Difference between revisions of "Oral-History:Donald Schilling"
m (Protected "Donald Schilling Oral History" [edit=sysop:move=sysop])
Revision as of 18:44, 30 December 2008
About Donald Schilling
Schilling received his Bachelors at City College (1956), his Masters at Columbia (1958), and his PhD at Brooklyn Poly (1962); his dissertation was on the phase lock loop, a control system for spread spectrum and other communications (particularly wireless) systems. He was Professor of electrical engineering at Brooklyn Poly until 1969, professor at City College of New York, 1970-1992, then worked full-time at his company, SCS, until 1996, when he founded a new company, Golden Bridge Technologies, where he has worked since. His research has included much work on code division multiple access (CDMA); 1960s research on Song Adaptive Delta Modulation for NASA and the Army, for digitizing voice or video; meteor burst communications; spread spectrum systems (chirp radar, direct sequence spread system); PCS; 3G; and work developing wireless standards internationally. He has had extensive involvement with IEEE, ComSoc, and other societies, and received the Armstrong Award in 1998 for lifetime achievement in wireless communication. He describes the major developments in the communications field since World War II; describes technologies such as spread spectrum, multipath, interference cancellation, and adaptive power control; describes the growing importance of wireless communications and predicts its future, and discusses his colleagues at Brooklyn Poly during the 1960s.
Schilling's colleague Raymond Pickholtz provides further explanation of spread spectrum systems and CDMA in Raymond Pickholtz Oral History.
About the Interview
DON SCHILLING: An Interview Conducted by David Hochfelder, IEEE History Center, 7 September 1999
Interview # 356 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc., and Rutgers, The State University of New Jersey
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, Rutgers - the State University, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Don Schilling, an oral history conducted in 1999 by David Hochfelder, IEEE History Center, Rutgers University, New Brunswick, NJ, USA.
Interview: Don Schilling
Interviewer: David Hochfelder
Date: 7 September 1999
Place: Sands Point, Long Island, New York
History of communications systems since World War II
Give a thumbnail of the progress in communications and engineering since World War II, but especially since the founding of the Communications Society as part of the IRE in 1952. Secondly, if you could explain how your career has fit into those developments in communications technologies since that time. Third, if you could talk about your role in the communications society and also the role of the society in the progress of the technology.
Since the end of World War II, the USA had a developed telephone system, wired, and this is the basis of all our communications—home telephone—which in America was fantastic compared to the rest of the world. We were a shining light. I believe credit for the shining light is due to Bell Labs/AT&T and the telephone companies that were a part of AT&T before the divestiture. All of that caused the USA to have a telephone system that was beyond reproach; it was really a great system. We were connected to the rest of the world after a while by satellite or undersea cable. Satellite never sounded good, and still doesn’t sound good due to echoes. Undersea cable has almost no delay, of course, and sounds considerably better. As time progressed, if we don’t consider the fax but stick to the telephone, the biggest change was the cellular system. Walkie-talkies and different types of wireless systems were used by the military and commercial, however, these were not cellular systems, available to the masses.
In the ’60s and ‘70s, particularly in the ‘70s, a tremendous change occurred in communications. It has been twenty years since the end of the second World War, but it was at that period that AMPS’s (analog FM) cellular communications took hold, due to AT&T and Motorola. That was a great system, even though the voice sounded terrible, and many of the calls got cut off so that it took perhaps four calls to complete a call. The most beautiful thing was that I could drive along in my car and takes care of my business calls at a fair and reasonable price. People from around the world got involved in this because most of the world did not have a really good wide area telephone system.
After that, more changes took place. Digital came to be known as "second generation." Second generation systems had to do with time division multiple access (TDMA) and Qualcomm’s narrowband, code division multiple access (CDMA). Most of the world went with TDMA, and that received an awful lot of attention and sales throughout the world. TDMA constitutes approximately ninety percent of second generation sales, with CDMA getting maybe five or ten percent. The TDMA sounded no better than the AMPS, but you could get about three calls for every one, so it allowed the cellular provider to make a little bit more money. The same thing was true with the CDMA system. The CDMA system today is first reaching mature status in my estimation, TDMA systems sound better than the CDMA systems.
In 1990, PCS (Personal Communications Systems) began. The idea of PCS was to provide a better quality of voice and higher data rate than analog or basic digital. It was at a different frequency band: 1840 to 1990 MHz, which is above the military band and below the microwave band used by radio broadcasters. In referring to PCS I am including both the American version and the European version, which is typically called GSM. In the PCS system the narrow band CDMA is simply raised to a new frequency band. In the United States those frequencies were auctioned off, and many of the companies that bought the frequencies have gone Chapter 11 or some such thing.
Now PCS is starting to take hold and starting to compete with the 900-megahertz band wireless communications systems. The next thing that happened was the start of Third Generation Wireless (3G). That is where the future is. 3G is supposed to provide better quality of voice and higher data rates of up to two megabits per second, and Internet access via packet switching. PCS never achieved those goals, but provided essentially the same quality of voice and the same data rate at a different frequency band so as to be competitive. PCS tries to provide more features, like call forwarding, etc., but it is basically the same thing as first generation cellular. When 3G comes to pass, it will be another great step forward in communications.
