PREVIOUS	Jan. - Feb. 2005 (13.10-11)	NEXT

Customer Owned Networks Emerging

Mesh Wireless and Open Access Fiber Architectures Laying the Ground Work - Developments are New Termites Gnawing at LEC Foundations

Security Issues Become Worse - Enterprises Trapped Between More Destructive Attacks and Need to Use Public Net To Do Business

How to purchase this issue. $400 or $1600 group.

Mesh Wireless as Infrastructure

New technology and ways of doing business are growing while the old have not yet died. Consequently telecom is still stuck in a difficult-to-map transition window.

Some trends - that we see more strongly in the last two months than before - are the continued strong growth of wireless capabilities with mesh 802.11b networks. Entire cities are beginning to contract for such networks. They range in size from Taipei (Nortel) to Cerritos California, with Philadelphia having installed some mesh but not yet contracted for the whole city (Tropos). While the Cerritos network apparently works well, we don't yet know how well a multi-thousand-node mesh will work over an entire metropolitan area. The consensus seems to be that it WILL work - if not immediately - with some more tweaking.

This is significant because it now looks possible to replace wireline metro architecture with wireless. If so, we have another huge termite gnawing away at the foundations of the LEC. Such metro mesh networks, however, are still walled-gardens in that, if they do VoIP, they rely on gateways into the PSTN and they must backhaul to the Internet to be really useful. Nevertheless, the movement is there and it seems to add more force to the idea that advanced telecommunications is another infrastructure to be supplied by local government. More and more, it is seen to be like highways, sewers and the electric grid. Namely it is something necessary to make any local environment suitable for commerce.

There are however uncertainties. Among them is the business model. Is access to the networks to be given away or charged for? Coffee shops give it away as an amenity for their customers. Cities, having many many expenses, wish to charge. How this will shake out remains to be seen. As cities, liking mesh architecture, begin to sign up, LECs will use all their political muscle at the state level to shoot them down.

In this context more than 30 members of our on going symposium debate the issues with several wireless interviewees -- Nigel Ballard on Wi-Fi hotel networks, and Wi-Fi coffee houses; Dave Hughes and Simon Buettrich on ad hoc mesh networks in Berlin and the Djursland peninsula of Denmark, and Ron Sege CEO of Tropos on his company's commercial mesh network hardware and protocol.

Security

Another looming issue is security on the public Internet. The second half of our symposium discussion centers on wireless and security - primarily security for enterprises in both wireless and wireline. We have a vigorous discussion between VJ Gill, Tony Li, David P Reed, Melissa Davis, and Frank Coluccio on what is perceived to be a rapidly worsening problem for corporate enterprise networks. These enterprises are caught between the jaws of the dilemma of the need to escape the massive damage being done to their internal networks by a Microsoft OS enabled collection of worms and varied malware and the need to use the internet for a wide variety of communication with the general public and with suppliers.

Customer Owned Fiber Networks

Andrew Odlyzko correctly points out that the much-vaunted optical core will remain a small, very much commodity part of the overall equation. We see that new business growth is at the edges and increasingly is to be found in customer owned networks be they municipal or enterprise.

Malcom Matson's OPLAN, an open access fiber network sold as real estate with long term financing and without the expectation of large near term profits is explored as the foundation that Matson believes needs to underlie the higher level network architecture and to replace the guaranteed income of the Local Exchange Carriers.

In this context, the January symposium discussion centers on municipal or statewide fiber network architectures as seen from an economic and policy point of view. Obstacles being wrestled with include the difficulty of formulating a carrier broadband business model that will allow the carrier to survive. Projects like Utopia seem to be more sustainable long term because the networks are customer owned and controlled as opposed to being in the hands of a third party.

But there is no uniformity. Many state's school network needs are organized via school district purchases from the LEC. In such a case there is little leverage to be had. In Utah however a statewide consortium using e-rate funds is bringing gigabit Ethernet to every school in the state. See pp. 34-35 above. We also note that Provo Utah also has its own fiber network that is not a part of Utopia.

