensive fibre and 5G convergent access infrastructure, so completing the circle and essentially positioning itself as a technical and service equivalent to its main rival, BT.    

We’ll look at this transformation to the PON option and discuss its opportunities and its potential impact on the market in a follow-up article.

- Ian Scales, Managing Editor, TelecomTV