LEXIS-NEXIS® Congressional Universe-Document

Search Terms: telecommunications act of 1996, House or Senate or Joint

Document 275 of 383.

View Related Topics

JUNE 24, 1999, THURSDAY

SECTION: IN THE NEWS

LENGTH: 3755 words

HEADLINE: PREPARED STATEMENT OF
MR. RUSS DAGGATT
VICE CHAIRMAN
TELEDESIC
BEFORE THE HOUSE COMMERCE COMMITTEE
TELECOMMUNICATIONS, TRADE AND CONSUMER PROTECTION SUBCOMMITTEE

BODY:

Teledesic 1445 120th Avenue, NE Bellevue, WA 98005
Thank you Mr. Chairman and Members of the Subcommittee. It is a pleasure and an honor to be here to speak to you today. My name is Russell Daggatt, and I am the Vice-Chairman of Teledesic LLC. At Teledesic, we are in the process of building a satellite network that will provide people in every part of this country and the world with affordable access to broadband communications services.
As this Committee and other organs of the government consider how to promote advanced telecommunications, it is of utmost importance that you continue to support universal access by all Americans, as well as the new technologies that promote it. I want to emphasize to you that this is not just a matter of economic or regulatory significance, but it is of profound social import as well. When I first joined Teledesic over 5 years ago, it was necessary to explain not only why broadband communications was important, but also what it was. At that time we described our system as an "Internet in the Sky" and our service proposition as "global, broadband Internet access." As difficult as it may be to recall now, five years ago the World Wide Web had not yet made its presence felt and the Internet had not emerged as the consensus network model. This was before Netscape was started - before there were any commercial Web browsers. Since then, various different notions of "broadband" have been put forth. We had to endure the "video on demand" period, followed by the "push technology" craze. Five years later the World Wide Web has become a daily part of most of our lives and the Internet an increasing necessity.
As evidenced by the plethora of different companies and technologies represented on the panel here today, many of which didn't even exist just 5 years ago, clearly there is no shortage of interest in providing broadband communications. Fiber optics, coax, copper, terrestrial wireless and satellites are all required to serve the global insatiable demand for bandwidth, but it is important that the government polices continue to support their development.
Access to capital is indisputable element to the current global broadband build-out. The Telecommunications Act of 1996 established the regulatory certainty needed by the capital markets to fund an unprecedented number of new competitive start-up telecommunications providers. Due in large part to the ground-rules established by the Act, in the US alone Wall Street investors have committed with tens of billions of dollars for competitive infrastructure. Therefore, it critical that Congress not take any action that could upset the capital markets that are providing the investments necessary to bring broadband services to all Americans.
It is not an exaggeration to say that building the infrastructure to provide broadband Internet access globally is the single biggest business opportunity on the planet over the next few decades. In most of the world, no telecommunications infrastructure exists at all. Where such infrastructure does exist, it consists largely of 100-year- old technology -- twisted-pair copper wires with a circuit-switched architecture. Even in relatively developed countries, when it comes to a telecommunications infrastructure optimized for networking computers, we're virtually starting from scratch. Estimates of the amount that will be invested in telecommunications infrastructure globally over the next decade start at around $2 trillion and go up from there. Although the Teledesic network will only be one star in the global constellation of services required to meet the demand, it will enable a unique capability to provide broadband Internet access to all those areas of the world that would not be economic to serve by other means. This would include rural Virginia and Louisiana as well as remote parts of Africa.
I would like to encourage the Committee to continue to support the development of competitive communications infrastructure. The industry is currently at a critical juncture in meeting the challenge set forth by Congress of bringing advanced communications capability to all the people of the United States. All we need to do is keep the trains on the track.
Thank you Mr. Chairman and Members of the Subcommittee. It is a pleasure and an honor to be here to speak to you today. My name is Russell Daggatt, and I am the Vice-Chairman of Teledesic LLC. At Teledesic, we are in the process of building a satellite network that will provide people in every part of this country and the world with affordable access to broadband communications services.
As this Committee and other organs of the government consider how to promote the development of advanced telecommunications, it is of utmost importance that you continue to support the goal of universal access by all Americans, as well as the new technologies that will make this universal access a reality. I want to emphasize to you that this is not just a matter of economic or regulatory significance, but it is of profound social import as well.
When I first joined Teledesic over 5 years ago, it was necessary to explain not only why broadband communication was important, but also what it was. At that time we described our system as an "Internet in the Sky" and our service proposition as "global, broadband Internet access." As difficult as it may be to recall now, five years ago the World Wide Web had not yet made its presence felt and the Internet had not emerged as the consensus network model. This was before Netscape was started - before there were any commercial Web browsers. Since then, various different notions of "broadband" have been put forth. We had to endure the "video on demand" period, followed by the "push technology" craze. Five years later the World Wide Web has become a daily part of most of our lives and the Internet an increasing necessity for things we associate with a high standard of living-from education and health care to economic development and public services.
As evidenced by the plethora of different companies and technologies represented on the panel here today, many of which didn't even exist just 5 years ago, clearly there is no shortage of interest in providing broadband communications. Fiber optics, coax, copper, terrestrial wireless and satellites will all play a role in serving the insatiable demand for bandwidth.
When trying to understand which technologies will be most efficient for servicing which needs, it is important to understand that in the traditional circuit-switched telecommunications model, you can break the network out into "access" or end-user connections and "transport" or backbone elements. The two elements have very different economics. In the Internet model, a third major element comes into play -- "quality-of-service" -- which sort of summarizes the whole. It is important to understand all three in comparing the economics of a wireline technology like optic fiber with a wireless access technology like Teledesic. The capabilities of optic fiber are truly amazing and growing more so every day. Optic fiber is certainly in the "miracle technology" category. In point-to-point applications, the economics of fiber absolutely overwhelm any other technology.

