A Practical Navigator for the Internet Economy

Exploring Googin's Six Equations of the Real Time Global Corporation, TinyOS, and Free Space Optics

From Wireless Fiber to an OS for Wireless Motes -- An Examination of a Growing Global Web of Interconnected Technology

How to purchase this issue. $200 single copy or 600 group.

Some Thoughts on Modularization, Interconnectedness and a Dearth of Innovation

After a thirty year burst of innovation digital computers and digital telecom are now commodities. Huge disruptive breakthroughs can no longer be seen. The technology will “improve.” It will get smaller. And in bits-per-dollar cheaper. But the huge changes are likely to move toward biology, bio-engineering and integration of digital and nano technology into living systems.

At a recent meeting we talked with Robert Lucky, a distinguished researcher who spent a large part of his career at Bell Laboratories. He was saying with dismay that he and many senior colleagues had been discussing whether there was any expectation of any new developments within telecom that might serve as an enabler for any of the industry innovation that marked the 1990s. The conclusion was that, if anything were out ‘there,’ they simply didn’t see it.

This is one symptom of commoditization. Another may be seen as publication of research. From roughly 1970 to 2000 research papers in electronics, IT and telecom had been published primarily by industrial research laboratories. Perhaps 80% from the labs to 20% coming from the universities. The trend has now reversed. Bell Labs hardly exists as it and many others have been pruned to enable its parent corporation’s bottom line. A sad tale. We don’t have enough wisdom to judge whether it was an inevitable one.


Aside from commoditization and pruning back of research, miniaturization has by no means run its course. The most clear trend that we can see is an explosion of growth in CPUs. The quest for computing power developed huge supercomputers. Progress in miniaturization helps to enable further commoditization. The next explosion will simply be in the sheer number of computers. They will come in all shapes and sizes and will surround us utterly.

To the extent that growth may be found it will be in the creation of the software and networks that will knit them together. They will indeed be knit. Having talked about the hardware design and networking of tiny wireless “motes,” we go on in this issue to examine the TinyOS that has been written to enable these motes to gather and process and transmit data.

It is modularized and open source. Development here will consist in part of writing hardware specific application specific modules to plug into the basic operating system. The usefulness will lie in the interconnectedness of these complex systems. The frontiers look to be systems and process design as well as the integration and the management of complex systems highly modularized that ‘know’ what they need to do.

Introduction pp. 1 -2

It’s the interconnectedness that matters. The really significant new technologies are likely in areas such as bio-engineering. Computing and telecom are becoming “faster” and “better” whatever “faster” and “better” means. The changes are incremental and the markets mature. The significance now is not so much in moving more bits to anyone particular desktop as it is in the business process changes being brought to a globe made up of interconnected desktops.

Interview with Roxane Googin pp. 3-9

I want to talk today about what I will metaphorically term six simultaneous equations. These are areas of technology where profound change is underway and where, when the processes are completed in a year of two, will drastically reshape global business. They are : 1) The semiconductor industry's move to 90 nanometer geometries,300mm wafers and copper interconnects using "low-k" dielectrics, 2) movement from direct attached to network attached storage, 3) circuit switched narrowband to IP broadband telecom, 4) client device evolution from PC desktops to ubiquitous wireless and handhelds, 5) software move from procedural 3G to object-oriented code, using real time web services, and 6) the triumph of blade servers and Linux over single-system-image servers and RISC/Unix or Windows. When these fall into place, we are likely to see a major productivity spurt as well as a major increase in the export of technology jobs to lower wage areas. These changes mean that we are headed toward a "real time corporation" where everything is integrated and everyone is working from the same data. . . . .

You become a "real time" organization by crystallizing business processes. These are the actions and tools that people use to perform their roles and understand what they are expected to do. You encode all this in software and also run your business on what is referred to as "one version of the truth." This is one horizontal, flat data layer that does not exist separately in every application. Rather it exists one point in time. It is shared by everyone and is always available. This is a radical transformation.

What is interesting is that to move from the old vertically oriented paradigm to the new horizontal real time corporation, all six of these transformations must take place. There is also a seven equation - job migration. With the six equations maturing in a way that will bring the real time corporations into existence, the result will be to move skilled technical labor out of the high wage West into the lower wage East. This will dampen economic growth for the US and Europe. But the rest of the world will likely be quite pleased.

