March 29, 2011

Paul Baran (1926-2011)

A quiet giant died Saturday. The fact that you are reading this can be sourced to Paul Baran’s breakthrough concept of what became known as packet switching. Baran’s idea of a distributed digital network became the backbone design of the networks that define today and are redirecting the future.

During the Cold War U.S. defense planners became concerned that the central switching of the telephone network was vulnerable in the event of a Soviet attack. Because the message to launch an American counterstrike used the telephone network, an enemy first strike could easily destroy the critical switches necessary for a launch message to reach its destination. The Defense Department commissioned Rand Corporation, a California research institution, to come up with a solution.

In 1964 Baran, then a young Rand engineer, published “On Distributed Communications.” His idea was that instead of an end-to-end circuit that could be broken by destroying a switching point, the launch message should be digitized, broken into multiple pieces (“packets”), each containing digital instructions as to its destination, and routed through a hub-less network. Should an attack take out one link the network would simply reform using connections that bypassed the problem. The fishnet-like digital network of today was born.

Baran’s distributed network became the cornerstone of the doctrine of Mutually Assured Destruction (MAD) that by keeping the world on the brink of nuclear destruction successfully avoided annihilation. The ability of the U.S. command and control network to survive a preemptive strike and respond in force was so important that the U.S. government openly shared the details of the new network with the Soviet government so they could understand the folly of a first strike.

Paul Baran overturned the common knowledge about networks that had existed since hunter-gatherers had followed the paths of their prey. Networks had always been a point-to-point open pathway. The new network design was especially an anathema to those who ran traditional networks. The first-strike risk that Baran’s plan overcame was the consequence of the telephone company’s centralized architecture. Yet AT&T, the monopoly long distance provider of the time, refused the Defense Department’s request to solve that problem by building a distributed network. In the end the government built the network itself; after multiple iterations it became ARPANET and ultimately the model for the Internet.

I had the privilege of discussing the history of those days of discovery with Paul Baran many times. He would laugh as he recalled the challenge of trying to explain the distributed network concept to engineers at the telephone company. He said it was as if he was “speaking Swahili.” The engineers at the telephone company had been raised in an analog world and only thought in terms of setting up and taking down end-to-end circuits. The idea that connections could be made without maintaining a constant open circuit was beyond the scope of their imagination. The idea that messages would be sent by constantly establishing and then destroying a connection was incomprehensible. At one point in the effort to try and get AT&T to build the new network it was necessary to enlist the scientists at Bell Laboratories who were doing pioneering work in digital technology (but who understood analog) to act as interpreters so the analog engineers at AT&T could understand the concepts Baran was advancing.

One of the most endearing qualities of this giant of a man was his modesty. “The Internet is really the work of a thousand people,” he once explained. He likened the developments that followed his as like “building a cathedral.” “Over the course of several hundred years, new people come along and each lays down a block on top of the old foundations, each saying, ‘I built a cathedral.’” Paul Baran did not invent the Internet, but his discovery did enable what became the Internet. If we are to follow his cathedral metaphor, someone had to lay the cornerstone.

It was my privilege to call Paul Baran a friend. Imagine the unique honor of learning about digital networks from the man who developed the concept that made them possible. Consider the marvel of sitting with Paul when he would say, “I’ve got a new idea” (he started seven companies, five of which went public). He was a quiet man in his demeanor, but oh so loud in what he accomplished and contributed. We stand on the precipice of great new things because what Paul Baran left us.

March 1, 2011

Shooting Behind the Ducks

“Europe Loses the Mobile Phone War” the front cover of The Economist declared the same week as 60,000 people converged on Barcelona for the annual Mobile World Congress.

The root cause of this development, the magazine observed, was that the mobile telephone had become a hand-held computer. Whereas telephones were all about networks and devices, portable computing is all about software and services. “This is where America, in particular Silicon Valley, is hard to beat,” the magazine concluded. “Companies like Apple and Google know how to build overarching technology platforms.”

The movement from a telephone to a computer has brought to the wireless carriers a double-dose of the earlier experience of their wireline brethren. Like the wireline carriers who chafed at their lines being used to enable high-margin Internet businesses, wireless carriers are channeling then-AT&T CEO Ed Whitacre’s famous 2005 comment, “For anybody to expect to use these pipes for free is nuts.” The double-dose that hits wireless carriers, however, is how the flood of new wireless apps (particularly video) over which they have no control is forcing them to spend billions of dollars on new “pipes” in order to keep their existing customers.