In the "transport" networks, the cost per bit of a loaded system (including all the up-front, fixed costs) will be very low, nearly infinitesimal. Unfortunately, the challenge is in extending broadband to the access networks, to make this technology available directly to end-users. infinitesimal. For this reason, distance will largely disappear as a pricing criterion in telecommunications (putting aside legacy regulatory distortions to the market). In the transport network, fiber dominates.
Unfortunately, the challenge is in extending broadband to the access networks (point of end-user interface). In the traditional circuit- switched networks the rule of thumb was that, on average, about 80% of the network cost is in the access portion. But that only takes into account those areas that have access (which does not include the vast majority of the Earth's surface and the vast majority of the world's population). With the economics of fiber coming to dominate the transport networks, with packet networks replacing circuit networks, and with the ubiquity of access increasing (or, more accurately, the lack of ubiquity decreasing), it is probably reasonable to assume for all relevant purposes that almost all of the network cost is in the access network (especially as the Internet model redefines "access". That is where the economics of wireline vs. wireless get more interesting.
The relative economics of wireline access technologies versus a wireless approach (including a satellite approach like Teledesic) are a function of both density and intensity of usage. The density part is pretty obvious. The cost to connect a customer with a wireline technology depends on the length the cable and the number of users it serves. That leaves most people and areas around the world unserved today. You can say that there is no demand for broadband telecommunications access in rural, remote and undeveloped or underdeveloped urban areas, but that is a bit circular in its logic. Any activity that requires an advanced information infrastructure today, almost by definition, must migrate out of those areas that don't have such an infrastructure. It's dictated by the economics of wireline access.
The intensity of usage determines the relative economics of access technologies. In the connections to most individual offices and homes, most of the capacity of a fiber connection would sit idle most of the time. The average residential subscriber in the U.S., for example, uses the phone for only something like 20 minutes a day. Internet applications are making usage patterns even more bursty and intermittent. It might be necessary to burst up to broadband speeds for only for a few seconds or a few minutes for a particular application, but the total number of megabytes sent and received over a day or week might still be very small. For example, I live in the heart of Seattle and have a DSL line. I work on my computer at home maybe 10 hours a week between evenings and weekends (pretty high usage). But the total number of megabytes I send and receive is pretty small (even though I want high speed when I do burst). With wireline technologies like fiber, all of that awesome capability must be rigidly dedicated to a particular end-user at a particular location, whether or not they need it at that moment. Given the very significant cost of extending fiber to individual offices and homes, i.e. using fiber as an "access" element, the cost per bit is most definitely not infinitesimal. Nor are the increases in the capabilities of fiber of much relevance -- even on a neighborhood level, the capacity of the fiber is not the limiting factor in the economics of its deployment. Even at Teledesic's headquarters, with 150 or so very data intensive users sending and receiving very big files, the company collectively bursts up to the full capacity of its T-1 line for a very few moments during an average day.
Wireless technologies, including satellites systems like Teledesic, that offer bandwidth-on-demand can provide a more economic access technology in a wide range of settings by dedicating only the bandwidth required by a particular application at a particular moment. Because demand for broadband services will generally be uneven and diffuse, it won't be possible to justify fiber buildout for most of the world's geography and the vast majority of its population. Even in the highly-developed urban areas, the early adopters who want a T-1 connection at their homes, for example, are likely to be fairly randomly distributed throughout each of those areas. On a neighborhood level, few areas will have the aggregate demand for two-way, interactive, broadband network connections that would justify the full area build-out that wireline technologies require. Where an existing coaxial cable or copper access infrastructure can be upgraded, the economics improve. But you don't see many (if any) overbuilds of those existing networks, which says something about the economics of the access network.
