One thing that should be clear to everyone at this point is that one access technology doesn’t fit all, at least with respect to broadband in general and broadband Internet in particular. Thus, the question isn’t when we’ll converge on one (fiber is the one most mentioned) but what indicates the best, or even a suitable, access technology. From there, we’d need to know how operators would be impacted by whatever multiplicity in access options we could expect to see.
In theory, you could use any technology that can deliver the desired speed to provide Internet access. In practice, the cost of a given technology will likely determine whether it can offer that speed goal at a cost that will provide the ISP with a satisfactory rate of return. Since access providers don’t normally deploy to individual sites but to areas, the economic characteristics of the area and the requirements of the technology will combine to determine validity of a technology choice.
Two factors seem to determine the deployment efficiency of a given area. The first is the economic density per unit area, which I define as GDP per square mile. The second is the access efficiency of the area, which I’ve found to be proportional to the miles of roadway in the area, since that relates directly to the right-of-way that’s available to place or run things. Years ago I created a metric to reflect both, which I called “demand density”. My analysis of various market areas suggested that, if we set the US value of demand density as 1.0, demand densities above roughly 4.0 would support almost any access option, and those below roughly 0.7 would be difficult to support profitably with any physical-media access technology.
When demand density for any market area is above roughly 4.0, that means an operator could make a mass-deployment in the area profitably. As demand density falls, what happens in the real world is that pockets of the area, large a first then getting smaller and more sparse later, will still support broad deployment. In the US, for example, you have areas (like the northeast) where an area is likely to be largely suitable for any access option, and others where towns and cities will be suitable but other suburban and rural areas will not be. In some low-population areas, few or even no hot spots of opportunity will exist, in which case only over-the-air broadband options will be suitable.
Most communities with populations of 20,000 or more will be suitable for physical-media deployment if the household income is average. Communities with more than 10,000 population may be suitable if incomes are higher or if business density is high. Thus, suburbs of major cities are likely suitable, but deep suburbs where lot sizes are large may be an issue. Even in countries with very high demand density, there are always pockets where physical media will not be profitable, but operators may accept this because the area is so profitable overall.
As demand density increases, options for physical media increase. Today, copper loop is generally not considered suitable for competitive broadband delivery. Fiber is of course the preferred strategy, though CATV cable can provide effective and profitable broadband. Operators I’ve talked with do tend to prefer fiber overwhelmingly for areas dense enough, and cable MSOs are staying with their CATV for home connection, though some are taking fiber to the home in select areas and nearly all propose deeper fiber deployment.
FWA is arguably the hottest access technology. It’s a logical extension to the FTTN approach that originally used copper loop for the last mile, but with a much-improved broadband service potential and lower cost to pass or connect sites. FWA is particularly valuable in deeper suburbs, and non-dominant access providers are interested in it for small communities and target penetrations of competitive areas. It’s particularly useful in markets where average demand densities run from roughly 1.0 to 2.5 because these markets can rarely be served entirely with fiber.
Cellular 5G can also be used for fixed access, and can thus offer combined mobile/fixed presence in a market, but this avenue for broadband is limited by spectrum availability, particularly for small alternative providers, and in most cases it can’t match fiber, CATV, or FWA in speeds. It’s achieved success in the US for operators as a means of offering fixed broadband where they also provide mobile services, but most operators doubt it can sustain full competition with other options for delivery of streaming video.
Of 88 operators who offered comment on this, only 11 believed they would settle on a single fixed broadband technology. Another 41, including nearly all the cable MSOs, thought secondary options would serve less than 15% of their customers, and all the remainder believed they would offer two or more technologies “regularly” in order to meet their business needs.
No operator proposed offering “major site” enterprise business services via anything but fiber, but all 88 operators said they believed that residential broadband technologies would likely serve at least two-thirds of their business branch sites and almost SMB sites within five years. Those who currently deploy enterprise VPN services using carrier Ethernet said they believe the smaller sites would convert to residential technology in all but the major metro areas, though they would rarely suggest when this would be largely complete. Those who would comment on timing usually gave a timeline of 3-5 years. All operators acknowledged they had some SD-WAN sites already.
The shift of business sites to residential broadband infrastructure, say operators, has a tendency to elevate interest in the technology options and their optimization. Operators admit that SD-WAN in particular has put pressure on the technology aspects of the competitive dynamic. Since managed service providers (MSPs) often control SD-WAN deployments and competitive CSPs often use SD-WAN to gain access to customers, inferior Internet access technology can bite in business services that are overall much more profitable than residential broadband. This competitive impact would then increase provider interest in optimizing access through multiple technologies.
This, of course, could increase opex if network operations and service/customer experience management tools in use are technology-specific. Roughly half of all operators say they are looking harder at technology-independent tools and practices in 2024 than they had in the past. Harmonizing these areas to simplify operations overall may become a major goal, even a requirement, by the end of the year.
