Tuesday, June 19, 2007

Why new network business models and architectures are critical

[The following web sites point to some of the challenging problems facing users and application providers in trying to deliver innovative solutions over the current broadband last mile Internet architecture. Despite all the advances in Internet technology, Moore's law, routers, optical switches, etc the current Internet is still riding on an infrastructure that was designed over 100 years in the case of DSL/telephone and over 40 years ago in the case of cable. Not only is this tree and branch architecture outdated, the resulting business model is fundamentally shaped and distorted by the design assumption in that architecture.

The Internet architecture for the long haul and enterprise markets has been radically reshaped over the past decade because of rigorous competition. Many large enterprises multi-home to several competitive providers and also do their own direct remote peering because of the availability of dark fiber in most metro markets. As a result companies with new competitive business models largely dominate the enterprise Internet market such as Level3, Cogent, Equinix, etc.

In the last mile it is a different story.

Cablecos and Telcos are now wrestling with the data deluge of Internet video distribution over their networks whether P2P or HTTP. Many have quietly and surreptitiously implemented various policing and shaping mechanisms to limit the growth of this traffic. Although it is applied on a non-discriminatory basis in terms of the customer or the provider, most cablecos and telcos still don't understand the broader implications in terms of network neutrality and are surprised when the public takes umbrage at such tactics. This is especially true when they block VoIP over cell phones.

The final frontier is the last mile architecture to our neighborhoods and homes. I am pleased to see that Australia is taking a proactive step in this direction with their recent $AUS 2 billion broadband announcement and the formation of an open committee to specify the FTTN architecture.

I have always argued that university/research community needs to take a more proactive role in both the research and, more importantly, the deployment of alternative last mile architectures. GENI (and the European equivalent PAN) is a critical step on the technology research side. But we also need new examples of actual deployments and business models. I know of a couple universities that plan to build virtual 3G wireless networks so that students will be free to integrate WiFi and pico-cells with their regular cell phones and also be free to develop new applications and services without getting permission of the underlying wireless provider. A couple of institutions are also exploring innovative last mile broadband networks for their universities and dormitories, and be liberated from the bandwidth limiting tyranny of either the service provider or the campus CIO -- BSA]

From Dewayne Henrdick's list
From: Ken DiPietro

Will carriers spoil the online video party?

BBC story on the Internet is being overloaded http://news.bbc.co.uk/1/hi/programmes/click_online/6756899.stm

From Dewayne Hendricks list
Early adopters want Wi-Fi mobile phones, T-Mobile wants to kill VOIP


Research firm In-Stat reports that there’s a substantial market niche
for Wi-Fi enabled mobile phones:

A recent survey of US early adopters found that almost half of those
respondents plan to replace their cell phones want Wi-Fi capability.
To meet the growing demand, there is an avalanche of dual-mode phones
in the pipeline. By the end of this year, the Wi-Fi Alliance will
have certified more than 100 different models of Wi-Fi/cellular
phones . . . widespread Wi-Fi deployment and the variety of Wi-Fi/
cellular handsets offers Wi-Fi/Cellular based systems a significant
head-start in the market,” says Allen Nogee, In-Stat Principal
Analyst. “Other technologies, such as WiMAX and Ultra Wideband, are
also poised to enter the handset market, but Wi-Fi fills a unique
niche that WiMAX and UWB cannot match.”

An alternate architecture for university dormitory networks http://www.canarie.ca/canet4/library/customer/Customer_Owned_Network_for_university_dorm.ppt