On the route going this way, the FCC, in the early 1980's set aside some bands called the ISM bands (Industrial Scientific Medical bands), 902 to 928, 2400 to 2483, and 5700 to 5825 MHZ. These bands were originally set aside to be only for spread spectrum use, to encourage the use of spread spectrum. A lot of companies took advantage of this band to develop spread spectrum equipment to sell in the marketplace. It really aided tremendously in the development of CDMA and the application of CDMA for LANs and things of that nature.
That represents the status of wireless communication. I left out the satellite because LEOS (Low Earth Orbit Satellite) systems may have started in the early ’90s or late ’80s, but none of them really work well. I read in the Wall Street Journal recently, that the first one to come out, Iridium, just went Chapter 11. Motorola funded that. In my personal opinion, Motorola made a terrific error in judging the marketplace. The satellite has to come from such a large distance down, even when they are LEOS, and you can’t go back up with low power from a remote. In order to have communication with a LEO, you have to have direct communication —you have to “see” the satellite. To see the satellite in New York would require a considerable number of satellites, as you don’t even see the sun, but a reflection, as it bounces down. It is true that in places like the desert or the ocean, or places where there are no people, you can see the satellite. However, simply, if there are no people, how do you get subscribers? The satellite works well where there are no people, and works terribly where there are people. Therefore, it is a terrible system. My opinion is, it will be changed so that satellite communication will be to a cell or base station on top of buildings, which will then go to the LEOS, and then go to remote users. In that case, why do that? I could either go by satellite to keep it all wireless, or I could go with a regular telephone system, or with microwave. It is just another option that people have. It may be good or it may be bad, but that is the way that business will have to be structured. The Iridium system appeared not to structure their business that way, and it appears to have failed. It was a tremendous expenditure of money by Motorola and others.
That is how I see the wireless branch of communications going. Wireless, I think, has proven itself to be the most important branch today.
Secondly, please talk about your career, and how your career has fit into the history of communications and engineering in the past thirty or forty years.
I began by my career graduating from the City College of New York in 1956; receiving my Masters from Columbia in 1958; and my Ph.D. from Brooklyn Poly in 1962. My dissertation was on the Phase Lock Loop, which was the control system that is used in all spread spectrum and other types of systems. My dissertation emphasized the demodulation aspects of FM signals specifically. Mischa Schwartz was my advisor.
Phase lock loop systems are very, very popular in any communications system, particularly wireless ones. They have a number of applications. One is when a signal arrives, the frequency of that signal is not known. A local oscillator must be locked to that incoming signal. A phase lock loop is the technique by which the local oscillator is synchronized to the incoming carrier frequency. Phase lock loops can also be used to demodulate FM signals. An ordinary FM discriminator can also be used to demodulate an FM signal, but an FM discriminator assumes that the carrier frequency is fixed, like you have on your radio station. But if it comes in from a wireless system where there could be a Doppler shift, then the frequency is unknown. Then part of the demodulation process must be the synchronization of a local oscillator to the carrier frequency. The phase lock loop shines here as it allows this carrier frequency to be adjusted. Hence, it is used often for wireless communications where the carrier frequency drifts due to a Doppler shift.
After my Ph.D., I became a professor of Electrical Engineering and taught all sorts of courses in electrical engineering at Brooklyn Poly until 1969. In 1970 I moved to the City College of New York, where I taught until I retired in 1992. During that time I co-authored 12 texts and more than 200 papers. I advised more than 75 Ph.D. students, many of who wrote their Ph.D. dissertations on CDMA related topics. In the 1960's I was involved with adaptive delta modulation, a voice digitization technique which was used by the military at that time; it is still used to perform analog to digital conversions. We came up with a new type of delta modulation system called the Song Adaptive Delta Modulation system that allowed us to digitize voice or digitize video. I did the work initially for NASA, then after NASA I did it for the Army, and then for several other government agencies.
After the adaptive delta modulator I became involved in meteor burst communications, which was a wireless system used to transmit signals beyond the line of sight, using the meteor channel. Meteors are formed all the time, and we bounced our waves off of the meteors. If you do it right, you can get continual voice. A lot of this was done for the Air Force in Anchorage, where they transmitted between Anchorage and Kozebue, which is a city in Alaska right off the Russian border across the Bering Sea. It was a lookout station, and there was no real ionosphere to communicate back and forth, so they communicated off the meteors. We built the equipment and demonstrated that continual voice could actually be achieved if it was done properly. We had an adaptive system. I edited a textbook for John Wiley on Meteor Burst Communications, which dealt with that whole area.
I had a consulting company that was involved with spread spectrum systems. I had been involved with spread spectrum since the 1950s. My first area of spread spectrum was the chirp radar, which is a system in which you change frequencies continually and linearly with time. That was hard to build initially, but we did that.
In the 1970s I got involved in teaching courses in spread spectrum with Professor Raymond Pickholtz. As time went on, Professor Ray Pickholtz, and Professor Larry Milstein, and I did more and more consulting work for the Army in spread spectrum. We were very interested and active in the whole area of direct sequence spread spectrum, frequency hopping spread spectrum, and hybrid systems combining the two of them.
