From will they ever learn department, we are once again seeing attempts by incumbent carriers to skirt rules around network neutrality.

They tried and failed with UBB. Now they are at it again with “speed boost” technologies. The two technologies at question are Verizon’s “Turbo” service and Roger’s “SpeedBoost”. There are very few technical details, but it appears in the former case that users will be able to purchase additional instantaneous bandwidth to the detriment of other users on the same shared service. Whether this will make a difference to actual throughput is another matter because the slow video may be due to server problems and not network congestion. And if you are in elevator with very poor connectivity, you will unlikely get any faster download speed, no matter how many times you press the turbo button. But will Verizon give you a credit if you don’t get the advertised speed boost? I doubt it. Similarly the Rogers’ service, while still free, seems to imply faster speeds if they detect you are streaming a video, particularly from their own on-line service. Will users who are not streaming video, but using other real time applications get the same benefit such as VoIP or Telepresence? I doubt it.

The carriers continue to have this brain dead idea that bandwidth is a scarce resource—which is only true to the extent that were the ones who created this artificial scarcity. Building a business case around an artificial scarcity is as stupid as trying to make a premium market from air we breathe. Customers aren’t interested in buying bandwidth or quality of service to enhance their user experience. Just as with electricity they want and expect that just about any appliance or application will simply work—with no need for special speed boosts and other gimmicks. Imagine negotiating with the electric utility for a little extra power when you needed to turn on your stove or TV.

It is last mile packet loss which has the biggest impact on the customer’s user experience—NOT bandwidth or congestion. The Internet (TCP/IP) is designed so that packet loss is used as a signaling tool to reduce packet throughput. Regardless of where the packet loss occurs the Internet is designed to slow down any data stream, that is affected by a lost packet. However the rate to which a data stream is slowed down is greatly dependent on distance. This is why moving caching boxes as close as possible to the user affects end-to- end throughput, particularly if there is ongoing packet loss.

Although bandwidth and congestion can be a factors affecting packet loss, there are much more clever ways of reducing the impact of packet loss, especially in wireless environments. There are two much simpler solutions. The first is to locate caching/cloud servers as close as possible to the end users. Something that companies like Akamai and Google do already—at no charge to the carrier. Decreasing wireless distance from the wireless node is the other critical factor. This is why integrating WiFi with 3G/4G is so important.

A good example of a carrier that “gets it” is Free.FR in France. Free.FR is redefining what the idea of a carrier in the 21st century is, thanks to these innovations I have been talking about and pioneered by R&E networks like SURFnet. Integrating a blend of Wi-Fi, 3G and its all-fiber backbone, Free will offer unlimited voice, texting and data over the mobile networks. Free.fr deploys their own set-top box for automatically sharing a portion of one’s broadband connection via Wi-Fi with other Free.fr customers. Over five million set-top boxes means Free.fr has a free Wi-Fi cloud covering major cities such as Paris. Even when away from home, you can easily get broadband instead of resorting to an expensive 3G network. Their set top box will also allow extreme local caching, to further enhance the user mobile experience. This is the future of broadband. Not silly gimmicks like TurboBoost or SpeedBoost.