DOCSIS 3.1 vs. FTTH: Is That the Real Question?

Once again, cable companies are facing intense competition from Telcos in the neverending Residential Bandwidth war. Even though cable has lately taken an essential advantage with the deployment of DOCSIS 3.0 technology, which allows them to offer 300 Mb/sec packages as a regular service today, largely exceeding typical xDSL services (even VDSL2), now Telcos are answering with PON/FTTH deployments.

This movement, which has been delayed as much as possible due to its high cost, is currently being rolled out by Telcos as the only way to keep the pace in this war. Laying fiber to the home requires heavy investment in replacing the old copper cables with fiber (around 70% of the investment needed on a green field FTTH network is utilized in laying fiber and passive optical nodes in the outside plant). However, once it’s done, the advantages are pretty obvious: Multi-gigabit speeds based on a passive, low maintenance access network, ready for the foreseeable future. It’s for sure that most telcos will not replace all their copper plant with fiber in a short time frame, as it may take several years, so cable companies need to evaluate their next move carefully.

What do we really know about fraud in DOCSIS networks? Read on to find out 5 Quick Tips about Fraud in DOCSIS Networks

Protocols and Architectures

First, let’s clarify that it is NOT FTTH vs. DOCSIS 3.1. FTTH is a Point to MultiPoint (P2MP) network architecture, DOCSIS 3.1 is a transport mechanism. Cable companies taking full advantage of DOCSIS 3.0 can already provide services close to the gigabit speeds: Theoretically, you can bond 32 channels, 38Mbps width to get 1.2Gbps (or 24 Euro DOCSIS channels, 50Mbps ) but this seems unpractical or, more specifically, a costly approach.

DOCSIS 3.1 initially offers 10Gbps/2Gbps asymmetrical service with plans for a symmetrical service in the short term, and it can take advantage of HFC already deployed infrastructure. But this is only partially true, as new CMTSs will be required to use full DOCSIS 3.1 capabilities. Instead thru an SW upgrade, some CMTSs can offer partial DOCSIS 3.1 support, still better than DOCSIS 3.0. Also, the outside plant may require some changes as cable distance from nodes to homes needs to be reduced, and finally new cable modems need to be installed at home.

On the other hand, deploying a new fiber network can be future proof regarding speed. Current GPON standard allows for 2.5 Gbps/1.2Gbps asymmetrical services, but the next step in PON evolution, XGPON is around the corner with also 10Gbps/2Gbps asymmetrical service that will match DOCSIS 3.1 capabilities. Even more, some kind of multi lambda approach, like WDM or technologies like OFDMA and several others under study for PON networks can easily extend fiber speed well beyond these numbers, as it currently happens with other optical-based transport technologies. In fact, plans for long-term evolution to 40Gbps are well underway. The optical technology itself is very efficient, simple to maintain (being based on passive elements on the outside plant) and more cost-effective than DOCSIS in terms of infrastructure ($ per bit/sec). But again, it initially requires a very expensive layout of a whole new fiber network.

Alternatives

In the telco world, dealing with aging copper twisted pair networks seems the path is just one, deploying fiber on those areas where it makes economic sense. For the rest of the network, VDSL/VDSL2 that offers significant improvements over older xDSL standards can be the interim answer. This twisted pair technologies cannot cope with cable service competition but can help Telcos extend the life of current networks until they deploy FTTH services network-wide.

For cable, things are a bit more difficult as DOCSIS 3.0 still provides very competitive speeds. The upgrade to DOCSIS 3.1 will give them the fastest residential service nowadays at a significant cost, and replacing the HFC network with fiber will provide a future-proof infrastructure but at the greatest initial cost of all options (even though later on operating costs are significantly less).

The question here would be, do you really need gigabit service in your country/region? Is that a truly competitive differentiator? Keep in mind that most US homes are still comfortable with services in the 50/60 Mbps range. There’s already fastest services being offered with even some Gbps ones, but that is still not a real requirement for most residential users. In most Latin American countries, the average is even lower, in the 20/30 Mbps at most of the leading countries. Again, there are also faster services, but those are not exactly the most popular packages.

The Final Solution

Fiber, and particularly PON networks, seems to be the long-term answer for residential internet services. In the meantime, your next move will heavily depend on several factors like average ARPU (are your customers willing to pay a premium for a Gbps service), homes networking gear (most residential routers and switches cannot take advantage of Gbps service), competitive pressure, and the budget available to invest in network upgrades. With the sole exception of Greenfields where fiber would be a natural choice, the most likely scenario would be a mixed one with fiber in new neighborhoods or very high ARPU areas and some limited investment in intermediate technologies in the rest of the network.

Managing new opportunities DOCSIS 3.1 provides requires both standard monitoring and new predictive tools. Learn what’s needed to monitor DOCSIS 3.1