Feature: Letter from Dr. Schreiber

31 July 2000

Hon. Edward J. Markey
2108 Rayburn House Office Building
Washington DC 20515

The recent hearing re COFDM vs "8-VSB"

Dear Congressman Markey:

You may recall that I appeared before your subcommittee at the start of the HDTV Inquiry. At that time, I was director of the MIT Advanced Television Research Program. Of the various things that I said, the one that got the most attention was "HDTV is not about pretty pictures; it is about jobs and money." Although we are much less worried today about jobs or money, the shift that is underway in terrestrial TV broadcasting from analog to digital is still very important for the future health of our economy as the importance of information technologies grows.

The FCC was quite correct in deciding that over-the-air (OTA) broadcasting must shift from analog to digital. In my opinion, there is no other way to provide the spectrum that is needed for all the wealth-creating wireless services that we hear so much about. The current NTSC system, using 50-year old technology, is simply too wasteful of spectrum, requiring an allocation of 67 6-MHz channels to provide no more than 20 programs of mediocre technical quality to each viewer. By using digital transmission and the best current technology, it would be possible to provide 20 HDTV programs to each viewer in the country with an overall allocation of only 20 6-MHz channels. Alternatively, for lower technical quality, but still higher than that of NTSC, we could allocate even a smaller amount of spectrum.

Although the FCC deserves a lot of credit for understanding this aspect of OTA broadcasting, it made a serious blunder (no kinder word suffices here) in accepting the "8-VSB" modulation method that was proposed by ACATS. This error was partly technical and partly political. Reed Hundt placed much too much faith in the "free" market's ability to design TV standards that would properly serve the public interest, convenience, and necessity. The design of the system was left entirely to the industry, without adequate supervision by the Commission. In particular, the Commission failed to insist on realistic testing. As a result, we have a system that is too unreliable to be used. While this is not the only reason for the failure, so far, of the transition to digital broadcasting, it is a problem that absolutely must be solved for the transition to be successful enough so that analog broadcasting can be turned off without a public outcry.

I was most interested in what transpired at the recent hearing. While one demonstration surely is not sufficient to conclude anything, there have now been many demonstrations of the ease of reception of COFDM (the system demonstrated by Sinclair) under many different kinds of conditions. There have been many other examples that clearly indicate the difficulty of receiving the 8-VSB transmissions on simple antennas, especially in downtown areas. A number of those testifying in favor of 8-VSB gave false and misleading statements on these matters that were, unfortunately, not challenged by members of the Subcommittee. It should be borne in mind that the system approved by the FCC was submitted by ACATS in 1995 -- more than five years ago. One would think that any problems in receiver design would long since have been found and fixed if possible. In my opinion, the 8-VSB scheme will never work well enough, no matter how much time is allowed.

Digital OTA broadcasting using COFDM started in Britain in November 1998, the same time as in the US. Nearly one million subscribers now use the service and there have been few complaints. That penetration, taking account of the different populations, is 100 times greater than in the US.

It is not as if COFDM was unknown to the American system proponents. The FCC as well as the system proponents in the Grand Alliance were fully informed about the advantages of COFDM -- about its much better performance in the presence of multipath (ghosts), its ability to support single-frequency networks that would completely solve the problem of finding spectrum for LPTV stations, and its ability to provide more service in a given spectrum allocation than single-carrier systems such as 8-VSB. For a variety of reasons, all specious, ACATS turned down COFDM.

In order not to make this letter too long, I have placed in an appendix some material relating to the history of COFDM and my own involvement in it. For the sake of full disclosure, I should say that I have some patents in the field, assigned to MIT, but I do not expect to make any money from them, no matter what happens to digital broadcasting in the US.

I would like to get the substance of this letter into the hands of Mr. Tauzin and whoever on his staff is following this matter, and I solicit your suggestions as to how to do this.

Sincerely

William F. Schreiber


Appendix: Some OFDM History

I first heard about OFDM on a trip to Europe in the late 80s, and called it to the attention of the chief engineer of the FCC on my return. OFDM was invented at Bell Labs in 1965, and the "C" (coded) was added, by CCETT (a French government lab.) in the middle 80s. By that time, it had been tested for audio in Europe and Canada with good results. Virtually all the labs then working on it had come to the conclusion that it was the right system for DTV broadcasting because of its good multipath performance. When I first described it to the FCC point man, he said that such a system could not possibly work. (A famous mathematician once "proved" that FM was impossible because it has an infinite spectrum.) Eventually, however, the FCC changed its mind and directed ACATS to investigate COFDM, which it reluctantly did.

I was sufficiently impressed by the possibilities of COFDM that I decided to take two more PhD students after my formal retirement from MIT in 1990. The project was funded partly by Scitex, an Israeli company for which I had been a consultant, and partly out of patent royalties due me at MIT, i.e., out of my own pocket. Eventually, the two students, Mike Polley, now at TI, and Susie Wee, now at HP, simulated a complete system. It was a multiresolution system with three levels of quality, using both OFDM and spread spectrum. The base-level signal -- about NTSC resolution -- had a 6-dB threshold. It worked with 0-dB echoes, and is described in my paper "Advanced Television Systems for Terrestrial Broadcasting," Proc. IEEE, 82, 6, June 1995, pp 958-981. I have a few copies of a complete report, including this paper and the two theses, for anyone who is seriously interested.

The group I was then working with at MIT decided that it would be useful to have a meeting of all those working on the subject as a means of informing the FCC and the various DTV system proponents of this new technology, then relatively unknown in the US. We had the assistance of Ken Davies of the Canadian Broadcasting Corp and Gary Tonge of the Independent Broadcast Authority in the UK in organizing the meeting and inducing the Europeans to come. All the American system proponents were invited as well as the FCC. The meeting was held at MIT in October 1992. Every lab in the world working on COFDM was represented, but almost no system proponents or FCC people came. I still have a number of the refusal letters; they were all "too busy."

The next year, a committee representing ACATS did go to Europe in accordance with the FCC directive. My opinion is that they were simply going through the motions and were fully determined to find nothing that would change their development plans. One of the stated reasons for the turn-down was that their own system had already been fully developed, and COFDM was in its infancy, to the extent that no equipment could be purchased to be tested under US conditions. Now, seven years later, some of the VSB proponents are asking us to wait while it is further developed. VSB was approved by ACATS in 1995, so one would think that in the ensuing four years, whatever work needed to be done to eliminate its problems would have been done by now.

William F. Schreiber, 13 July 1999

***



Copyright 1998 - 2004 [ITVT] | Swedlow. All rights reserved