Oral-History:John Proakis
About John Proakis
Dr. John G. Proakis is Professor Emeritus and Research Professor at Northeastern University. He was a faculty member at Northeastern University from 1969 through 1998 and held the following academic positions: Associate Professor of Electrical Engineering, 1969-1976; Professor of Electrical Engineering, 1976-1998; Associate Dean of the College of Engineering and Director of the Graduate School of Engineering, 1982-1984; Interim Dean, College of Engineering, 1992-23; Chairman of the Department of Electrical and Computer Engineering, 1984-1997.
His professional experience and interests are in the general areas of digital communications and digital signal processing and more specifically, in adaptive filtering, adaptive communication systems and adaptive equalization techniques, communication through fading multipath channels, radar detection, signal parameter estimation, communication systems modeling and simulation, optimization techniques, and statistical analysis. He is active in research in the areas of digital communications and digital signal processing and has taught undergraduate and graduate courses in communications, circuit analysis, control systems, probability, stochastic processes, discrete systems, and digital signal processing.
He is the author of the book Digital Communications (New York: McGraw-Hill, 2001, 4th ed.) and co-author of the books, Introduction to Digital Signal Processing (Upper Saddle River, NJ: Prentice Hall, 2007, 4th ed.); Digital Signal Processing Laboratory (Englewood Cliffs: Prentice-Hall, 1991); Advanced Digital Signal Processing (New York: Macmillan, 1992); Digital Processing of Speech Signals (New York: Macmillan, 1992, IEEE Press, 2000); Communication Systems Engineering, (Upper Saddle River, NJ: Prentice Hall, 2002, 2nd ed.); Digital Signal Processing Using MATLAB V.4 (Boston: Brooks Cole-Thomson Learning, 2007, 2nd ed.); Contemporary Communication Systems Using MATLAB (Boston: Brooks Cole-Thomson Learning, 2004, 2nd ed.); Algorithms for Statistical Signal Processing (Upper Saddle River, NJ: Prentice Hall, 2002); Fundamentals of Communication Systems (Upper Saddle River, NJ: Prentice Hall, 2005).
About the Interview
JOHN PROAKIS: An interview Conducted by Michael Geselowitz, IEEE History Center, 10 June 2021
Interview #857 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
Copyright Statement
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, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. 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:
John Proakis, an oral history conducted in 2021 by Michael Geselowitz, IEEE History Center, Piscataway, NJ USA.
Interview
INTERVIEWEE: John Proakis
INTERVIEWER: Michael Geselowitz
DATE: 10 June 2021
PLACE: Virtual
Geselowitz:
Okay so this is Mike Geselowitz from the IEEE History Center and as part of our oral history program I’m here virtually interviewing John Proakis and particularly as part of our IEEE Signal Processing Society collection. John, I’d like to start with your early years, your origin, your early education, and how you got interested in technology.
Proakis:
Okay. Well, I have somewhat of an unusual background. I came to the U.S. in 1946. I was eleven years old at the time. I should tell you a little bit about my parents. My father emigrated to the U.S. in 1915 and stayed and worked here till about 1929, when the depression broke out. He returned to Greece, got married and because of the lack of jobs in this country, he started to farm in Greece. I was born in 1935 and my sister was born a couple of years after me. At that point, my father had plans to return to the U.S. with the family. He had become a U.S. citizen prior to 1929. However, his parents apparently became ill during this period. So, he stayed in Greece a little too long, and the war broke out. We were on the island of Chios in Greece during the war.
Geselowitz:
Oh, of course, Chios, yes.
Proakis:
Right. Chios is near the Turkish coast and we were under German occupation during the war. My father kept his American citizenship a secret of course. The Germans then left around September 1944 and a couple of days later the British fleet sailed into the harbor of the island. In any case, the earliest we could get back to the U.S. was 1946 and the U.S. government at that time provided transportation for the whole family, my sister, my brother and me and my mother and father. I was eleven years old. It was the summer of 1946. We started school in the fall of 1946. I knew not a word of English so I was put in the first grade and within one year my English improved, so I actually finished the first four grades. The second year was spent in the fifth grade and I skipped the sixth and finally went into the seventh grade. In the process I lost only one year in this transition, I was a very good student. I really loved mathematics.
Geselowitz:
Excuse me for a moment. Where were you living in the U.S.?
