The Problems With American Broadband

My father and I share a special relationship. We always had a good relationship, but while I was in college our time together became more meaningful. I don't recall exactly how we got started, but we began to go out to lunch once a week and just talk. At some point I discovered that I had a new close friend as well as a father.

We no longer have our one on one lunches, but we do see each other quite often. And, almost invariably we talk technology, much to the consternation of my daughter who thinks we should have other topics that are worthy of discussion. Technology seems right, though, as it was the Apple II he purchased for the family that put me on the path my professional career has followed for almost two decades. Incidentally, I thought the purchase one of the most wasteful and imprudent I had ever seen. Shows what I know.

Anyway, over a recent weekend he got our conversation going with what could be considered a loaded question. He had read an article proclaiming American broadband to be a colossal laggard as compared to that enjoyed by the Koreans and Japanese in particular. He asked why this was and what it meant for the United States.

The question is "loaded" as I, in my role as a Chief Technology Officer, for many years spent a significant part of my time and effort in fighting with the "phone company." Perhaps I should say "phone companies" as the one in particular to which I refer has in fact been four different companies during my tenure. While the names and logos may change, the actual company has remained remarkably the same, I am sorry to report. I'll save my tales of telecom woes for perhaps another column, or many columns as the case may be, but suffice it to say that what I experienced firsthand as a somewhat sophisticated customer of telco services did not leave me with a warm and fuzzy feeling. It was inconceivable to me that the incumbent Bell operating companies had any ability to be a driving part of a telecommunications-rich future. To make matters worse, it is clear that technological progress is outstripping our ability to craft enlightened public policy.

By "enlightened public policy" I essentially mean policy that is created for the public good. The "public good" is a concept that is heavily codified in this nation's earliest documents as well as a concept that has been lived through a significant part of our relatively young history. Sadly, I fear that the concept of "public good" is becoming a secondary citizen as the concept of "corporate good" ascends to first class status. This is probably a natural result of the rise of large and powerful modern enterprise, but it is a natural result that has not been adequately challenged and is certainly not in the best interests of the populace at large. Intellectual property rights is one of the principal issues of our day, though most people probably do not yet realize this truth.

The relevance to telecommunications policy is that policy-wise we are confused as to whether or not telecommunications, at least the wired part, is a public good or a corporate good. Of course, practically it has always been a hybrid, but the pendulum that stood relatively still for decades has had trouble making up its mind recently. The "public good" aspect of telecommunications has its roots in the concept of the "common carrier," which is a term that dates back to public policy governing the movement of goods and people in carriages and boats. The desirably of the common carrier for the public is that there is guaranteed access at standardized rates for the public at large to the carrier.

The debate between addressing the public good and the corporate good is easy to understand as it places our commitment to the public good in conflict with our commitment to the preservation of property rights. The communications providers argue that by virtue of their being the parties that invest the financial resources into the building of networks, they should have sole access to those networks. That is, ownership should allow them to keep other service providers off of their networks.

Being a common carrier is not something that communications companies find appealing. This is because, like their carriage operating ancestors, a higher level of regulation is imposed on a common carrier. Traditional telephone service has been determined to be subject to common carrier regulation, but thus far Internet service has been a murky area. The FCC heretofore has declared cable Internet access, for example, something different than traditional telephony and therefore not subject to common carrier regulation. And, herein lies a critical part of the problem.

The designation of telephone service providers as common carriers was an important part of telecommunications public policy that made good economic sense. The cost of wiring customers to a geographically dispersed network was, and still is, a prohibitively expensive undertaking. This is due not only to the wiring itself, but to legal issues such as getting rights of access to poles and other property. Thus, it made sense not only to designate the telephone companies as common carriers, but also to allow them to operate essentially without competition for telephone service. It was more important to get everyone connected to the network than it was to have competition due to the realistic fear that we could end up with lots of competitors pursuing the lucrative customers and none pursuing those that are difficult to reach.

Data networking, which is fundamentally what telecommunicatiosn is increasingly all about, has long been architected in levels or layers. Our beloved TCP/IP, for example, consists of four layers: data link and physical; network; transport; and, application. Data link and physical refers to the actual physical connection of a host (computer) to the network. The network layer defines how devices are logically addressed and found on a network and internetworks. "Logical" is the key term as multiple protocols can share the same wire, or physical layer, which has its own way of finding other machines on the network. Above that one has the transport layer, which manages connections and communications between hosts. Finally, the application layer is that which is presented to the user in terms of performing a specific type of function on a network.

You may have heard the terms "dumb network" and "smart network." Data networks have historically tended to be "dumb." That is, the network itself knows how to move data around, but it is not particularly knowledgeable about what it is moving. The "intelligence", so to speak, is contained in the hosts on the network. Thus, you may use your computer to surf the Web. Your browser knows what it is doing and the Web server knows what it is doing, but the underlying network simply acts as a conduit for the service. In a sense, the "dumb network", properly architected, as the Internet is, becomes a kind of highly adaptable "platform" for application development.

The issue presented here is the depth of the layer of the network that is presented to potential developers. In traditional telephony, both users and developers see the application layer; that is, the making and receiving of phone calls.

Try building applications on the telephone system, which is better known as the public switched telephone network, or PSTN. This is not as easy as the "intelligence" is built into network itself, making the network a rather closed system. The PSTN is great for making phone calls, but a non-starter for new applications, the kind of which we routinely see on the Internet.

All of which has largely been okay until recently. The "smart network" that is the PSTN has connected people via telephones with an almost incomprehensible level of reliability, at least in this country. Despite their other (many) failings, the incumbent Bell operating companies have a done a better of great job of delivering dependable telephone service to virtually anybody who wanted it. By comparison, the world of the Internet has been a reliability failure. While being pretty darned good, the Internet itself does not even come close to the reliability of the PSTN. In fact, and this is not a joke, my broadband connection is down at the moment while I can clearly hear my wife having a telephone conversation downstairs.

The problem for policy-makers is convergence. Voice has historically been primarily an analog affair and data obviously has been a digital affair. Everything but voice has always been all digital and voice is quickly becoming digital from end to end (voice has largely been digital for years, though only at the center). When voice is digital from end to end, it becomes much more efficient to deal with it the same as other data, provided that the voice data can be transmitted fast enough so that conversations are comprehendable. Your patience is having to wait for your grandmother's next word is going to be considerably slower than pulling up that page from E-Bay. It is efficiency as in "cost" efficiency that is the key here. Data networks, like the Internet, can move data much, much more cost-efficiently than the PSTN. The problem for policy-makers is that this convergence is incongruent with the public policies that have long regulated telephone service and the PSTN, which is subject to the concept of "common carrier." Public policy on telecommunications needs to be reconsidered in light of the changes already happening in terms of how we build communications networks.

While we founder in crafting better policy in this time of massive change, the key to success continues to lie in the concept of "common carrier" and the application of this important concept to the new types of networks that are springing up. We know that these networks will be built on a variety of technologies, but we know clearly that they'll all be digital networks. Consequently, we know that they'll be built within the confines of the layered data networking model where the physical manifestation of the network will be separate from the logical manifestation. With this division being as clear as it is, we have the perfect opportunity to apply the concept of "common carrier."