Post crash technology looks unable to support carriers that exist to run networks and sell bandwidth. As Pete Kruckenberg wrote: "The business model of the ISP/bandwidth provider boils down to this: figure out a way to deliver 100% more bandwidth every year without receiving more revenue (or only a fraction more). Which means that the ISP has to figure out how to cut costs (or decrease quality) by 50% every year. Forever."

After the 1984 breakup of AT&T, the changes in technology and economic conditions underlying the technology meant that telecommunications infrastructure could not longer function as a natural monopoly. Multiple networks were possible and the possibilities of what could be done proliferated faster than many could understand. Planning and implementing telecommunications infrastructure quickly became very complex because at every level of the protocol stack there were many options and at every level of geography there were different jurisdictions.

We went from a national monopoly to long distance companies plus local monopolies. But within this context a bank, a drug company, a steel mill - any business in fact and soon any government - any enterprise could buy and interconnect its own telecommunications infrastructure into the deregulated national fabric.

As Kruckenberg points out "An enterprise can justify building their own network for at least two reasons: the service quality of the best-effort IP network is not good enough, and they can build/operate their own network (or so they think) for less than a better-effort IP service (if it were available) would cost. Or they can stay with public-cloud ATM/FR/MPLS because they don't have bandwidth enough to justify building their own network."

The enterprise view is generally more short term and results oriented than the local government point of view. There the watchword is economic development. Network as infrastructure. Again however the sophistication required of local government is great. Will a city build the network and run it? Or will it contract the running to a third party? Will it offer content as well as services? How will it choose technology and what architectural and operational limits will it impose? Utah and Iowa both have different approaches to operations. Various entrepreneurs are trying to sell municipalities on networks with fiber or wireless or some combination of the two at the same time as the incumbent phone companies screaming unfair competition try to get state legislatures to prohibit local governments from building their own telecom systems.

No Agreed on Model for Network Ownership

If the vertically integrated monopoly of service and content is the most outmoded and old fashioned, the other extreme is represented by the creation of a network infrastructure as an open access utility. Like sewers and the electric grid the city provides the fiber and support structures on the basic of a long term investment in its future, open to those who want to compete in the offering of services. One of the most fully worked out proposals was a June 2004 paper by British entrepreneur Malcolm Matson.

Matson is advocating 'community' or 'location-centric' public local open access telecom transport networks (OPLANs) where the primary 'value' in such networks, remains with those using it, rather than the owners. He advocates the structuring and funding of OPLANs as commercial real-estate - with passive, portfolio institutional investors providing the capital in exchange for a relatively low and safe financial return. As such his ideas fit within the realm of the network as a 'utility ' infrastructure where ownership of content and the network that delivers the content are rigorously separated. Under such a scheme the OPLAN becomes a fourth utility. (The highway grid, electric grid, water and sewer grids and a telecom grid.)

His ideas fall within the general realm of the municipal networks of which Jim Baller and others are such forceful advocates, and the architecture advocated by, among others, Francois Menard, and Matt Wegner. PacketFront and World Wide Packets make equipment designed to facilitate the ability of municipalities or other communities running their own networks. Increasingly it would seem that that the municipal utility approach is the only alternative to getting broadBand from the telco or cable co. Moreover it is one that makes the most sense, assuming that the network is looked at from the point of view of supporting local economic activity rather than as a service with a payback intended to sustain stock holder investment.

Malcolm's June 2004 paper describing his open public local access network (OPLAN) is the most thorough elaboration of these ideas that I have happened upon. It is presented as a side bar of pages 37-38 of this issue. There does not yet seem to be any generally agreed up business model for locally owned networks other than the one being developed by Malcolm's company, OpenPlanet Ltd. To his credit Malcolm goes into these issues in more detail than I have seen elsewhere.

[snip]

Security Again

Meanwhile for enterprises especially in the last months of 2004 the security environment grew critical. Melissa Davis reported her recent experience: "One of the discussions turned to the topic of the Intel proposal for building a new Internet based on 21st century requirements and standards. There was the consensus in the room, no minority from the working task groups that the present Internet is "too far gone." The phrases "gang warfare", metaphors of "urban street culture", and "Mad Max Culture" were used.