These changes are underway. They are all happening now and, in all phases of information technology, they are the most traumatic that have ever been seen. Telecom is not alone. It is not the entire "elephant." The problems of telecom are shared across the global scope and scale of computing and information technology. In this scenario telecom becomes a computer backplane facilitating machine-to-machine communication. Business processes replace voice as revenue generators and all intelligence has to move to the edge. . . .

We all know that US capitalism beat Russian communism, even though both countries had similar geographies and human resources. Who knows why for sure, but one common belief is that our open economy was simply able to out-produce their top-down system. The same theory applies to business. For business competitive success, one key differentiator is the free flow of ideas. Similarly, on a global scale, the next winner is the society that can best manage its information flows. Corporation by corporation, we are in a race to see which companies can become best of breed in transitioning into real time global businesses. It does not necessarily mean that the US will remain number one, rather number one will be whomever understands these issues the best. . . .

So Who Wins?

COOK Report: The only thing left then is integration of the various processes and education customers as to how the tools and processes may be used to greatest advantage. And through outsourcing producing new technology at the cheapest possible price.

Googin: And there is labor at the cheapest possible price. Once you put a business process into a computer it becomes location independent and you can ship it off to India or wherever the labor portion of the work is done at the most effective price. Right?p> Who will win? Someone asked the question about Dell. Well when Toyota makes a car in Tennessee who wins? Tough question. But I will tell you if a country like the US has a screwed up regulatory scheme and lets the old companies give us crappy broadband at high rates, then we are going to be seen as incompetent and hard to deal with.

COOK Report; Because these process are so complex and computationally demanding? Global business process that have to go through American infrastructure will be undertaken at a disadvantage?

Googin: If all you have to work with is a T1 it is like sucking on a straw forced into a pin hole. The right hand is not going to know what the left hand is doing - ever. You will be seen as uncoordinated. Its like having a bad nervous system. Does that make sense? The Chinese are installing broadband. Everyone else is putting in their own big networks. So what I am saying is that on a global stage America cannot compete if it doesn't have its own.

Tiny OS pp. 10 -15

Editor: Without an operating system wireless sensor networks cannot function. We interview Jason Hill creator of Tiny OS. Hill: While the Berkely Motes had all the necessary physical components, they were missing the software that could enable them to gather data, process what was collected and transmit it elsewhere. This was the catalyst that started the Tiny OS research team.

To make the devices useful then we clearly needed an operating system (OS) and the first step in building one was to outline the characteristics that were needed for an OS in an environment such as this. The first step was to switch to an event-based operating environment where the entire system is based on events as they move through the node, instead of the traditional programming environment that runs multiple and simultaneous threads as on your PC, where you double click on the Outlook Icon and it starts an Outlook execution process. You could have a radio event where a message arrives, or a sensing event and your application begins to process those things in a serial fashion. ....

COOK Report: With one "ear" then you are listing to your environment and with the other you are deciding how to organize and what to do with that which you are hearing?

Hill: Correct. The event-based program is a mechanism for doing this. The processor is assigned the task of looking at all these event queues so that when a radio message comes in it grabs it and handles that event and does the processing on the message and then goes back to look for the next event. It can hop from sensor analysis to doing a bit transfer on the radio, and then to doing a quick update on its routing protocols to find out which is the best parent mote for some communication task. . . .

Event based programming then was one of the key design points that went into the development of Tiny OS. A second one is that it is highly modularized. Every piece of the system is broken up into what we call components where small functional blocks are used to handle these events and then issue commands to other components. For example if you wanted to build something like Great Duck Island, you would have a small component at the top that was there to understand what the different sensors meant at a very high level. In other words: "this is a temperature sensor and that is a pressure sensor and here are the valid ranges."

COOK Report: Does one component then listen to your sensors and another to your routing protocols and so on?

Hill: Exactly. In the simplest sense there is a component that understands the valid range of values for a sensor and understands how to calibrate it. It may be tied to a given version of a Honeywell magnetometer placed on a specific board as well as which IO lines were connected where. And then there is a higher-level component that does DSP filtering of a generic sensor reading. In this case it doesn't care whether it is sitting on top of a magnetometer or a light sensor or an accelerometer searching for earthquakes. . . .

Free Space Optics p. 18

Editor: For Broadband wireless back haul free space optics which is to say data delivered over laser pulses through the air is the undisputed broad band champion. Speeds up to a giagbit per second are in commercial use. When fog prohibits line of sight transmission, the system does an automatic fall over to millimeter wavelength radio.