In the 1985 timeframe, we were building some equipment for the Navy. Millicom, a company that was involved in first generation cellular, wanted to get involved with second generation cellular and PCS. They were involved with PCS in England, and they did not like the way the TDMA systems worked. They were told by a number of companies to contact me regarding spread spectrum. They came to my lab, and I showed them how direct sequence spread spectrum worked and could be used to transmit and receive voice. The beauty of CDMA is that it mitigated the effects of noise due to multipath disturbances or other users on the band. Nothing helps thermal noise.
Spread spectrum technology and wireless systems
Development of spread spectrum technology
When did spread spectrum technology first come into existence?
You can go back to the ’40s. The May 1982 IEEE Transactions in Communications, Part I, is a special issue on Spread Spectrum Communications, guest edited by Charlie Cook and Fred Ellersick (who unfortunately is deceased), Larry Milstein and myself. The first paper by Bob Scholtz, called “The Origins of Spread Spectrum,” gives a very interesting history of spread spectrum. Professors Pickholtz, Schilling, and Milstein did a tutorial called “The Theory of Spread Spectrum Communications” in the same issue. Scholtz’s is the best and most thorough history. In another issue Bob Price interviewed the movie actress Hedy Lamar, who had a patent on frequency hopping spread spectrum. I have a copy of the patent. It is fascinating work concerning torpedo tracking that she did with her husband.
Very little is not known about the basic principles of spread spectrum. The patents being filed offer specific implementations and specific aspects that generally are not known. However, going back far enough with a good enough literature search will probably reveal that almost everything patented has appeared in the open literature at one time or another.
AMPS and TDMA systems
Let me explain the different types of wireless systems. One is called AMPS and is an FM system. The European version is called TACS. It is a very simple FM system whereby you have a particular radio frequency that you transmit on, like an FM station, and I transmit on a different FM station to you. So you listen to my FM station and I listen to your FM station, and we can demodulate each other’s FM stations. We communicate with basically two FM radios. Such systems are called Frequency Division Multiple Access (FDMA) systems and are very wasteful, because you are using two whole channels with one voice. Suppose I let multiple people share a voice channel. I digitize your voice, compress it, and then burst it out using say QPSK (Quadriture Phase Shift Keying). Such a system is called a Time Division Multiple Access (TDMA) system. The American system uses three voice channels. Three people store their voice digitally, and burst it out first, second, third, repeatedly, in that order. The European system is called GSM. In GSM, eight people burst it out in the same bandwidth. And you can do that. It is a matter of digitizing the voice efficiently and properly timing the transmissions. When you digitize a voice efficiently you lose a little bit of the quality. In my opinion, there is no real significant difference in the voice quality between TDMA (including GSM) and AMPS. The AMPS channel is so narrow that the FM loses its great benefit, which occurs when using a wide band; AMPS is a narrow band FM system.
Is Qualcomm's system also a TDMA system?
No. Qualcomm employs CDMA (Code Division Multiple Access). The beauty of CDMA is that the voice is coded, and everyone gets a digital address on top of its voice. Because each digital address is different, like when we put our letters in a mailbox, they are all able to be sorted because the address is different. So, too, with CDMA. Everyone sends their message at the same time over the "ether," which, if you like, is a gigantic mailbox in the sky, and you can pick out your own from the mess, because you are looking for a particular address. The other addresses, correlate with a low correlation and represent very little noise. If you know what the different codes and addresses are, techniques, called Interference Cancellation Techniques, allow you to “subtract” off those addresses and reduces the interference even further. The techniques are currently being developed.
There are two types of spread spectrum. One I just described: Direct Sequence Spread Spectrum. The other is called Frequency Hopping (FH) Spread Spectrum or Frequency Hopping Multiple Access, generally developed for the military for tactical applications. SINCGARS is a particular military radio that uses frequency hopping multiple access of spread spectrum. The message is digitized but it is not spread, but then you jump to other frequencies periodically so that an enemy, not knowing which frequency you are going to jump to next, has trouble listening to your conversation or jamming your conversation. The "friend," who is listening to you, knows which frequency you are going to and knows the sequence of frequencies you would use. FH is different from Direct Sequence in that Direct Sequence always uses the same frequency to send a digitized signal on top of which is a digital address. A hybrid system does both: the digital voice is put on a digital address, which jumps from frequency to frequency.
Hybrid systems are a little too complicated for commercial use. In my opinion frequency hopping is not good for commercial use because when two users land on the same frequency there is a lot of interference. The direct sequence solution has really taken a-hold in the commercial world for spread spectrum use. Qualcomm was the first one to truly commercialize it, even though they use narrow band spread spectrum, which in my opinion is not the way a good spread spectrum system should be built. However, they were allotted a certain amount of bandwidth, they operated within their bandwidth, and they gave us probably the best system possible, within the confines of that bandwidth. The third generation (3G) systems have wide bandwidths. To give you an idea, the bandwidth of the Qualcomm system is about 1.25MHz. The new 3G systems are at the minimum 5MHz, and they can go to 10MHz and even 20MHz. 3G is a minimum of four times wider with a possibility of 1.25 -- 20, which is like 16 times wider. That is quite a difference.
Can you explain a little about multipath?
When any type of radio signal is sent through the air, that signal spreads out from the antenna. Rays emanate and bounce off of buildings that are in their way. Eventually many of these rays are collected at the receive antenna. Because the number of bounces differ for each of these rays, they all come in at different times. These received "rays" are called the "multipath signals." Since they come in at different times, some of them add and some of them subtract in phase. You can actually get an amplitude variation, and some of the variations cause the signal to be very small. For instance, on your wireless phone you can get no signal, and then you take one step and you can get a good signal. That is because of the multipath effect. Only a very small fraction of a wavelength, a half of wavelength or so, is needed to significantly change the amplitude.