Proakis:
Okay, my father spent most of his life working in a steel mill so he looked for work when we returned in 1946. He looked for work in Baltimore and Pittsburgh and in the little town of Weirton, West Virginia, which is in the northern panhandle of the state, He found work there and so we settled in Weirton, West Virginia. My father worked in the steel mill for the rest of his life basically till he retired. I went through the school system in Weirton. In high school, I did extremely well in science and math and it was fairly clear that engineering was a good career path for me.
I have a very good friend who had recommended the University of Cincinnati which had a co-op plan. Part of the problem really was that my parents had very little money to send us to college. My father’s salary about the time when I started college, was around $5,000 a year, so you can understand that college expenses were of primary concern. The University of Cincinnati seemed like a great place for engineering studies. They had the co-op plan that made it possible for me to earn enough to pay for tuition and living expenses while I was in school. It took five years, typical of engineering co-op studies to complete the B.S. degree.
I should tell you something quite interesting In my class at the University of Cincinnati in electrical engineering there were about forty-five to fifty students and three of us applied to MIT, and all three of us were accepted, which I thought was very unusual because MIT is such a hard place to get in. I had an assistantship and so I studied from 1959 to 1961 in the MSEE program. I love probability, statistics, and basically applied mathematics, so I did extremely well at MIT.
By that time, I had been in school for seven years; five in Cincinnati and two at MIT and I thought I should get some industrial experience. I had several job offers. One was from Bell Labs, in which they offered to pay my salary and full tuition to send me to NYU for advanced studies. Another one was from RCA Labs in Camden., New Jersey. In any case, I chose to go to work for Lincoln Lab at MIT and that was probably the best decision I had ever made. I was in Group 34 at the time. It was headed by Paul Green, a very well-known fellow in telecommunications and Bob Price was one of the engineers in that group. Bob and Paul really were tremendous mentors for me and I learned so much from them.
At Lincoln Lab, I was put to work on a very interesting project at the time. A major problem during the 1960s was sending critical messages to submerged submarines. These were submarines that of course carried nuclear weapons and they generally did not like to come to the surface to receive messages, so the big problem was how do we get critical messages sent to them while they are submerged. Lincoln Laboratory at the time proposed to the navy that they use VLF communications, which is in the range of 15 to 20 kilohertz The navy had at least a dozen of these VLF stations around the world. In VLF signal propagation, the ground and the ionosphere serve like a wave guide and signals can travel around the world. One of the problems was, of course, to make sure that these messages were secure. So, we designed and built a highly secure digital communication system at VLF. I say we, actually I was a very junior person on this project. I didn’t do any of the design. We had a brilliant guy by the name of Paul Drouilhet and several other senior people that actually designed the system. It was a spread spectrum system and to appreciate what this means, the bandwidths available in VLF are something like a hundred hertz, so when you spread the signal over the band the actual bit rate was something like 2 bits per second. That is, we used spread spectrum for security and reliability, and so the bit rate was very low. To send a critical message required about 20 minutes. The other issue was that we wanted to make sure that the enemy, the Soviet Union at the time of course, could not knock out one of these transmitters that are around the world. The idea was that we would transmit the same message simultaneously from all the available VLF transmitters. It was a multichannel system basically, with all VLF stations transmitting the same spread spectrum signal at the same time, and that made the system hard to jam or to destroy. That was it. I was sent to a couple of the transmitting sites to install the transmitter exciters. One was in Hawaii and that was quite a trip. Another VLF transmitter was in San Diego. There was another VLF transmitter in Maine. Then, in March of 1962 I was sent to sea on a surface ship (Liberty ship) to actually test the system. Another colleague of mine was sent on a submarine to do the testing as well. At that time, the Liberty ships were deployed in the Atlantic Ocean, about 300 miles off the U.S. coast, to monitor all incoming aircraft coming into the U.S. from overseas. Typically, one ship was about 300 miles east of Cape Cod. Another was east of New Jersey, another was east of South Carolina, and so forth. They would basically rotate around, spend three weeks or four weeks at sea, and then come into port for a week. On a surface ship we actually towed an antenna that was submerged by about 20 to 30 feet below the surface and so we were doing all our recording with the submerged antenna. That was the project that I was involved in for about a year and a half at Lincoln Laboratory. The other part was actually doing some theoretical analysis working with Bob Price and some of the other engineers. Bob did some brilliant analytical work on the performance of multichannel communication systems like the one that we were actually deploying, and his work was ultimately published in IEEE journals. Because the multichannel system was a phase-coherent system, one of the issues that I worked on was how to obtain a phase reference from an incoming received signal and this work resulted in one of my early publications in an IEEE journal. That pretty much sums up my two years at Lincoln Laboratory.
