Print 

CAIDA and the Commons - an Idea Whose Time Has Come

Emphasis on Backbone as Neutral Infrastructure Yields National Muni-Network Cooperative Research Effort Designed to Enable Local Network Self Sufficiency

How to purchase this issue. $150 or $450 group.

February 23, 2007 Ewing, NJ -- The April issue contains an introduction to the Common's Project, and interview with Mark Spencer in which Mark explains the creation of Asterisk and Digium

Executive Summary

The Commons

pp. 1

From KC at the Cooperative Association for Internet Data Analysis and Sascha Meinrath of the Champaign Urban Community Wireless network comes an innovative project named COMMONS (Cooperative Measurement and Modeling of Open Networked Systems) initiative.

 

We report on the release of a final report from a workshop that CAIDA held on December 12 and 13. The workshop brought together leaders of community networks (mostly wireless) for discussion with network researchers and traffic specialists at CAIDA headquarters at the San Diego Supercomputer Center.

The outcome of the meeting is a proposal to use a wavelength on the National Lambda Rail research network as a transit backbone for muni wi-fi and fiber networks. The networks can do this and fit the research purposes of NLR, if they let researchers study their traffic for the purpose of things like understanding the resiliency of data flows and the ways in which video and other p-2-p traffic is affecting the network. The concept proposes to give the community owned networks a cost effective means of interconnecting with each other and, at the same time, it proposes to assist the national commercial networks by removing traffic from their backbones and lowering the amount of capital they need.

The proposal treats the network as basic infrastructure to be run on behalf of the members of the co-op or Commons." Unlike the case with the msos and telcos where the costs as well as the traffic are very large the commons founders hope their new network will serve as a way to transparently ascertain what the true costs are of providing this service.

They have established a well thought out and brilliant program that removes what needs to become an infrastructural 5th utility like service from the proprietary control of service providers whose major interest is to use old tools to preserve scarcity. The program right now exists on a shoestring budget and is looking for funding that would enable a significant ramp up of membership.

They clearly can help communities to help each other. The kinds of data they seek to amass should be exceeding valuable but to ensuring a stable and resilient internet and to their ability to run efficient and cost effective service but the success of their out reach plans is not a given. It will depend on the cooperation they get which in turn will depend on funding and their ability to explain to communities why it is in their interest to participate.

Mark Spencer and Digium

pp. 16

 

In an interview Mark Spencer explains how growing up with personal computers has shaped his life. Using computers to create software became a way of shaping the environment for Mark and access to the Internet allowed him to get in on the ground floor of Linux. About to turn 30 he is a member of the first generation of internet users to have always had the capability of using their own tools to shape their world in ways that their elders could not.

While studying for his BA at Auburn University he participated in a co-op work program and choose ADTRAN, a make of telephony hardware over Motorola because ADTRAN was less regimented. ADTARN was open to open Source and Linux in a way that Motorola was not. Mark was motivated by the desire to experiment and explore the power that the world of Linux based tools gave him. He had no master plan. When he decided he wanted to start a business offering Linux based support, he put together a small start up nest egg. When he realized he needed a sophisticated phone system, he did what he always did. Borrowing some hardware gear from ADTRAN, he hooked it up to his PC and wrote his own software to create his own PBX which he called Asterisk. In the Spring of 2000 he was on the verge of graduating from Auburn. It was also at this time that at the peak of the bubble when he was approached to sell his nascent business he attended a Red Hat sponsored meeting that introduced him to the Zapata telephony project. He went back to ADTRAN for advice and ADTRAN said don’t sell – we will invest.

Asterisk meanwhile was being rapidly adopted and supported in the open source world and by 2002 it was becoming clear that the sustainable business model was Asterisk support.

When I asked about sustainable business models, he said: “in the case of telecom there certainly were some logical business models. There were several logical factors behind it. It might be worthwhile to look at some of those factors.

First of all the telco market is absolutely huge. Much bigger than operating systems and data bases where open source has enjoyed a significant amount of success already. Second: the people who directly use Asterisk seem to be very technical people. What this means is that as a percentage of users, the number that can actually contribute is actually very high. So this means that the project is more likely to have success. Third there is a very serious cost difference. This is not just an Internet Explorer versus Firefox issue. These are real dollars. Fourth there is a need for extreme personalization and customization in telecom. People are very finicky about how their phones work.

