Footnote to multimode's history

Stephen Brown's May 2000 column, "History Lesson: Fiber's current boom made by genius and AT&T divestiture" (page 16), brought back some memories. Although I agree with much of what he wrote, I would like to correct one point and comment on several others.

The statement that "at the time [1980] of the [Northeast Corridor] filing, AT&T could manufacture neither singlemode nor multimode fiber" is incorrect. Multimode fiber had been manufactured at the Western Electric Atlanta factory since at least 1976 and was used in a number of installations prior to 1980. There was never any question of Western being able to make the fiber.

Mr. Brown indicates that I "glossed over the details when [I] wrote a historical description of the project in 1986." What he refers to is a chapter entitled "Fiber-Optic Transmission Technology and System Evolution" in a 1986 book, which was a technology and applications overview and was not intended to be a history. Somewhat more of the history is in a February 1995 article in Optics and Photonics News. But as explained there,

I had only limited access to the high-level discussions that were going on at AT&T at the time of the Northeast Corridor filing. (I am flattered as being characterized as an "AT&T luminary," but in reality, I was a middle manager at Bell Labs.)

Nevertheless, let me both comment and speculate on what happened following Corning's argument to the FCC that a viable domestic fiber industry depended on there being more than one fiber supplier for the Northeast Corridor system. Although there were sentiments within Bell Labs and AT&T for buying some of the fiber on the outside, Western Electric was strongly opposed, and their viewpoint prevailed. So as not to delay the installation, the FCC approved the first phase of the project, but made the second phase contingent on AT&T going out for competitive bid (for the system, not just the fiber). Fujitsu was the low bidder, and the procurement organization within AT&T seemed ready to award them a contract.

In that February 1995 article, I wrote (with carefully chosen words) "after high-level discussions between AT&T and the executive and legislative branches of the federal government, AT&T Long Lines selected Western Electric, who had submitted the lowest bid of the domestic suppliers, to provide the second phase of the Northeast Corridor." I was not a party to these discussions, and I do not know who initiated them, but it is not unreasonable to assume that AT&T stimulated some of the congressional letters.

I agree with Mr. Brown (and indeed that was one of the points of the 1995 article) that the Northeast Corridor had a profound impact on the subsequent history of fiber optics and the AT&T divestiture. At the time of the AT&T antitrust suit, the prevailing sentiment in the industry (if not in the Department of Justice) seemed to be that AT&T should divest Western Electric. It is ironic that AT&T again argued to maintain its manufacturing, only to unilaterally divest the manufacturing some 12 years later.
Ira Jacobs,
Professor Bradley Department of Electrical and Computer Engineering Virginia Polytechnic Institute and State University Blacksburg, VA

Having spent over 25 years as a design engineer in the telecom marketplace, 15 years with Scientific-Atlanta and 12 years with Telesync, I read Stephen Brown's article in Lightwave magazine, entitled "History lesson: Fiber's current boom made by genius and AT&T divesture" (May 2000, page 16) with some interest. I cannot support his conclusion and final statement, "Clearly, had there been no divestiture, there would be no fiber boom today," since you have clearly overlooked pre-fiber technology history and demands of the marketplace.

In 1980, the backbone of the U.S. long-distance service was digital microwave radio, and at that time, 90-Mbit/sec radio was king, with 135 Mbits/sec becoming available. The 90 Mbits/sec was composed of two DS-3s, each carrying 672 voice channels-1,344 calls for 90 Mbits/sec and 2,016 calls for 135 Mbits/sec. Do you remember all of the hilltops with towers and microwave horn antennas? Why didn't AT&T just run another digital-radio route for the Northeast Corridor? The answer is that digital radio was max'ed out, the towers couldn't hold more antennas, the radios and support equipment were expensive to maintain and calibrate, and demands on the network were growing faster than digital radio could grow to meet [demand].

There was another significant problem and that was transoceanic phone service: Existing coaxial cable/repeaters were expensive and couldn't carry enough traffic, and the satellite system was not satisfactory for voice traffic, as the half-second or so roundtrip delay was quite annoying in conversation. Along came fiber optics, with the potential to solve bandwidth demand and time-delay problems with much lower maintenance; Reagan the free-market advocate; and a marketplace that needed all the bandwidth it could get and more. Everything seemed to come together at the right time.

We could argue whether divestiture speeded up or slowed down fiber development, but it would have come regardless, since market demands required the bandwidth that fiber offered, which existing 1980s technology-digital radio/coaxial cable-just couldn't meet. The geniuses behind fiber development should of course be recognized, as I think fiber optics may go down in history as one of the greatest inventions of the last century. Lest we can't forget that old adage, "Necessity is the mother of invention."
Robin Larson
Vice President, Engineering
Telesync Inc.

Thanks very much to Prof. Ira Jacobs and Robin Larson for their comments.

Robin Larson's main point, "it would have come regardless, since market demands required the bandwidth," is on the money, but the timing of the "would have come" is still a big variable. My own opinion is that divestiture really quickened the market for fiber.

Prof. Jacobs's remarks about Western Electric are valuable additions to the historical record and confirm an opinion I've heard before about how AT&T operated before divestiture: Western Electric was the tail that wagged the dog, a subsidiary that did not like being one and that often successfully placed its interests above those of the larger corporate body, to the ultimate detriment of AT&T itself.

Also, it is a surprise to me that Western Electric was manufacturing multimode fiber in 1976, a fact that may not have been widely known then. In 1980, another AT&T subsidiary, Mountain Bell, had applied to the FCC to build a fiber-optic system using NEC fiber equipment to link NORAD's Cheyenne Mountain facility and nearby Peterson Air force Base. The Department of Defense objected to NEC's participation, and the company dropped out of the project. But it is still peculiar that Western Electric was not Mountain Bell's first choice.

As for Prof. Jacobs's remark that he was in "middle management" and therefore not an "AT&T luminary," my response is that being a luminary does not require being in top management and being in top management does not require being a luminary.
- Stephen Brown

More in Market Research