My business these days is building corporate and government "Walled Gardens" with dramatically collapsed Internet Gateways, elimination of rogue gateways, concentration of heavily fortified DMZ's. These are a dead-weight cost on business, and one of my customers logs 10,000 intrusion hits a day. We have reduced actual intrusions from the outside to .001 actual intrusions.

So, this is a technology and economic issue to me, and one that goes to the heart of "connectivity for the masses." We have drivers' license laws, drunk driver laws, laws for truckers to restrict fatigue/time on the road, aggressive driving laws, speed limits. There are the fatality issues, but also the pervasive upward pressure on auto insurance that affects everyone and drives up costs. We have gun laws, public health quarantine laws.

Hence, the looming political regulation threat is there. I am not full of answers, but trying to find another way, a technological way that is economically feasible.

This is an issue that should appeal to both libertarians and utilitarians."

Will the Users Interconnect without Carriers?

While a fear of chaos may impel some centralized movement on security issues, centrepital forces impelling action and innovation at the edges are still dominant. The technology of User controlled light paths developed by Canarie has matured very greatly in the last four years. This is a highly disruptive technology that for the time being works only in walled gardens apart from the general internet. By design it is predicated to assess whether user owned and operated networks can successfully interconnect and operate without carriers in the middle. I interviewed Bill St Arnaud for 75 minutes on this subject on November 11. The private COOK Report network economics mail list has already undertaken some discussion of this subject which will be the focal point of the March 2005 issue.

Contents

Network Ownership Grows at the Edges End Users Where Possible Cope with Uncertainty by Rolling Their Own Networks

Telecom & Information Technology Are in Difficult to Map Transition Window p. 1

Interviews on Wireless Issues

Wi-Fi Business Models Range from Installation in Hotel Chains to Free Coffee Shops :

Nigel Ballard Explains Criteria for Major Hotel Chains and Describes Personal Telco Economics - in Both Cases Free p. 8

Wireless Community Networks in Europe

Dave Hughes Describes the Wi-Fi Community Network on the Djurlands Peninsula of Denmark and Meshcube Radios of Freifunk in Berlin Highlights p. 15

Symposium Discussion September 12 - October 12

Looking at Open Access Fiber Architectures; QoS Evolution in Context of What Gives Networks Value;

OPLAN as an Economic Model of Open Access, and Wireless in Community Network Evolution p. 22

What Will the Edge Pay For? Not a Lot p. 45
Some More Thoughts from Andrew Odlyzko - The End of Telecom? p. 46
Who Will Capture VoIP? IXC Alliance with CableCos for VoIP? p. 47

Internet Ethics in the Context of the Capitalist Jungle p. 49
Some Architecture Issues p. 50
What is the Right Thing for a Service Provider to Do? p 51

Expand Network Availability Without Restricting Services p.51
'Mutuality of Interests' Yielded Wi-Fi Standards p. 52
The Bit Rate Limiter p. 53
Dealing With Metcalfe's Law p. 55
Open Access Remarks on OPEN ACCESS from the Media Access Provider POV (point-of-view) p. 56

Business Values Are Based on a Different Matrix or Axiology p. 57
Metcalfe's Law - Is Universal Access Really a Desirable Goal? P. 58
Economic Utility of Additional Connectivity p. 60
VoIP as a Pot of Gold? p. 61
Metro 802.11/16 Networks p. 63

Interview

Tropos Networks Uses Hardened Mesh Wi-fi for City-wide Infrastructure

Interview with CEO Ron Sege Describes Mesh Routing Algorithm and Economics of the Technology p. 66

