AfterDawn: Tech news

Research shows data caps dont effectively target an ISPs most expensive customers

Written by Rich Fiscus (Google+) @ 08 Dec 2011 5:19 User comments (3)

Research shows data caps dont effectively target an ISPs most expensive customers ISPs have been arguing for years that the heaviest data users should pay more for their service because they put a bigger strain on the network. This is the primary rationale for tiered data plans which cap downloads at varioius sizes and charge overage fees for additional data.
A new report from Benot Felten challenges that assumption based on analysis of traffic data from an unnamed North American ISP. His conclusion based on that analysis is that heavy data users do not, in fact, put a disproportionately large strain on ISP networks.

While the study itself must be purchased if you want to read the details, he has posted a free summary of his key findings.

  • The top 1% of data consumers (hereafter Very Heavy consumers) account for 20% of the overall consumption.
  • Average data consumption over the period is 290 MB, while consumption for Very Heavy consumers is 9.6 GB. Thisroughly equates to data consumption of 8.7 GB and 288 GB per month, respectively.
  • However, only half of these Very Heavy consumers are customers of the highest service tier (6 Mbps), which implies that half of them have bandwidth usage restricted to 3Mbps (the next service tier) or lower.
  • 61% of Very Heavy data consumers download 95% of the time or more, but only 5% of those who download at least 95% of the time are Very Heavy data consumers.
  • While 83% of Very Heavy data consumers are amongst the top 1% of bandwidth users during at least one five minute time window at peak hours, they only represent 14.3% of said Top 1% of users at those times.


So what does all that mean? The key points are the first and last. Despite the fact they consume 20% of the data flowing across the network, the heaviest data users account for less than 15% of the heaviest data consumers during peak times.

In other words many other users have just as great an affect on peak usage. This is important because peak usage is essentially how you can measure the bandwidth cost for the ISP. That cost remains the same even when no one is using the service. Unlimited downloading during non-peak hours doesn't cost the ISP anything extra.

Further down in his post he sums this up, writing:

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. Furthermore, many users who are "as guilty" as the ones who are over cap (again, if there is such a thing as a disruptive user) are not captured by that same net.

Previous Next  

3 user comments

18.12.2011 11:15

So...........what.........?????.........sounds like the solution is to lower prices, remove caps and just let things be??

28.12.2011 13:19

Not quite. The solution is to remove caps, because they do absolutely nothing, and invest in infrastructure upgrades.

The problem with congestion is not on the internet end. The internet infrastructure is not even remotely close to congestion. All of the congestion happens on the ISPs end. Because the ISPs refuse to spend the money to upgrade their own data centers and the nodes that their costumers connect to.

What the report is saying is that data consumption is absolutely meaningless because the user is already being bandwidth capped based on the service tier he purchases. Using the articles stats...

A "heavy user" who consumes 288gbs of data a month is still capped at the 3 to 6mbps of bandwidth as per his service plan. Even though he consumes more data, he can not consume more than the 3 to 6mbps of bandwidth assigned to him at any moment. The "average user" who consumes only 8.7gbs of data a month consumes that very same 3 to 6mbps of bandwidth that the "heavy user" does, albeit during peak times.

The "average user" may only be using his connection at "peak hours". While the "heavy user" is using his connection 95% of the time. However, congestion only happens during "peak hours". Therefore the heavy user's usage has no impact during "off-peak" hours. While the "average user" consumes the same amount of bandwidth as the "heavy user" during peak hours. Meaning that the "heavy user" and "average user" have the exact same impact on congestion.

The "heavy users", despite consuming 20% of the total data in a given month, only account for 1% of the total users. So take a typical ISP node and assume 400 houses are connected to this one node. That means 4 connections are "heavy users" and 396 users are "average users".

Given the fact that they are allotted the same amount of bandwidth, even if you remove all "heavy users" from the node, it will make absolutely no difference. You still have 396 "average users", of which ~42% are consuming the same amount of bandwidth during "peak hours" (per the link, also quoted below).

Quote:
42% of all customers (and nearly 48% of active customers) are amongst the top 10% of bandwidth users at one point or another during peak hours.
Thus, data caps have no affect on congestion. If congestion is happening it means that the node is oversold / oversubscribed, and needs to be upgraded.

P.S. The ISP that gave their data up for analysis is Sonic.net. Their CEO has a blog, and he has come out on this very same topic numerous times. Here's his blog post on this very topic.
This message has been edited since its posting. Latest edit was made on 08 Dec 2011 @ 13:20

39.12.2011 8:45

I think it is more complex than that...

My ISP will throttle the heck out of you if you down load heavily over a long enough period during prime time. You get a 10X period of grace over the weekend. I am sure most other big ISPs do the same. They train the heavy users to use the most when the lines are light. This is good for everyone.

The grace period is also governed by your tier. I seen nothing wrong with any of this. The heavy users that do not pay their fair share do so because they grab what they can, when they can. I don't see anything wrong with this either.

Comments have been disabled for this article.

News archive