Skip to main content

Reflections on infinite bandwidth at NAB2005

Author, consultant and engineer Robert Lucky.

How would HD transmission be if the RF spectrum were infinite? What would the product of the ATSC have been if there were no restrictions on bandwidth?

What would happen to justification for government regulation of broadcast industry — especially in light of the First Amendment — if multipath and RF interference didn’t constrain the prevailing view of the spectrum as a scarce resource? How could there be broadcast licenses? What impact would that have on the business of broadcasting?

These and many other questions swirled in HDTU’s head upon entering the NAB2005 Engineering luncheon where noted engineer, consultant and author Robert Lucky was to present his keynote speech, “Is Bandwidth Infinite?”

Lucky, who has spent his career at Bell Labs and Bellcore, invented the adaptive equalizer and holds 11 patents. For more than 20 years he’s written the “Reflections” column in the IEEE Spectrum magazine and has been honored with the Marconi Prize and the IEEE Edison Medal. Although not a government official, he chairs the FCC Technical Advisory Counsil, charged with keeping the commission informed of the latest technology developments and their potential impact on spectrum usage.

This edition of HDTU breaks with the traditional Q&A form of “Sound Off” to present excerpts from Lucky’s NAB address. In the next edition, HDTU will conclude coverage of Lucky’s insights with a few more excerpts.

HDTU: Lucky on interference and regulation.

Robert Lucky: Marconi’s daughter Gioia told me several times that her father said to her that the most important day for radio was not in 1894 when he sent the first pulse, but April 15, 1912. That was the morning in the North Atlantic the Titanic sank. And because of the role of radio, both good and bad, in that Titanic tragedy Congress immediately stepped in and started regulating airwaves.

And what seemed infinite to Marconi in 1894 now looks like we’ve used it all up, there’s nothing left. But as you probably know, the fundamental conundrum of regulation is that there’s almost nothing left but almost none of it is used.

Estimates have been made that at the busiest time and the busiest place only 15 percent of it is used. So we are talking almost all gone, and the demand today is unprecedented for more spectrum, more capacity and yet there is nothing left to allocate and yet what has been allocated is sitting there unused.

HDTU: Lucky on 3-D space and spectrum.

Lucky: I heard a talk last week of a fiber-to-the-home person, and they were saying that a single fiber has multiple terahertz of spectrum available in a fiber whereas the radio spectrum is only between 50GHz and 60GHz wide.

So fiber has very much more capacity than does the radio spectrum, and I was thinking: Yeah, but the fiber is one-dimensional and radio is three-dimensional. And every path —every ray — that we would draw through space has that same 50GHz range. And there is an infinity of paths that can be drawn through space, so why isn’t the capacity of space infinite?

But you know what we do is we take a spectrum allocation chart, and we reduce this all to a big “X,” and we say that’s gone. That’s gone. Like the whole thing is gone. But yet there is all this space out there and time that’s not accounted for in a simple chart.

HDTU: Lucky on multipath and high-low.

Lucky: Now in the world I grew up in, multipath was bad. Today, multipath is good. The more you have, the better off you are because that’s just energy that you can use. It’s more good stuff. Do the processing, put it back together again, and you improve the system. The ultimate in multipath is what everybody is working on in the theoretical community and that is high-low. High-low, multiple input and multiple output.

Suppose I want to send a data stream to someone in the back of the audience. I put up an antenna, they put up an antenna, and we use the spectrum that we have to send the data stream out. Now in high low, I put up eight antennas, so I have eight antennas that I use, and this is not a phased array. I am going to send different data streams in each of these eight antennas in the same bandwidth and back there in the back of the audience you are also going to have eight antennas not pointed in any particular direction whatever, and you will receive in each of those antennas all of the eight different streams I sent out.

And I can represent the transmission between me and the back of the room by a matrix of eight by eight that says, “This is what you do to operate on signal 3 to go into antenna 4.” That kind of thing. So I take that matrix and using super computer-like processing, which we have available today, I invert the matrix. I make a diagonal matrix, and I have eight independent channels all sending different data in the same bandwidth between me and the back of the room.

And under good conditions I have eight times the capacity that I had before.

To be continued…

Back to the top