Nonetheless, it is fair to say that Teledesic is not likely to be the broadband access technology of choice for most users in developed urban areas. Which is just as well, because any satellite system is ultimately constrained in its "capacity density" -- that is, the amount of capacity it can focus in a given, concentrated area.
The Internet model introduces a third element to network economics, which I would argue, is the most important -- quality-of-service (QoS). In the traditional circuit network, QoS is not an issue. For each voice conversation, an end-to-end connection is established that is dedicated exclusively to that conversation (or data session). Of course, this assures a very high service quality, but it is also very inefficient. Packet-switched networks like the Internet, however, carry traffic from multiple sources that move over the same network infrastructure, making them up to 10 times more efficient than circuit networks. The economics of a packet network kill those of a circuit network. Because packet network traffic has to compete for network resources, network congestion becomes a big factor in how efficient packet networks can be. But in a packet network traffic has to compete for network resources. Network congestion becomes a big issue. The ability to establish and enforce priorities, latency guarantees and other service quality parameters becomes the distinguishing characteristic of a packet-switched network. In fact, with the Internet today, QoS issues are a bigger deal than bandwidth per se.
A critical point here: QoS is an end-to-end concept. It is not enough to provide QoS guarantees only part of the way to the destination, because the connection is only as strong (or fast) as its weakest link. It is also not enough to take the traffic from the end-user and dump it into the Internet cloud. Even fiber access is of limited value if it only connects to that Internet cloud (which, itself, includes abundant fiber). QoS is only meaningful as an end-to-end concept -- it all has to be tied together with enforceable service guarantees. As a result, in the Internet world, the concept of "access" changes. In the traditional voice world, access only requires a connection from the end-user to the nearest central office where a circuit connection can be established with any other circuit network. In the Internet world, that's not enough.
Teledesic is an end-to-end access network. Teledesic defines access in terms of the Internet model-in other words, in QoS terms. "Access" is the connection from the end-user to the nearest point of presence (PoP) that can provide the end-to-end QoS required by a particular application. In some cases, that might be only a kilometer. In other cases, it might be 1000 km ... or 5000 km. Let me explain this in more detail, because it is a critically important concept.
If you want to see where the applications are going to come from for the broadband networks of the future, look to where there are broadband networks today ... in the local area networks (LANs). What are the applications running over these LANs? Enterprise Resource Planning applications, SAP, Peoplesoft, SNI, BAAN, Oracle Financials, and the like. These are very demanding applications, particularly when it comes to latency. They were designed to run on LANs, not on the public Internet. Yet, increasingly, enterprises want to connect all their sites as well as their customers, suppliers, and the homes of their executives. They want to be able to run these enterprise applications not just at one isolated site, but everywhere to which their networks extend. This demands very high QoS guarantees from the network service providers.
UUNet (now part of MCI Worldcom) was the first major service provider to offer a product with a guaranteed maximum latency (of 150 ms). Others - Sprint, MCI, AT&T - soon followed with similar products. (Recent service level agreements I've seen have latency guarantees of 80 ms or less.) In every case, however, the service provider can only provide these guarantees where they actually control the network end- to-end. That is pretty limited availability, even for the largest service provider. MCI Worldcom, for example, only serves through its own facilities something like 40,000 sites worldwide.
This leads to another important point: It is not enough that there is some carrier nearby that might be able to provide a particular service guarantee. In a competitive world it matters very much whose network infrastructure is available. For example, Teledesic is headquartered in a suburb of Seattle. Let's say there is a USWest PoP a kilometer away from where Teledesic is located. That doesn't necessarily do Teledesic any good if it is a customer of, let's say, France Telecom.