If spread spectrum is wide band enough, each of the received multipath signals will look like another user’s address (another user’s signal) and cannot correlate with your original signal. It is only when the spread spectrum is very narrow that the multipath signal's addresses look alike and correlate. This produces noise to the multipath. If the digital addresses do not overlap the multipaths look like other signals. When a spread spectrum signals overlap and are correlated, that is bad. They are no longer correlated, when they are offset by what we call one chip.
Now, a chip is the shift to another frequency?
No, we take each bit of the data signal and put it on that digital address. That digital address is a binary signal at a very high data rate relative to the original information data. Typically, ten to one, 100 to one, 1000 to one is really what you would like. For example code voice to fifty kb/s (which means each data bit lasts less than twenty microseconds). A binary signal multiplying this data signal, where each bit of the binary spreading signal lasts for two microseconds, gives you a ten to one spread in bandwidth. You need ten times more bandwidth to transmit the resulting data signal. Because of that ten times more bandwidth, if you are off by one of those bits of the binary spreading sequence, it looks like another signal. There is very low correlation. It is almost orthogonal. The bits of the binary spreading signals are called chips, because they seem to "chop up" the data bit. There is a whole theory developed about that. My last book, The Principles of Communication, has a whole a chapter on spread spectrums. Now many of books on spread spectrum out are available.
Could you tell us a little bit about methods that a CDMA communications system could use to correct for multipath?
The easiest method is not to correct, and just make your system wide enough so that the multi-path just represents noise.
Another technique is called RAKE. It is not an acronym; it is really a rake—sweep all of the multipath signals together and try to align all these signals; speed up ones that are delayed and slow down those that are not delayed. The signals then all come in, in phase, so that they all add up. You take advantage of the multipath in that manner. It is easier to say it than to do it. However, every system now employs RAKE in one form or another. For instance, GBT (Golden Bridge Technology), a company that I am associated with, uses a matched filter receiver, which is a system of delay lines. When the incoming signal lines itself up with the spreading sequence, you get an output of either one or minus one, which represents the data bit. If it is sixty-four to one spreading factor, there are sixty-four of these spreading bits (chips) per data bit, and when the data signal lines itself up, you get an output which is 64. When a multipath signal comes in, you get another output. When the next multi-path comes in you will get another output. In other words, since the matched filter was just "sitting there waiting," every time a multipath signal enters and leaves you will get an output which is 64. All these different outputs from that same matched filter are stored in the computer in memory, and they are all added up. That is the simple way of doing it. This technique is used in the 3G standard.
The third way is called Interference Cancellation (IC) which does not really handle multipath as well as it does interference from other users. I would not call IC a multipath cancellation, but a regular interference cancellation of other users' signals. Interference cancellation can be used if I know all the different codes; then I know what is present and I can subtract them all.
What about adaptive power control?
Adaptive power control is different. When many users transmit to the base station, a problem arises with a direct sequence spread spectrum, which is called the "near-far" problem. I allude to that in the 1982 COMSOC tutorial article. The near-far problem means that the strongest signal wins, because if the signal that comes into the matched filter, or correlator is very strong you get a strong output. A wrong signal, that is a hundred times stronger than the right signal, even though it is attenuated, due to low correlation, may still give you a larger output. The base station monitors each of these signals that are received and tries to get each user to transmit in such a way that they are received with the same power, so that one will not outdo the other. That is called power control. This has to be adaptive because as you move, the power at which you receive the signal is changing. The base station receives all these signals, looks at their power levels, and adjusts them all higher or lower on a continuing basis. There are papers on adaptive power control that go back to the 1970s. A very famous one is by Ormendroyd from the University of Bath, published in 1982 in a conference cosponsored by the Communications Society. It is a very well known paper. So all of these patents on power control that come after it have to be specialty items. You cannot patent the principles of adaptive power control, since there is open literature.
SCS and FCC demonstrations
If we could talk about your role in the IEEE, and also the IEEE and COMSOC's role in developing communications technologies.
In addition to teaching courses in spread spectrum since 1970 my company SCS, became involved with Millicom in 1989. Millicom had come back to the United States and wanted us to join with them to develop PCS, and they wanted this to be wide-band CDMA. We put voice on a wide-band CDMA system and demonstrated it to them, and they liked it very much. We made our presentation to Mr. Al Sikes, who was the chairman of the FCC at that time. He allowed us to put the frequency band between 1840 and 1990 MHz. The low side was chosen to be just above the military band and the high side was just below the FM band that was used by newscasters. We went right inside the frequency band used by the microwave band users. We were able to demonstrate to the FCC that using our approach, we did not interfere with the microwave users and they would not interfere with us. This was demonstrated, subsequently, in Orlando, Dallas, Houston, and San Diego. We also demonstrated in China, etc. that is all over the world. We really put this through the paces. Obviously we received an experimental license in order to do this, but other companies got experimental licenses as well. Then there was a "shootout." By the time there was a shootout, Milicom had decided to withdraw from the project. SCS then merged with Interdigital Communications Corporation (IDC) and I became executive vice president and also vice chairman of the board, and a large shareholder. At that time I retired from City College and went to work fulltime at this job. Primarily we were selling wireless loop systems, which is a wireless telephone system there was no wired infrastructure. For example, how do you run cable to 25,000 islands in Indonesia? In China there is not enough copper in the world to supply wired communications, nor enough people to install that copper, if there was such copper, in a lifetime. In a capitalistic society you must communicate, and in order to communicate quickly, we had to provide them with wireless communications. Interdigital’s business was doing that initially with TDMA, and then they wanted our broadband CDMA system, and that was under development. IDC also received about 50 of my CDMA related patents. They now have the CDMA product working. They are also very active 3G technology.