We interview fSona Corporation one of the leaders in this area.

COOK Report: How did you acquire the technology you are using now?

Carbonneau: When I started this company, I approached British Telecom (BT) because they had an R&D program on free space optics transmission that included a prototype that they built. They had also done outdoor testing in London. They had used 1550 nanometer wavelength because they had assumed that the adsorption window at that wavelength would be superior to other wavelengths in foggy wet weather. The other reason for their choice of 1550 was that this wavelength was inherently eye safe. Therefore they could use higher power to transmit, since, at this wavelength, the laser did not penetrate tissue. Higher power would extend the range of the device. They felt that as a telco they could not risk the liability of being out in a public space with any wavelength of light that could potential damage the eye. . . .

COOK Report: What if you were trying to connect the edges of two metro areas? Can you create a free space optics pop where without a building you just interconnect a pair of your devices?

Carbonneau: Yes. This has been done in Belfast.

Corcoran: We can literally daisy chain our link heads back-to-back. We have single mode fiber that feeds them with 100% digital regeneration. There is no intermediate device needed. Theoretically you could string 100s of these together with the only concern being latency that is virtually undetectable and jitter that would be only a nominal factor. Latency and jitter is typically introduced only when there is an intermediate device that is typically used by microwave radios. In our case we don’t need this because we do our own digital regeneration.

Our customer’s do daisy chain for link budget purposes. Let’s assume you are down in Santa Monica where you do get west coast fog. There you will need repeaters every 700 meters to ensure five nines of connectivity uptime. Digital regeneration is also done for line-of-sight purposes. Lets assume that you are doing SONET protected ring and you need to turn a corner. You put one link head in and slam in a second redirected at 90 degrees and connect them with single mode fiber. Do this and you turn a corner. Internal digital regeneration of the signal gives us this kind of flexibility.

Advanced Optical Networks p. 25

In trying to take a really global view of where we are headed it has become difficult to stay on track with things that we have covered in past. The course being pursued by CANARIE with CA*Net4, by the StarLight optical exchange in Chicago and by Surfnet in Amsterdam is very important. These researchers are continuing to do amazing things with end-to-end user owned and controlled optical light paths. Give the amount of fallow fiber in existence they are getting very good bargains on the necessary fiber. As announcements on the CAnet-News mail list have shown during 2003 steady progress is now being made on the goals of end user control of lambdas that was one of the goals of the project when it was announced in 2001.

With the exception of Chicago most of this development is going on outside of the US. However some development is beginning to get some funding here. Oak Ridge National Laboratory has announced funding from Department of Energy for Science UltraNet.

ICANN, p. 27

Karl Auerbach: If you look at what ICANN really and truly does you will see that it has little, if any, real role relating to internet technology. Rather it is an organization that, for the most part, imposes the business goals of a selected and limited set of priviliged "stakeholders" onto the operation of businesses that sell domain names.

Moving ICANN from the blind-oversight of the US Deparment of Commerce to the UN or ITU ill only widen the stage for those privileged "stakeholders". A move to the UN or ITU, by itself, will not improve the security of the net or any nation.

Without major structural reforms (such as I suggest at http://www.cavebear.com/rw/apfi.htm) ICANN will remain a non-technical body that regulates and governs internet


Whereís the Technology Taking Us? - Thoughts on Interconnectedness and the Emergence of the Real Time Global Corporation p.1

An On-Going Revolution - Six Equations Driving the Global Evolution of Information Technology and Telecom Toward the Real Time Corporation and Migration of Jobs Offshore p.4

An Operating System for Wireless Sensor Networks - Jason Hill Describes the Origins and Future of Tiny OS p. 10

Intelís Tiny Hope for the Future p. 17

Free Space Optics -- An Exploration of the Technology and Economics of High Speed Wireless Backhaul p. 18

Advanced Optical Networks Make Significant Progress - A Quick Snapshot Shows Major Divergence from the Commercial TCP/IP Internet - Grids and Self Managing Systems are Near Term Influences p. 25

ICANN and Internet Governance - Karl Auerbach Debates Vint Cerf on ICANN’s Performance Under Threat from Verisign p. 27

Interview, Discussion, and Article Highlights p. 32

Executive Summary p. 39