Many predictions prove quite false. Consider the 2009 study of pandemic impact on internet access by the U.S. General Accountability Office. That study warned that, in a serious pandemic, “increased demand during a severe pandemic could exceed the capacities of Internet providers’ access networks for residential users and interfere with teleworkers in the securities market and other sectors.”
Pointedly, the report said “DHS (Department of Homeland Security) has not developed a strategy to address potential Internet congestion.” As it turns out, internet service suppliers, in the normal course of supporting their businesses, built networks robust enough that measures to deal with internet congestion simply were not needed.
It is worth noting that the report suggested “an influenza pandemic could result in 200,000 to 2 million deaths in the United States.” Though any loss of life is lamentable, that prediction has not proven close to correct.
“Increased use of the Internet by students, teleworkers, and others during a severe pandemic is expected to create congestion in Internet access networks that serve metropolitan and other residential neighborhoods,” the report warned.
Of course, a pandemic did happen in 2020, did result in the shutdown of most of the economy, people did have to stay at home, away from school and work. But the feared internet access disruption never happened. In fact, the percentage of people required to stay at home vastly exceeded the 50 percent figure the GAO assumed. It was virtually 100 percent in most parts of the United States.
For the week of March 15 to March 21, 2020, as people were ordered to stay at home, internet access services in 200 U.S. cities maintained service levels, though 13.5 percent of cities had seen average speed dips of 20 percent of typical ranges, according to Broadband Now.
In late March and early April, consumer traffic was up possibly 30 percent to 40 percent in affected countries where stay at home policies were in effect.
An earlier 2007 DHS study was said to “confirm that the increased traffic generated in neighborhoods during a severe pandemic is likely to exceed the capacity of the providers’ aggregation devices in metropolitan residential neighborhoods.” That has not proven to be the case.
Notably, the GAO report said that at 40 percent of absenteeism (workers forced to stay home), “at the 40 percent absenteeism level, the study predicted that most users within residential neighborhoods would likely experience congestion when attempting to use the Internet.”
The Covid-19 pandemic caused close to 99 percent stay at home behavior.
The point is that predictions always are hard to make. In this case, ISPs built robust networks that were able to handle the absolute worst case scenario for internet usage caused by a pandemic, with only a slight slowing of peak speeds.
If nothing else, the stay-at-home orders put into place to combat the Covid-19 pandemic have stress-tested consumer internet access networks.
That raises the question of whether data caps for fixed line services actually are necessary for the oft-stated reason of preserving quality of experience by reducing potential congestion.
Most consumers seem to understand that mobile networks likely require more management, have less bandwidth and are more prone to actual congestion than fixed networks. At least that is what a Government Accountability Office survey might suggest.
Fixed networks might be another matter. To be sure, use or non-use of usage caps is a business issue, not a technology issue. The pandemic performance pretty much confirms that. To wit, bandwidth caps are not needed, on today’s fixed networks, to prevent congestion. The consumer networks were able to handle the sudden and unexpected demand creating by nearly everyone staying away from work and school.
The next question is whether a significant number of large ISPs will decide there is market advantage to be gained by dropping the bandwidth caps, increasing them substantially (it should not matter, in terms of congestion), or abandoning caps altogether.
The performance of networks suggests it is technically safe to do so. The issue is whether business policies will change, or not.
No comments:
Post a Comment