Is the ‘Bandwidth Hog’ a Myth?

Sharing, you learned in kindergarten is the right thing to do. So when broadband providers said that “bandwidth hogs” are slowing the network for everyone, you probably thought it was fair to share those resources, well, fairly. But it turns out that the providers weren’t telling the truth (there’s a shock) about the causes of network congestion.

A new study shows that the so-called hogs aren’t the cause of congestion  — the villain is  peak use by folks who don’t come close to hitting the download limits that are causing the problem.

“Our analysis confirms that data consumption is at best a poor proxy for bandwidth usage,” writes Benoît Felten, chief research officer of Diffraction Analysis, which conducted the study. “The correlation between real-time bandwidth usage and data downloaded over time is weak, and the net cast by data caps captures users that cannot possibly be responsible for congestion.”

The caps, first imposed by Comcast a few years ago, and more recently by AT&T and smaller broadband providers, come asmore of us are using the Web to download and view data-heavy HD videos and movies. Most of the new pricing strategies revolve around data caps: a level of monthly data consumption that triggers pay-as-you-go mechanisms at steep per megabyte rates.

The providers haven’t been eager to prove the truth of their claims. Indeed, in 2009 Felten offered the services of his company free of charge to evaluate how the hogs affected network performance, but it took several years for a mid-sized U.S. DSL provider to agree.

That company’s name hasn’t been released but the results have been. (The full report is proprietary, but you can read a summary here.)

• The number of users is fairly constant during most of the day, the study found, but average bandwidth use spikes in the late afternoon and early evening. So who’s using the bandwidth? It turns out the answer is – nearly everyone. During the three peak hours of the day, nearly half of the company’s active users were sucking up huge amounts of bandwidth. If that’s the case, slapping extra charges on a few so-called data hogs wouldn’t make much difference.
•  
• Much of the use by “hogs” is at non-peak hours, which means that they are not having a significant effect on network performance.
•  
• The researchers also looked at usage by customers who exceed their data caps and were charged extra.  Of those, 78 percent were part of the top 1 percent of all users during peak hours. This means that roughly 22 percent of the heaviest data users were being blamed for peak-hour congestion that they couldn’t have caused.
•  

If data hogs aren’t causing the problem, what is? Bandwidth use by all users is obviously rising, and the sensible strategy to cope would be to expand the networks. But the carriers and cable companies would rather raise prices and pocket the extra revenue than endanger their precious profit margins by adding more capacity.

(Kudos to Ars Technica for noticing this study first.)