They want their phone system to behave in precise and predictable ways. We have one enterprise customer with old key system installed through out their enterprises at 1000 different locations.”

 

I asked him: How then do the pieces fit together? There is the basic software and then the compatible open source modules designed to extend its capabilities? Then there is various hardware some of which probably runs the software just fine. Some of which does not. So is it a matter of knowing how to put all the pieces together and to advise a client on the trade offs between price capability and reliability as they make decisions on what they will select to build and then customize an asterisk based system that meets their needs? Digium in effect has people who can add the custom made clothing on top of the open source core?

Spencer: Yes. That is a good summary. The existence of these kinds of opportunities means that there is a business model there between the hardware and services needed to implement the software. Add testing to what you just said. Telecom is a case where all this is quite feasible because it is such a key business function that people are willing to spend money to make sure their phones work and work well.

One of the things that is very special about Asterisk as a technology is that it does talk the new voice over IP protocols. As well as all the “old school” protocols. That is it can do both new telephony and old telephony. We do not only H.323 but things like T1 and E1 emulation. PRI. GR303, analog phones.

You can use asterisk to talk many different protocols. You can use it with a variety of different technologies which is a very powerful feature in and of itself.

COOK Report: So one advantage of an open source product like this is that it can leverage its usage from the weakness of non interoperable proprietary systems.

Spencer: True. On one occasion we actually used Asterisk to bridge two Avaya H.323 products that otherwise could not talk to each other that. Could not talk to each other. They were different models of Avaya acquired as the result of the acquisition of another company.
A major story here is the can do generation represented by Mark Spencer. He and his fellow geeks are commoditizing many of the bread and butter products of Lucent, Nortel, Ericsson and Alcatel.

Symposium Discussion

FTTH in France p. 25

French Communities are now specifically allowed by French law to become telecommunications network operators, or even service operators if there is insufficient private initiative.” In Paris on December 20th the French Regulatory authority gave a workshop intended to help the municipalities get up to speed on their new legal abilities to shape their own digital future. According to Olivier Jerphagnon: The local authorities are worried about the competitive model. They fear that there will be 2-3 operators controlling market like for French mobile market and that prices will not go down as opposed to unbundled ADLS market that has been very successful in France in seeing triple-play offerings come early and prices go down dramatically.”

“What happens to CityNet may have a ripple effect on Europe. Also, Sweden and the Netherlands want France to be more vocal at Brussels (Germany and France are the two elephants). Finally ADSL growth is still strong in France because it is used to bring broadband to the rest of France (beyond major cities). Most fiber deployments from local authorities so far have been to help connect alternative carriers to France Telecom's historical central offices that manage and control the copper lines.”

Doc Searls comments favorably of the Danish company Indienet. http://indinet.dk, “a Danish netco that only addresses the question you ask above -- and answers it with managed pure-Net service connected to the best backhaul they can find. Sometimes the connection is just an ADSL line. Sometimes it's fiber. But they start at the edge (usually a co-op apartment building) and connect toward the middle (sometimes pulling up cobblestones and laying conduit under streets on a Sunday morning). Indienet.dk isn't getting rich, but they've been growing for the last five years and are part of a market ecosystem that has put Denmark at #1 in broadband penetration.

The ATT BellSouth Merger Shenanigans p. 29

The republican controlled FCC played the Democratic Commissioners for fools. ATT's lobbyists submitted a 20 page legal brief to the FCC by which ATT appeared to capitulate to its net neutrality critics and to promise to follow a regime of uninhibited net neutrality for two years after the merger's approval. The document was circulated among carefully selected consumer groups and the deliver to the FCC on Thursday December 28. Copps and Addelstein read it on the 28th and dropped their opposition to the merger claiming victory. It was made public on the evening of the 28th and by noon on Friday the 29th had been ripped apart by analysts like Dave Burstein. Too late - Copps and Addelstein had left town for the new years weekend. Martin and Tate put out their comments and did the same.

The joke was on the public as a whole as Martin and Tate basically said they thought ATT was stupid to have agreed to any kind of a compromise and that they would do their part to ensure that the the network neutrality provisions were not enforced. Scott McCollough points out that the final order is not yet released. Fred Goldstein summarized how Copps and Adelstein were played for fools in the final order of the Triennial Review leaving us with no expectations of anything other than a consumer hostile inefficient colossus.

Google & Amazon Offer Additional Business Models p. 34

Google has used a commercial version of G Mail to remotely offer some university campuses email accounts for their students an only a few days. It brings the application live in far less time than the local IT shop can. Further more in Manhatten Google is setting up a version of the same operation to server email to enterprises.

Amazon with its S3 remote storage and Elastic Computing cloud are making available storage and processing power on the internet that can be access and used remotely from anywhere on the globe.

According to Doctor Dobbs Portal “Amazon EC2 is a massive farm of Linux servers at your finger tips. It enables you to bring up a single server or many server instances -- all installed with pre-configured Linux images -- with a single command. You can use pre-packed Linux images provided by EC2 -- public images are currently available that include Apache and MySQL -- or build new images and upload them to Amazon's S3 storage service.” This in turn is leading to references like Calls in the Utility Computing Cloud” - Experiments in on demand ultra scalable telephony Using Amazon EC2 and S3.” Entrepreneur are moving in and one new start up is Digisence that offers SME off site back up using Amazon's S3.

Verizon Sells Rural NE Lines p. 41

In January Verizon sold off 1.5 million rural phone lines to New England based FairPoint Communications. In southern New Hampshire, where FairPoint will take over the high-speed FiOS fiber-optic network available to 80,000 households, customers who may have expected to see FiOS TV offerings won't likely see "triple-play" bundles of voice, television, and Internet in the near future. "Clearly, video will be a consideration, but we don't want to get distracted by that," as FairPoint takes over, Leach said. "We are going to increase high-speed data right out of the box."

Doc Searls commented “folks on the right and the left should both recognize high-speed Internet market demand as something that primarily arises from local businesses (including small and home offices) - and that this demand will only increase as more consumers become producers and look for high-speed connections that serve production as well as consumption. Everybody needs to see that serving this demand is good for the economy, for society, and for individuals. Next, state, local and regional governments need to create regulatory environments that are friendly to high-speed Net build-out by everybody, including individuals and businesses of all sizes.”

Joost p. 44

Dirk Wilhelm van Gulik: Right now - Joost simply turns your PC into an instant and interactive, on-demand TV. No set top box is required - as it is a mac/linux/win application. And it is quite simply a TV - full screen, the concept of endless channels, rather than search & down load. However by pausing and going back- or forward in time - it is also catch up TV. On top of that it mixes in the usual web 2.0 sundry to also allow search and tagging; or third parties to create their idea of interesting channels. So in short - nothing all too special.

What may be of interest is the fact that, while the EPG, search, advertising, command and control system is very traditional and central - the distribution of the raw (though encrypted) video fragments is not.

P2P is used for that. Based on the same foundations as Skype. This gives the company a different type of distribution economic. And also gives the network operators something of a change in expected bandwidth utilization. The reason for the latter is that, once you take care of the duplicates and re-runs across channels there is actually not all that much TV.

Cell Phone Woes p. 47

According to law professor Tim Wu's testimony: wireless carriers are "aggressively controlling product design and innovation in the equipment and application markets, to the detriment of consumers. Their policies, in the wired world, would be considered outrageous [and] in some cases illegal."

I write about Motorola's financial woes that are generally viewed as an artifact of its cell phone policy: Motorola's cell phones are essentially limited in the United States, to what, 4 customers now? In Europe to another 4 or 5? And in Asia to how many: ten or 15?

Talk about Soviet style planning! Gosplan - Gosudarstvennyie Plan - the State Planning Agency. For the last 25 years we have had the ITU as head of global systems standard regime. That system standards regime is put in place to keep Telco control over their networks and billing systems in place. Under the straight jacket of this global planning regime, in the United States there are essentially four ministries that tell Motorola what it must do if it is to be able to sell its phone to the contracting arm of the ministry - that is to say to the telco. Each ministry delivers Motorola a muti-volume encyclopedia size directory of specifications. Every phone, to be eligible for purchase by the state ministry, must meet way more than five thousand different requirements.