Golden Bridge Technology and wireless standards
I left Interdigital at the end of 1994, and I basically took a year off. In 1996 I became chairman of Golden Bridge Technology, a company that wanted to get involved with the third generation standard for wireless to provide that higher quality. There the idea was to use the matched filter as a receiver while Interdigital and Qualcomm use correlator receivers. To use a correlator receiver you really need a pilot signal to do it right. The idea at GBT was to use matched filters, which did not need the pilot signal, but which used a header, which was a little different philosophy. Each signal was preceded with a header, and so it was packet oriented because the matched filter detected a particular packet. This concept of packet orientation is extremely important for third generation and fourth generation technologies because it allows you to synchronize to each packet. This is a much more efficient operation and much more immune to the vagaries of Doppler and multipaths, etc.
I then worked with ATT and GBT, and formed the TIA Group 46.1, which was to standardize the Golden Bridge Technology in the TIA (Telecommunications Industry Association, a US standards body) for third generation use. We had a number of companies, including IDC, that agreed with our philosophy and they worked with us to develop the standard. Then we found that the Japanese standard body (called ARIB) and the European standard body (ETSI) also agreed with our design. Their design and our design were very similar. Our committee then decided to accept the ARIB and ETSI proposal, and we basically merged all of our standards into one and agreed to work together. At that point, we were joined by T1P1, which is another US standards group. What got me very upset was when Madeleine Albright, US Secretary of State, decided that the 3G proposal being developed in TIA 45 was the U.S. proposed standard and that wide-band CDMA was a foreign proposed standard. She made that comment in the newspapers. It was totally wrong. There were two U.S. proposed standards for CDMA, a narrow band standard that was being developed by Qualcomm, and a wide band standard -- TIA 46.1. The International Telecommunications Union (ITU) has combined these two and other proposals together into a single, multi-mode, 3G standard. If I look at my role in 3G, I was one of the prime movers in using this matched filter based technology. Indeed, I have been preaching wide-band CDMA for commercial systems since 1985, when everyone was telling me that narrow band CDMA, Qualcomm’s system, was the way to go. Everyone is now working on wide-band CDMA, and it will be adopted. The idea of packet transmission, communications with packets is being adopted. I am very pleased that what I have been working at for most of my life it seems to becoming a reality. As a result of my work in wireless communication, I received the Armstrong Award by the IEEE Communications Society in 1998. I am very proud of that award because I am very proud of the work that I did.
IEEE, Communications Society
My role in the Communications Society goes back to about 1968. I just donated all my journals to Clemson University except for a few that represent auspicious moments. In 1969 I became the Publishing Editor of the Transactions Editorial Board and Ron Slayton was the Editorial Manager. The vice chairman CDMSDC was Dick Kirby, and he became chairman in 1970. It was Dick Kirby who asked me to take over the job. Then I became the Editorial Manager in 1970, a year later. When I took over I appointed all of the Editors to serve with me and to help me. I changed the policy of COMSOC (it was called CommTech at that time) to be very responsive to the authors. Indeed the authors then started getting their reviews back within a month. I really laid down the law. I picked my own editors who were responsible to me. I changed around the whole philosophy. I initiated the COMSOC Magazine, and the first magazine was in 1973. It was a very small magazine at the time. Al Culbertson was the president. Steve Weinstein, was foreign abstract editor and later was Editor of the Magazine and helped significantly to "grow" the Magazine. The original idea of the Magazine was to produce articles without equations that the average engineer could understand. The people on the Board of Governors of COMSOC told me the only way that they would give me more money for my budget was if I produced a magazine that they could understand—not just for the "elite." Indeed, the Transactions is for the researchers. We put out two issues of the Transactions, one a regular issue covering all topics, and one dealing with a particular area in selected topics. Around 1980 I decided we should have a separate journal for these "selected topics," and we did that in the journal, JSAC. There was a resounding applause to it. Many, many people signed up for it, which had to be paid for separately, and for the Magazine, which expanded rapidly. After a while I had no trouble getting people to write for the Magazine even though it had no equations. It had no archival value, if you will, and the university people initially said they couldn’t write for it because without any archival value there would be no credit towards tenure, promotions, etc. But after a while, everyone wanted to write for it, so we had no trouble. As President in 1981,I set up several conferences, MILCOM and Infocom; and I also got COMSOC to get away from the idea of splitting their conference returns with AES and GRS. After that, COMSOC took the risk and all of the rewards. I stayed as editor until 1979 and then became vice president. I was a petition candidate as the Nominations Board of COMSOC did not like the idea that I was from a university—at that time they used no university people in these higher positions. They thought we did not know what we were doing and we were just theoretical quacks. At the time it required a few thousand votes to be a petition candidate, now they have lowered this number to about 100.