Symposium Discussion
October 11- November 9

Bandwidth Market p. 73
Fixed Bandwidth Contracts Do Not Permit Efficient Planning p. 74
Sebastian Buettrich of Freifunk and Djursland Wireless Community Nets p. 76
Wireless Urban Infrastructure in Search of a Business Model p. 76
Comments and Elaboration on the Berlin Meshcube Radios p.77
WiMax - Hype or Not? p. 81
Cognitive Radio Brains Organize p. 82
802.22 (Cognitive Radio) Deemed a Non Viable Competitor to WiMAX p. 83
Economics and Security of Wireless Architectures p. 84
Regulatory Issues p. 84

Security

The Security Issue - Making Mincemeat of WPA p. 85
General Network Security p. 87
Defending the Routers Against SYN and ICMP p. 89
Back to Wi-fi Security p. 89
Security - Can the Startups Beat Cisco and Lucent? p. 91
Machines Can be Made Secure - Humans Are Another Matter p. 93
The Typical Malware Trojan and Protection Tools p. 94
Security and Hot Spot Authentication p. 96
Who's Responsible for Security Failure? p. 98
Economic Penalties of Lax Security p. 99
Is the Internet Too Far Gone? p100
How and Where Do You Defend? Or Cast Blame? p101

Security Education and Other Human Factors Issues p102
Is It So Bad that We Must Abandon and then Replace the Current Internet? p103
Does Business Try to Abandon Internet I? p105
How Vulnerable is the Internet as a Whole to a Zombie DDOS? p107
Are There Systemic Design Weaknesses in Public Internet Operation p108
Security of the Operating System and Browser p.109

Highlights from Interviews and Symposium p.115
Executive Summary p.138

Side Bars

Symposium and Interview Contributors to this Issue

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

Rich Bader, CEO Easystreet Wireless friendly ISP in Portland Oregon
Nigel Ballard, Director Wireless Matrix Networks, Portland Oregon
Sebastian Beuttrich co-founder http://wire.less.dk, wireless tech for developing countries
Frank Coluccio, President DTI Consulting -designer of optical networks New York City
Melissa Davis, optical network architect Cisco -now with RS Information Systems
Peter Ecclesine, market analyst for wireless, Cisco
Jim Forster, Distinguished Engineer, Cisco
Vijay Gill, Director Peering America on Line
Steve Heap, CTO of Arbinet a bandwidth broker and VoIP Traffic terminator,
Dave Hughes, owner Old Colorado City Communications- no license wireless advocate
David Hughes, security expert
Pete Kruckenberg, Architect Utah Education Network
Patrick Leary, evangelist for Alvarion and winner of numerous wireless industry awards
Jon Lebkowsky, Principal Polycot Consulting, Austin Texas
Martin Levy, former VP Engineering Concentric, now independent consultant
Tony Li, Router Architect at Cisco, Juniper and Procket, currently consultant
Malcom Matson, British entrepreur and author of OPLAN
Francois Menard, Canadian policy expert and municipal fiber network architect
Andrew Odlyzko, Director Digital Technology Center, University of Minnesota
David Reed, internet pioneer, spectrum policy and wireless advocate, currently with H P
Larry Roberts, Arpanet pioneer and founder Caspian Networks and Anagran
Jere Retzer Sr Mgr, Next Generation Networks, Oregon Health & Science University
Chris Savage, telecom attorney and partner at Cole, Raywid & Braverman in Wash. DC
Ron Sege, CEO, Tropos Networks
Dave Siegel, VP Network Architecture & Long Range Planning, Global Crossing
Jim Southworth, former chair DSL forum, VP Concentric. Now Secure Pathways CEO
Paul Stapleton, Editor-Publisher of I$P HO$TING Report: The Financial Newsletter for ISPs and Hosters
Bill St Arnaud, Director Ca*Net4 , Canarie Canada
Jeff Sterling, Interconnected Associates, Bellevue, Washington
Ron Yokubaitis, Founder Texas Net and CEO Giganews
Darin Wayrynen, CTO GoodNet, Former VP Enginering Winstar, current ISP operator
Matt Wegner, Product Manager North America, Packet Front
Damien Wetzel, Network Consultant Paris, Formerly with Akamai and Internap