If Teledesic is a customer of France Telecom its traffic might have to go 1000 km, to a France Telecom PoP in the San Francisco Bay Area, in order for France Telecom to be able to provide the necessary end- to-end QoS to the destination. In this example, "access" becomes 1000 km, not one kilometer. In other settings, in other parts of the world, "access" might be 5000 km, or more, in order to route around missing or problematic links or to connect into a unified infrastructure. Again, I would emphasize, it is not enough to take the traffic and dump it in the Internet cloud, or to hand it over to a competitor - or, more typically, to half a dozen carriers on the way to the destination.
It is not enough that there is a potential fiber connection to the customer. In a competitive world, it matters very much whose fiber it is. For example, running under the street in front of Teledesic's former headquarters in Kirkland, Washington, there are six fiber cables. Each one of those six cables could carry more than 100% of all the traffic moving on all six of those cables. So why six cables-- because competitors don't want to rely on the facilities of their competitors. (As an aside, it is interesting to note than none of the many condominium complexes running along that street connects into any of those fiber cables. The cost to slice into any one of those cables is about $50,000.)
Another example: Let's say that the Swedish national carrier, Telia, has the task of connecting all of Volvo's sites around the world, including an operation in Sao Paolo, Brazil. That Volvo site might be Telia's only customer in Sao Paolo. It doesn't make sense for Telia to build out facilities in Sao Paolo just to serve that one Volvo site. But it may be that the local service provider is affiliated with a competitor of Telia. Or the local service provider might be seeking exorbitant fees to provide the facilities for Telia. Or the local carrier itself simply might not have the facilities to provide service with the necessary QoS (this is most likely the case in most parts of the world today -- even in most urban areas).
Teledesic is an end-to-end network. Teledesic will be able to carry traffic from any point on Earth to any other point on Earth. That doesn't mean that Teledesic will carry the traffic end to end in every case - rather, that it can. Which means that Teledesic (or its partners) can provide end-to-end QoS guarantees to any customer from any place on Earth to any other place on Earth. Teledesic's satellite infrastructure will not be the preferred technology for all of a customer's sites in all cases (or even in most cases). But it is what enables the universal guarantee.
Teledesic has unique economics that enable the provision of broadband access (with end-to-end QoS guarantees) at a cost independent of location and independent of user density anywhere in the world (including maritime and aviation applications). That ability to aggregate diffuse demand globally creates a very robust business model that does not depend on the conditions of any single market. That can make even a $10 billion investment seem quite modest. Companies like Level 3 are spending comparable amounts just to undertake yet another fiber backbone overbuild in the U.S. The ability to aggregate diffuse demand globally makes Teledesic almost perfectly complementary to (rather than competitive with) fiber, which requires heavily aggregated demand (rather than diffuse demand) in a single point-to- point location to unleash its economic advantage.
It is not an exaggeration to say that building the infrastructure to provide broadband Internet access globally will be the single biggest business opportunity on the planet over the next few decades. In most of the world, no telecommunications infrastructure exists at all. Where such infrastructure does exist, it consists largely of 100-year- old technology -- twisted-pair copper wires with a circuit-switched architecture. Even in relatively developed countries, when it comes to a telecommunications infrastructure optimized for networking computers, we're virtually starting from scratch. Estimates of the amount that will be invested in telecommunications infrastructure globally over the next decade start at around $2 trillion and go up from there.
Access to capital is indisputable element to the current global broadband build-out. The Telecommunications Act of 1996 established the regulatory certainty needed by the capital markets to fund an unprecedented number of new competitive start-up telecommunications providers. Due in large part to the ground-rules established by the Act, in the US alone Wall Street investors have committed with tens of billions of dollars for competitive infrastructure. Therefore, it critical that Congress not take any action that could upset the capital markets that are providing the investments necessary to bring broadband services to all Americans.
Although the $10 billion Teledesic network will only be a drop in the ocean of global bandwidth required, it will enable a unique capability to provide broadband Internet access to all those areas of the world that would not be economic to serve by other means. And -- at least as important -- it will provide a competitive overbuild in all the areas that do have an existing broadband Internet access infrastructure.
END

LOAD-DATE: June 26, 1999

Document 275 of 383.


Search Terms: telecommunications act of 1996, House or Senate or Joint
To narrow your search, please enter a word or phrase: