DOCSIS 4.0 vs. Fiber

Comcast and Charter previously announced that they intend to upgrade cable networks to DOCSIS 4.0 to be able to better compete against fiber networks. The goal is to be able to offer faster download speeds and drastically improve upload speeds to level the playing field with fiber in terms of advertised speeds. It’s anybody’s guess if these upgrades will make cable broadband equivalent to fiber in consumers’ eyes.

From a marketing perspective, there are plenty of people who see no difference between symmetrical gigabit broadband offered by a cable company or a fiber overbuilder. However, a lot of the public has already become convinced that fiber is superior. AT&T and a few other big telcos say they quickly get a 30% market share when they bring fiber to a neighborhood, and telcos claim aspirations of reaching a 50% market share within 3-4 years.

At least a few big cable companies believe fiber is better. Cox is in the process of overbuilding fiber in some of its largest markets. Altice has built fiber in about a third of its markets. What’s not talked about much is that cable companies have the same ability to overlash fiber on existing coaxial cables in the same way that telcos can overlash onto copper cables. It costs Cox a lot less to bring fiber to a neighborhood than a fiber overbuilder that can’t overlash onto existing wires.

From a technical perspective, engineers and broadband purists will tell you that fiber delivers a better broadband signal. A few years back, I witnessed a side-by-side comparison of fiber and coaxial broadband delivered by ISPs. Although the subscribed download speeds being delivered were the same, the fiber connection felt cleaner and faster to the eye. There are several technical reasons for the difference.

  • The fiber signal has far less latency. Latency is a delay in getting bits delivered on a broadband signal. Higher latency means that a smaller percentage of bits get delivered on the first attempt. The impact of latency is most noticeable when viewing live sporting events where the signal is sent to be viewed without having received all of the transmitted bits – and this is seen to the eye as pixelation or less clarity of picture.
  • Fiber also has much less jitter. This is the variability of the signal from second to second. A fiber system generally delivers broadband signals on time, while the nuances of a copper network cause minor delay and glitches. As one example, a coaxial copper network acts like a giant radio antenna and as such, picks up stray signals that enter the network and can disrupt the broadband signal. Disruptions inside a fiber network are comparatively minor and usually come from small flaws in the fiber caused during installation or later damage.

The real question that will have to be answered in the marketplace is if cable companies can reverse years of public perception that fiber is better. They have their work cut out for them. Fiber overbuilders today tell me that they rarely lose a customer who returns to the cable company competitor. Even if the cable networks get much better, people are going to remember when they used to struggle on cable holding a zoom call.

Before the cable companies can make the upgrade to DOCSIS 4.0, which is still a few years away, the big cable companies are planning to upgrade upload speeds in some markets using a technology referred to as a mid-split. This will allocate more broadband to the upload path. It will be interesting to see if that is enough of an upgrade to stop people from leaving for fiber. I think cable companies are scared of seeing a mass migration to fiber in some neighborhoods because they understand how hard it will be to win people back. Faster upload speeds may fix the primary issue that people don’t like about cable broadband, but will it be enough to compete with fiber? It’s going to be an interesting marketing battle.

4 thoughts on “DOCSIS 4.0 vs. Fiber

  1. To improve upload speeds of cables will lead the end of roads of cables! Because the usage based on the upload speed will finally come to true for mass and China will no longer have reason to control the upload speed.
    Cables want to delay that timeline but AIGC will give cables no chance to hold back anymore.

  2. During the meanwhile FWA is taking off, with AT&T joining the party:
    https://www.telecompetitor.com/internet-air-att-embraces-fixed-wireless-after-all/ – I think that will provide more near term competition for cable OR fiber in more rural areas.

    Also, although I very much appreciate your descriptions of latency and jitter above, since I care about reducing those so much, DOCSIS 4.0-*LL* is the technology that actually can get docsis almost into the range of fiber (GPON is a min of 250us, 4.0-LL is in in the 1ms range)… but, with L4S, the cable industry appears to desire to charge more for subscribers to gain access to it. Fiber does need to have an answer to L4S (mine is FQ with fq_codel or cake), I think…

    Improving the mid-splits will help the cable industry a lot.

  3. As much as I don’t want to promote big cable lol… DOCSIS 4.0 or even (3.1) is really good. If we’re talking bang-for-your-buck, swapping out a couple remote CMTS and all the customer modems is vastly less expensive and quicker to deploy.

    That said, I don’t like a mono-crop of internet service. I just don’t like how everyone is drunk on fiber. DOCSIS 3.1 or 4.0 offers REALLY similar latency as *GPON and similar speeds as well, and these ‘cable’ plants are really fiber networks with cheaper/easier to maintain last ‘feet’.

    Urban areas should, at least in my opinion, have fiber AND coax/copper services that are first class from at least 2 vendors because markets need choice/competition.

Leave a Reply