So your involvement with the Communications Society in an official capacity lasted about thirty years.
From 1969 until about 1983. I do not consider myself to be in any kind of official capacity with COMSOC at this point other than I am a member. As a past president I was on the nominations board. I guess that is official capacity. I was director of the Board of Governors for the IEEE for a year. And then after that I just stopped. My feeling is that when you are done doing something, let the people that are doing it, do their job and do not interfere with them. Your words of wisdom are your words of wisdom. Let them have their own wisdom—they are as smart as you are.
Communications Society and technological progress
Can you talk about the role of the IEEE Communications Society in fostering the progress of communications technologies?
ComTech was really telephone company oriented: it was run by people from the telephone company and it had everything to do with the telephone company. The COMTECH and subsequently the COMSOC technical committees, editorial board, and conference board were initially "telephone" oriented. Wireless spread spectrum was looked upon as frivolous. A small division of COMSOC dealt with wireless, but it was not given much substance. It was really put in a class of telemetry. As time evolved, this evolved. More University people became involved and their research was supported by the Military and included spread spectrum. It was a fight, like any kind of change is a fight. But it did evolve and it went more and more towards the wireless end. Particularly while I was Editor and through my Presidency, we tried to be in the forefront of all changes. Any change that looked hot and promising we put an editor in charge and we created a special issue and gave special attention to it. COMSOC has moved in that direction since then. You will notice the new COMSOC magazines indicate when something new is taking place and COMSOC gotten into it. I think that is why the membership has grown tremendously. During the time that as I was Editor through the time that I was President, the membership rose to about 35,000 members. Now it is really very large, over 50,000. We have become a major organization and an international organization. We have been accepted by the Japanese, and by virtually every country in the world. Certainly, if you have communications theorists in your country, those people belong in the Communications Society. Even though it may be expensive, it is the place to belong, and for a very good reason. Our Conferences and Journals are great. They are really to the point, on a wide diversity of topics. We have a Communications Theory workshop that has been going on from 1970 to this date, started by Bob Lucky, Jack Salz, and myself. Lucky and Salz were at Bell Labs, and I was at the City College at the time. Now COMSOC sponsors many more workshops and different disciplines. Workshops are different from conferences because they are small, and intimate, with about seventy-five experts in the field. That is very good. The combination of workshops, big conferences, and The Transactions Magazine have really made COMSOC what it is today—the international communications organization. COMSOC magazine has won all sorts of awards for its format, etc. When we first started, initially no one wanted to write for it and no one wanted to put an ad in it. Now look at all the ads and look at all the people that are writing for it. It is really very hard to get published in it.
You eluded to in the early days that Bell Labs and the telephone companies had a lot of influence in COMSOC. Has it changed to where more university people have gotten into it, or is it still industry oriented?
Today, a lot of theoretical people are getting involved and a lot of international figures are getting involved as well. Many of the people on the Board of Directors now are from other countries, so we have internationalized to a large degree. It is a much more a disperse group of individuals from their location, their profession, and what their specialties are. I think it is really very, very nice.
Ascendance of wireless as dominant technology
When did the shift occur from wire to wireless as the dominant concern? When did that happen?
That evolved with time. It followed along the course of events of telecommunications. As things change, you have no choice. You have to be wireless or you wouldn’t have an organization.
Over what timeframe do you think that shift occurred?
In 1982 it was probably fifty-fifty, but now it is probably seventy-five percent wireless.
So basically in the 1980s?
Brooklyn Poly researchers
Could you talk about the connection between your work and work of folks like Ray Pickholtz and Larry Milstein, and then the group that came out of Brooklyn Poly in the ‘50s and ‘60s?
Professors Pickholtz, Milstein, and I worked together since our Poly days. Milstein was my Ph.D. student. We were really good friends.
Brooklyn Poly graduated a lot of people. When I was going for my Ph.D. at Poly there were about 1,500 graduate students. When we graduated, one group went to work for the telephone company. They became wired communication guys. The other people were interested in AM, FM and things like that. You might have analog and data, but it was AM and FM. Then you have people that worked for the military. The military type of person was interested in spread spectrum. The people on the West Coast that worked for JPL and NASA communications were wireless. The military groups were wireless. It doesn’t work to go into battle with wires hanging around you! Those people were involved in spread spectrum. So there were two distinct groups of people that did not even talk to each other. Not until recently did they really get together and each one understood what the other was doing. I like to feel that we understood the FM and AM, but the telephone people did not appreciate spread spectrum. Now everyone understands spread spectrum. I think a large number that graduate from college, even with the bachelor's degree, now understand spread spectrum.
It seems like there was a real critical mass of people that came out of Brooklyn Poly in this area of communications at the same time.
A lot of people were attracted to Professor Mischa Schwartz being there in the 1960's. You have to have a nucleus, and once you have a nucleus people come. That was a heyday at Brooklyn Poly at that time. I met Dick Kirby at Poly when he visited an underwater water chamber I had built to demonstrate multipath. That was in 1964. We had transducers, and you can transmit from one transducer to another. The wave that emanated bounced off the walls of the water tank, and so you created multipath signals. We were concerned about how to get rid of the multipath. That is what attracted Mr. Kirby. His original work was ionospheric propagation. In ionospheric propagation you also have multipath. We all knew the other investigators that did this kind of work. JPL did spread spectrum. Deep space probes used spread spectrum. The military people all worked together at spread spectrum. Everyone worked in the same general area except for the telephone company people, which again, was a huge part of the population.
