Network Neutrality is a term used to encompas the need for a balanced playing field on the Internet. Proponents for neutrality claim that having a tiered infrastructure will create a hierarchy of “haves” and “have nots.” The perceived problems would be similar to U.S. politics’ main problem: everyone has an equal vote; but those with money have a slightly “more equal” weight. Proponents claim that this same “class separation” would occur were network traffic to become tiered.
Opponents to “network neutrality” note that traffic on the network is saturating the pipes which carry data from one computer to another. They claim that one way to reduce traffic is by means of economics. By tiering network traffic based on content type, services with high bandwidth demand would become more expensive; the theory of supply and demand suggests that these high-traffic services would become less common.
I agree with opponents of network neutrality in that there is a saturation limit being met. However, I do not feel that tiering traffic based on content type is the way to address the problem.
My family’s ISP is facing a much larger drain on its resources than bandwidth-heavy services such as video and audio. This problem is the constant, daunting issue of handling the high ratio of unsolicited mail, or “spam” to good, expected mail, or “ham”.
A better approach to the solution of network saturation (in my humble opinion) would be to provide a base level of service at no cost, and charge the participants a content-neutral fee for going over this rate limit. This is the way that ISPs are charged for inter-ISP communication, and with a little bit of engineering can be extended to customer-to-ISP communication.
Let me give an example of when network traffic fees would apply, and when they would not be applicable. When you have two or more computers in your home, this network type is called a “LAN” or local area network. Any traffic that passes between the computers on your LAN travels across your own wires (or, as is becoming more common, your own wireless network). On a properly configured network, LAN traffic is not subject to fees of any kind.
If your neighbor has a computer in their home and you wish to communicate data, such as images, internet phone calls, movies, music, etc, you will likely need to pass these files over your ISP’s network. All of the files that are sent from your system to theirs will be accounted as Internet traffic, and rate limiting will apply. Your ISP might provide a base service of, say, 56 kilobits per second at no cost. This is enough to have an internet phone call with video, or to listen to your favorite radio show. It is even enough to stream a low resolution video.
However, if you combine these services, ie, talk on the phone while a video is playing in the background, you may exceed your rate limit.
There are a number of ways to address the problem of surpassing a rate limit. The general term for handling this issue is called “traffic shaping.” The ISP can do one of the following to handle rate limit exceptions.
- Disable surpassing rate limit by enforcing a hard cap
- Charge a fee for traffic above the rate limit (say $100/Gigabyte
- Determine the average rate and charge a monthly fee based on
These are only three examples. There are many other ways to address traffic shaping which are content-type-neutral.
It is my opinion, as a small ISP operator, that the types of content-specific traffic shaping being discussed are unnecessary and will cause more harm than good. A more general, network-neutral approach to addressing network saturation should be considered.