Forecasting is a perilous business, and forecasts "always" are wrong, to some extent. That holds for estimates of Internet bandwidth demand as well, which historically tend to overestimate long term rates of growth.
“The evidence reveals a persistent tendency to overestimate in both number and value,” say researchers Aalok Mehta of the University of Southern California and J. Armand Musey, Goldin Associates, LLC principal.
“Of the past seven Cisco mobile traffic forecasts for North America, for example, overestimates were nearly twice as frequent as underestimates (19 vs. 10),” they say. “Overestimates are also on average of greater magnitude than underestimates (103 vs. 81 PB/month).”
In the case of mobile bandwidth, unexpected developments such as Wi-Fi offload could have had an impact, as users learned to substitute Wi-Fi networks for mobile network demand. True, aggregate Internet demand does not change simply because a different access network is used.
But the specific amount of mobile network demand does change. Such estimates matter for policymakers, Internet service providers and ecosystem participants, as matching supply to demand hinges on the predictability of demand growth forecasts.
With the important caveat that end user demand and “traffic growth” are different items (set high prices and demand will drop, for example), “everybody” might agree that mobile traffic will continue to grow, for the simple reason that more consumers are buying smartphones, which means there are more potential users of the Internet access feature, and more users watch video, the most bandwidth-intensive application of all.
Also, there is a physical dimension to mobile bandwidth forecasts that does not exist to the same degree in the fixed access area.
When traffic is confined to a waveguide (wire or optical fiber), there often are multiple ways to increase available bandwidth. Internet service providers can use multiple wires or fibers, for example.
Up to a point, that spatial division also is how small cells work. But the risk of interference arguably puts limits on the practicality of ever-smaller cells as a means of increasing effective ability to support higher volumes of traffic.
Though some might claim there are no theoretical limits to mobile bandwidth supply, most would admit there are some physical limitations, where it comes to commercially-viable supplied bandwidth, even if network architecture, improvements in coding and modulation, antenna design, commercial tariffs and offloading mechanisms do exist.
Precisely how much mobile or untethered spectrum might be required, and when, remain matters of some debate. Mehta and Musey correctly argue that over-allocation of capital in spectrum resources is unwise.
So is under-investment, says George Ford, Phoenix Center for Advanced Legal and Economic Public Policy Studies chief economist.
The issue in the mobile realm is the time it takes to release additional spectrum for commercial use, as well as the physically-available alternatives for releasing new spectrum.