As far as spread spectrum, were communications better at Brooklyn Poly?
That was wireless. That was Professor Schwartz. That was also in the 1960's. By 1970, Schwartz, Wolf, Pickholtz, Milstein, and I were gone. Schwartz to Columbia, Pickholtz to GWU, and Milstein and Wolf ended up at UCSD, I went to CCNY. It is interesting to note that for a while many investigators became a big proponents of wired. He felt that wired was the way to go and that wireless was obsolete. What do you need wireless for when you have a wired city as proposed by Goldmark in the early ‘80s. Great idea, but the next step after the wired is to cut the wires. Why build a house and run all the cables? While it is cheap when you are building a house to run the cables, what happens after the house is built and you want new cables? Things change. I want to move my computer around; I want it to be portable. It is very hard to have wires running all over the floors—people trip over them. Wireless is the answer. The idea that you cannot get high speed and such is the same thing as saying that transistors can never operate at high speeds and high power.
That is right, you mentioned over lunch that your textbook included a chapter on vacuum tubes.
Right. That is the first one I did on electronics. McGraw Hill required it and said no one would buy an electronics text (in 1966) without Vacuum Tubes.
International wireless technology in the 1990s
Based on your last remark, in one of your articles you wrote a joint article with Ray Pickholtz and Larry Milstein. In this article the phrase appeared that the 1980s was the decade of the computer. I think the article was written around 1990 or 1991. Then the 1990s would be a wireless decade or the communications decade. Do you think that has come true?
Yes. I do not know who wrote that. But definitely there is no question that the 1990s is the wireless decade. Wireless LANs to connect to your house or your office, cellular systems. You haven’t seen anything yet. The 2000s will be even greater. It will be a carry on with more finesse. Actually, cellular started in the ‘70s, but it did not take real hold until the ‘80s. The general masses of the population grabbed it in the ‘90s. It took hold more in Europe and Asia than it did in America where we have this super fine-wired system. In Asia, in Hong Kong, people are walking around with two cellular phones, one in each hand. They do not have as fine a wired structure. I’m not saying that Hong Kong is primitive by any stretch of the imagination, but as you go to more and more developing countries, it is simply because they do not have the wired structure that they are more dependent on the wireless structure. It works fine and the prices are lower, naturally, because the people are not as wealthy.
Is there one reason why Scandinavian countries like Finland have been prime movers in this technology in cell phones?
I do not know. If you look at Nokia, who you are referring to, and in Sweden you have Ericsson, those are two of the finest companies in the world. From talking to friends of mine, I think the Nokia phone has today far outdistanced itself from the Ericsson phone. I expect that Ericsson will meet the challenge, just like Motorola will meet the challenge and come back fighting. I think now you are going to see a lot more from the Japanese companies. The Japanese companies lost out in PCS and in cellular. The current standards: the GSM, narrow band CDMA, and even AMPS, are not Japanese. The Japanese are very bright and pretty stubborn, so they came up with their personal handy phones that few used outside of Japan so the companies could not make much money. In my opinion, to counteract that they have been the prime movers of third generation wide-band CDMA—the technology that no one else was using, and that they knew would be a good technology. They did something that was very smart: they paid western companies to build to their requirements, like Ericsson, etc. Once these European companies built and saw how well the system worked, they got on the bandwagon too. I was at a Standards Meeting at the very beginning of third generation where the Japanese representative said that if you guys don’t want to come along, we will do it ourselves. Generally Japanese are very polite. He was "un-Japanese" and more American in style, but it was what was needed to make everyone wake up and say, “He’s not kidding around. He’s a serious guy. We better join in with them or they are going to do it themselves.” The Japanese engineers are very bright, and they could do it themselves. They didn’t have to. The western companies joined them, and then only the U.S. lagged until we joined them, and then it became a worldwide standard. I think it was the right way to go, and it was the right way for the Japanese to go. In my opinion a lot of the credit for 3G goes to Dr. K. Tachikawa, who you are going to interview, who is the CEO of NTT DoCoMo.
Obviously, you know Dr. Tachikawa.
I think Dr. Tachikawa has blended his fine Japanese education with his fine American education and his understanding of American style to really put it all together for NTT DoCoMo. You will find he will be like most Japanese, very modest and not wanting to take all the credit. However, I believe that if he was not running DoCoMo it would not be the same story.
Predictions on the future of communications
By way of wrapping up if you could talk a bit about your ideas and your view of what the future will hold in communications, say in the next twenty-five years.