Bloomberg Business Week warned in December the relationship between wireless carriers and Internet companies “could turn into a cold war.” Certainly the buzz around Barcelona was all about whether the cold war would escalate to a hot one. European carriers, for instance, were calling on Internet companies to step up and pay for the infrastructure upon which they rely and on which their products are placing ever-increasing demands. “Without us, there is no Google,” France Telecom CEO Stéphane Richard told the International Herald Tribune.

The world’s wireless carriers have also banded together to challenge the apps providers at their own game. The 68-carrier Wholesale Applications Community (WAC) is a new carrier-based apps market which the carriers promise will be “open” in contrast to Apple and Google’s “closed” platforms. The fact that so many carriers could come together to provide an indistinguishable service reflects how they fear competition from each other far less than from the Internet companies. Once there were Holy Wars between carriers fighting for each point of market share through service differentiation. Now such differentiation efforts take a back seat to a collective response to the Apple iStore and Google Android Market.

When I asked a senior American carrier executive about these developments his comment was, “They’re shooting behind the ducks, don’t you think?” Not only was this a catchy metaphor, but it also had the added advantage of implicitly asking the question, “How do carriers get ahead of the Internet companies rather than always having to play catch up?”

One of the ways of adjusting aim is to use the information assets inherent in each carrier’s network. To date the ducks have been flying faster than the carriers in this area. Google stole the march on location-based information by getting each user’s phone to rat out its cell site information so it could be captured, stored, and subsequently reused by Google. Then, in a similar effort to disintermediate carriers the app stores collect (and seldom share) information on purchasers and their patterns. In an Internet Protocol (IP) world this kind of data about the user is the most valuable commodity – even more valuable than the data that drives the application.

The wireless success of the Internet companies has now recreated the problem that Google was created to overcome. There is too much available out there, including the 350,000 apps on the iStore and 150,000 at Android Market. Someone needs to help the consumer find their way through this morass to the best possible experience. Someone needs to leverage the personalization opportunities of IP to help the consumer have a better online experience. No one is better positioned to do this than the carrier who knows the behavior of each user across multiple platforms, as well as his or her demographic and billing information. With appropriate privacy protection, it’s an asset of value to Google, Apple and everyone else using the network.

Let the Internet folks do what they do best. It’s problematic whether any carrier or group of carriers could build up the head of steam necessary to out innovate the folks in Silicon Valley. Similarly, let the carriers do what they do best to manage the information necessary to optimize their network and the consumer’s experience. Efforts by app providers to know about their customer will always play catch-up to the information carriers already have. The carriers could help the apps world do their jobs better – and collect revenue from the apps companies – all the while helping consumers have a better experience.

In the process the carriers would also pick up an important bargaining chip that puts them on the side of the consumer. The Internet companies want the carrier to be the bad guy and recoup their investment on the backs of the consumer by increasing rates and charging for speed and throughput. The information asset strategy, however, creates an opportunity for carriers to turn the tables and stand with consumers. The wireless app experience can be improved and consumer costs constrained if Apple, Google and other app providers will purchase information from the carrier.

Nokia CEO Stephen Elop caught the mobile world’s attention with his February “Burning Platform” email to Nokia employees. The image of jumping from such a platform overshadowed another important message in the email which, although specifically referring to devices, could just as easily be referring to carriers. “The battle of devices [and networks] has now become a war of ecosystems, where ecosystems include not only the hardware and software of the device [and network], but developers, applications, ecommerce, advertising, search, social applications, location-based services, unified communications and many other things. Our competitors aren’t taking our market share with devices [or networks]; they are taking our market share with an entire ecosystem.” The bracketed references to networks are my additions to Elop’s text, but they don’t change the message. In the mobile telephone era it was all about networks and devices. In the mobile computing era it is all about the ecosystem. And the mother’s milk of that ecosystem is information.

There may be a cold war brewing between the wireless carriers and the Silicon Valley folks, but they each need the other to succeed. No one has more information about users and network activity than does the wireless carrier. No one is more expansive and creative in the development of new applications and services than the Internet folks. It’s time for a détente and the way to reach that point is through the strength of network knowledge. Leveraging that asset to produce a better product for consumers, new revenue for carriers, and new opportunities for Silicon Valley would shift the balance of power by redefining the target so as to no longer be shooting behind the ducks.