The physical components, firmware and software is planned right down to the pixels of the screen and the pantone color specified for the placement of the Ministry Logo. Each ministry specifies its own branding experience for the manufacturer. This goes all the way down to which web sites the phone is allowed to reach by the decision of the ministry contracting arm.

Use of the cell-phone as a device of mobile connectivity to each other and to the Internet and to its resources has exploded. More people have cell phones than PCs. Cell phones could be open and liberating. But they are not. What they can do is severely restricted by owners of the networks on which they operate.

How did this happen? In the 90s, as both the Internet and cell phones took off, the phone companies stepped in and bought on behalf of their customers. Buying in great bulk they enabled the device makers to use economy of manufacturing-scale to ramp up the feature set and push down the price. But the Cell Cos were no fools. They made what became a Faustian bargain with Nokia and Motorola and all the other large makers of handsets. Want a concentrated marketing channel? OK -no problem. We will buy in quantities of a million and later in ten million and more recently perhaps 50 million. If and only if you allow the phone to do very precisely what we require and nothing else.

The specifications imposed on the makers began early and grew quickly. The makers were lulled by the enormous numbers to be ordered and by the economies of scale that could be obtained at Asian manufacturing plants. So they bought into the program. They bulked up. The market took off. The carriers served as a bulk marketing channel buying in one blow for 30 million customers or with the “new” ATT for 80 million customers at a time. This ability seduced not just Motorola, but Nokia and many other makers. But we are now at the point where exponential growth can be maintained only by commoditization of the product. A feature filled race to the bottom to sell an imprisoned 'locked” product will produce more debacles of increased sales, decreased product capability and laid off engineers. No wonder you begin to hear wistful talk about “open phones.”

Russian Telecom Long Distance DWDM Modernization p. 51

Russia is becoming the telecom bridge linking Asia Europe. The general press has not a clue.

In the wake of the recent cable cuts off Taiwan, traffic to Europe was restored in part by being sent from Hong Kong over a new DWDN network built by the Russian firm Transtelecom. The Transtelecom site says: “The role of Asian countries in the world economy requires the creation of a direct telecommunication path between Europe and Asia. JSC “TransTeleCom Company” in collaboration with the Mongolian carrier “Ulan-Bator railway” and the Chinese national telecommunication company China United Telecommunications Corp. (China Unicom) offers the short fiber-optic path which connects the European and Asian regions of the continent.”

“The project called ERMC (Europe - Russia - Mongolia - China) is a consolidation of digital networks of the leading carriers of Europe and Asia.”

At the same time Russian conglomerates, generally energy related, are moving into major investments in European telecom. We review a list of moves coving the period from late last summer to the present.

For the rest of the issue you will have to subscribe.

Contents

CAIDA and the Commons - an Idea Whose Time Has Come

Emphasis on Backbone as Neutral Infrastructure Yields National Muni-Network Cooperative Research Effort
Designed to Enable Local Network Self Sufficiency

The Final Report; Three Inter-related Problems p. 2
The Commons Will Help ATT Unclog its Pipes p. 2
Open Problems; Proposed Experiment p. 3
Broader impact: Economics, Regulations and Policy p. 4
Questions and Answers p. 5
Open Sourcing the Net p. 7
Infrastructure Issues & Community-Generated Research Topics p. 7

A Discussion of Some Commons Planning and Operational Issues

Criteria for Commons Membership p. 11
Important Papers on Public Ownership of Local Networks and
Preserving the Wealth Creating Capability of the Internet p. 15

Mark Spencer’s Edge Enabled Innovative
Journey and the Founding of Digium

Access to Internet Enables Experimentation, New Business
Models and Asterisk’s Disruptive Emergence