I find progress is exponential. If you look five years ahead, those changes will be as significant as the previous ten years. If you look over ten years and see what kind of changes you’ve had, then the next five years will evidence those kind of changes. My mom is ninety-three, and she remembers when there were few cars and there were horse-drawn buggies. Look what has happened over the century. Now we are in super jets. Things have changed tremendously. I believe what you are going to see is the wireless world. I think we are going to see high speed Internet access on a wireless base. These are all obvious to the casual observer. It is not going to happen in twenty-five years, but more or less within five to ten years and you are going to have really fine quality wireless cellular voice communication. Your fax will be operated through the computer and it will be rendered obsolete by e-mail. The post office will largely disappear or be an Internet Hub. Banking will all be electronic. Shopping will be done through the Internet. Our houses will be controlled from a "super” telephone system. Everything will be on one remote, but it will be a super remote. Look at the computers. We used to do punch cards in the ‘50s, and now I show you my Toshiba laptop, which weighs 2 pounds, and look at that power. Computers are 700 megahertz with twenty-five gigabytes of hard drive. A cellular phone is going to have the twenty-five gigabytes of memory with integrated circuit chips that are microns in size. All of that computer will be in your wireless "telephone." You will have a nice big display. The phones will open up so you will have a nice size screen. You can always hook into a screen, and I believe you will have screens everywhere so you can just walk over with your handset and plug right into anybody’s screen and operate. My friend’s Mercedes has two TV sets in the back for kids—there are VCRs for each one so they can watch their own separate movie or you can take outside videos. That car is eight years old. Limousines are all computer installed. You can have any kind of movie you want. The digital videodisk is going to be very small to fit into your computer. Your cars will be all set up with screens and displays. We already have cars that will detect being too close to another vehicle. You are going to have automatic driving. All of this wireless communication with computers is all within twenty-five years. I do not think you will have to wait twenty-five years for much of this to happen. I expect to be around to see it and be able to enjoy it.
What do you think some of the technical challenges will be?
There are always technological challenges to make things smaller and less expensive. Once people know it is has to be built, they build it. I do not think that is ever a problem. Once you can demonstrate that there is a market, people will rise up and build it. People even rise up to build it with there is no discernible market if they think it is going to be better for business in the future—they go out and build it and hope for the best. The Japanese did that, and they built the WCDMA, 3G, spread spectrum system and it works. I do not see any really major challenges. If you want to see a major challenge, go above the speed of light. After all, the speed of light is 3*10^8 meters per second on the average. Does that mean that sometimes it is larger and sometimes it is smaller? If it is larger, how much larger? Can we go faster than the speed of light? I do not think so. My narrow knowledge, which is very limited and makes me limited, says you can not ever reach the speed of light. Maybe that was just the way I was taught to think by people who were taught to think that way by people who were taught to think that way. Maybe when it comes right down to it there is no such limitation. Speed is usually an average indicator. If it is average, that means that sometimes it could be larger and sometimes it could be smaller. That would be interesting, wouldn’t it?
Textbook writing; economic expansion
Yes, it would. Are there any concluding thoughts?
I look at my textbooks over here, and the whole idea of textbook writing has changed dramatically. Textbooks are now all on the Internet and being downloaded. Your whole idea of teaching is changing. Movie theaters are changing. I think there is going to be a lot more competition. I think all of this creates a better economy. That is the one thing I would like to address. In our economy we are talking about when will this bubble burst? The bubble is not going to burst for ages because we are reaching a new plateau in our industrial revolution. The industrial revolution was done in the 1800s, and then it leveled. There were some changes. But the big change occurred and that raised everything. Then you had a leveling out. Now we are undergoing another "industrial" revolution. The stock market is not going to level out until the transient produced by that industrial revolution has ended and things level out again. We cannot even foresee the end of this industrial revolution. My belief is that the stock market, which reflects the increase and the growth of our economy, is nowhere near its peak. People are talking about the stock market Dow Jones of 30,000. Oh yes, easy—maybe more.
People are complaining that there is a labor shortage—there is no labor shortage; just look around the world. Many of our companies have labor forces overseas, and there are loads of people over there that are unemployed. They are studying engineering, and most people do not have to be engineers, they have to be laborers. They have to be able to sit there and make integrated circuits, wire up boards. The more sophisticated we become, the fewer laborers we need, the more our economy can expand. The labor shortage in America is meaningless. We have to get away from the idea of labor shortage for building things. We cannot build things here—we charge too much. I can pay a dollar a day in Mexico, China, Vietnam, or Cambodia. Why would I want to pay five or six dollars an hour in America? The labor is not that much better, if it is better. The work ethic of someone that is starving can often be better that the work ethic of someone who belongs to a union and is getting paid a very handsome wage. You cannot draw those kinds of generalizations.
I see the economy expanding like mad for at least another ten years. That is about as far as I can see. I imagine that in ten years, maybe we will level out; maybe not. But for the next ten years I think there is going to be a solid run in our economy, a solid run in our stock market, a solid increase in salary for our people, and well being for everybody. I attribute it all to our wireless economy, communications, and our computers. That is what is leading all of this. If you did not have wireless communications, could you have all of this? I don’t believe so.
- 1 About Donald Schilling
- 2 About the Interview
- 3 Copyright Statement
- 4 Interview
- 4.1 History of communications systems since World War II
- 4.2 Career overview
- 4.3 Spread spectrum technology and wireless systems
- 4.4 SCS and FCC demonstrations
- 4.5 Golden Bridge Technology and wireless standards
- 4.6 IEEE, Communications Society
- 4.7 Communications Society and technological progress
- 4.8 Ascendance of wireless as dominant technology
- 4.9 Brooklyn Poly researchers
- 4.10 International wireless technology in the 1990s
- 4.11 Predictions on the future of communications
- 4.12 Textbook writing; economic expansion