Just Do It and Learn p. 16
Access to Internet Opens Door to MIT and to Linux p. 17
Networking Washing Machines and Discovering ADTRAN p. 18
Starting a Company to Give Linux Support p. 19
ADTRAN Invests in Linux Support Which Quickly Becomes
Asterisk Support p. 20
Asterisk Uses, Business Models and Open Source p. 21
What Digium Does and What Asterisk Does p. 22
Open Source Tension p. 23

Symposium Discussion December 17 - February 17

Rest of World Goes Fiber While US
Recreates Pre Divestiture ATT

France Prepares to Put FTTH into High p. 25
Looking for the Appropriate Service Model – Denmark and
Amsterdam p. 27

The ATT BellSouth Merger Approval Slides Through
by Means of Net Neutrality Scam

FCC Should Not Have Capitulated to ATT’s Late Night
Goal Line Sneak p. 29
The Day the Internet Became Cable Television: Dec. 29, 2006 p. 31
Bait and Switch: Kevin Martin’s Last Laugh as He Negates
ATT’s Merger Concessions p. 31

Google Transforms G Mail into an Enterprise Application -- We Note Amazon’s Storage and Computing Infrastructure p. 34

Google as the Big Edge and Other Edge Innovation p. 37
Video Onslaught: Bandwidth Business Opportunities in
non Telco Controlled Infrastructure p. 39
Cable Operators to Spend $80 Billion Worldwide through
2012 to Expand Network Bandwidth p. 39

Verizon’s Sells It’s Rural New England Business as
Joost and Vermont as e-State Ramp Up

Watering the Net Roots p. 41
Indienet – The Danish NetCo as an Example of What Should
Happen in the US p. 42
Vermont and Joost Explained p. 44
Time Given to Open Source Support p. 46

Cell Phone Woes: Carriers as Bulk Buyers Very Carefully Restrict Functionality of Cell Phones to Maximize Revenue Extracted from Their Customers

Poor Motorola p. 47
How the Handset Makers Lost their
Independence and Became Appendages of the Carriers p.49
Pleasure of Early Growth and the Pain of Commoditization p. 50

Russia Uses Oil Revenues to Build DWDM Links from China to Europe and Take Ownership Stakes in European Telecom Infrastructure

Russia Building State of the Art Fiber Web to Link European and Asian Economies p. 51
Will Sweden Sell Telia to Russia? p. 52
KPN, Germany, and Russia - Communications and Energy in Making
of Global Strategy p. 53
Altimo Shops from London to Indonesia – Kiril Babaev Always
on Stage p. 55
Sistema Wants a Share in Deutsche Telecom or Failing that in
Telecom Italia p. 56
Sviazinvest p. 57

Executive Summary p.58

Symposium & Interview Contributors to this Issue

Affiliation given for purposes of identification - views expressed are those of the contributors alone

Kevin Barron, Network Manager, Kavli Institute for Theoretical Physics, U. Cal Santa Barbara
Erik Cecil, Regulatory Attorney
Frank Coluccio, President DTI Consulting Inc., New York City
Vincent Dekker, Telecom Reporter Trouw, The Netherlands
James Enck, Eurotelco Blog and telecom Analyst Daiwa Securities, London
Tom Evslin, Novelist, Blogger and Retired Former CEO of ITXC
Fred Goldstein Principal of Ionary Consulting, author of The Great Telecom Meltdown
Lars Hedberg, Sceretary General Swedish Association of Community Networks
Olivier Jerphagnon, Strategist Calient Networks
Scott McCollough, Texas Regulatory Attorney
Sascha Meinrath Executive Director CUWIN and community wireless expert
Bill St Arnaud, Director Ca*Net4, Canada’s high speed research network
Chris Savage Attorney CRB, Washington, DC
Doc Searls Editor Linux Journal and Doc’s IT Garage blog, ClueTrain Co-author
Paul Sijben, VOIP expert and consultant from the Netherlands
Mark Spencer, CTO Digium
Dirk Wilhem Van Gulik, CTO Joost
Dirk van der Woude Civil Servant Amsterdam and fiber expert
Tom Vest, Senior Analyst, Internet Economics & Policy CAIDA
Tom West, CEO National Lambda Rail
John Waclawsky, Chief Software Architect, Motorola
Damien Wetzel, Network Engineer, Paris France