I still had an interest in getting a Ph.D. While I was at Lincoln Lab. I wanted to strengthen my background in statistics and so I took two courses from Harvard University. This was an interesting situation because it turned out that MIT, in spite of the fact that they have an outstanding math department, really did not offer any courses in statistics and they had no statistics department. I went to Harvard part-time and I took two statistics courses from their statistics department. During that time, I met a young assistant professor who was a graduate of MIT, and he was teaching communications courses in the Division of Engineering and Applied Science. His name is Donald Tufts. Tufts convinced me that if I wanted to go on for a Ph.D. that I should consider Harvard. I applied for admission and I worked for Donald Tufts, as a graduate student with a research assistantship from September 1963 until the end of 1966 when I completed my Ph.D .
My studies at Harvard were also a great experience. Tufts had done a lot of interesting work dealing with intersymbol interference on wireline channels and that was a very hot topic at the time. Around mid-1965 you probably are aware that Bob Lucky came up with his automatic equalization algorithm, and so there was a lot of excitement about how to build modems that had the ability to adapt to the channel conditions. So, I did some work on that problem in my thesis, with Tufts providing the guidance.
In the final year of my Ph.D. research, I was doing some part-time consulting work for GTE Sylvania. So, after I completed my Ph.D. in December of 1966, I joined GTE Sylvania full-time in 1967 and worked there until the Summer of 1969. The Applied Research Laboratory at Sylvania had some outstanding people. So, the work environment was very stimulating. I worked in a department that was managed by Bruce Barrow. I don’t know if that name has come up in other interviews.
Geselowitz:
Well actually it has. Not in technology as much as in IEEE history because he was stationed in Europe for something or other…
Proakis:
Okay.
Geselowitz:
…for a while and he realized that there weren't sections in Europe, that IEEE hadn't made inroads there—it might have been back in the IRE days even or maybe it was IEEE…
Proakis:
It was IEEE, yes.
Geselowitz:
Yes. And so, he helped them organize. Even though he was an American he got Region 8, the IEEE Europe Region going.
Proakis:
Okay.
Geselowitz:
That’s how I know of him. Now, where, where was the GTE Sylvania Lab?
Proakis:
GTE Sylvania Lab was in Waltham, Massachusetts.
Geselowitz:
Okay.
Proakis:
His boss was a fellow by the name of Seymour Stein. I don’t think you have interviewed Seymour Stein. He should be interviewed by the way. That’s my opinion. He was really a great mentor for me. He had written a book with Professor Schwartz and Dr. William Bennett. Schwartz, Bennett and Stein was a great book in the mid-1960s on communication through fading channels and wireline channels.
Geselowitz:
I’m sorry, is that Misha Schwartz from Columbia?
Proakis:
Yes.
Geselowitz:
Okay.
Proakis:
Then there was another book that Stein published with a fellow by the name of Jones, so there was a Stein and Jones book. It was very good as well during that time. These were two books that I read carefully and learned a lot from. In any case, Seymour was a department manager and I believe that Barrow was a lower-level manager from him at the Sylvania labs.
The work at Sylvania was almost entirely focused on military applications of communications, e.g., the design modems for various types of channels with error correcting coding, et cetera. I was very interested in the equalization problem for fading channels and did some interesting work there. Besides Lucky’s algorithm for doing adaptive equalization there was another algorithm that was developed at Stanford. The algorithm was based on a mean squared error criterion; whereas, Lucky’s was a different type of algorithm. The Stanford algorithm was developed by Bernie Widrow and a colleague of his by the name of Hoff, so there was the Widrow-Hoff equalization algorithm. I started looking at that for applications to fading channels. That work resulted in a paper that was co-authored with a Sylvania colleague, James Miller, and published in the IEEE Transactions on Information Theory. This paper provided some very basic results on the use of the Widrow-Hoff algorithm on fading channels, which had not been done before. It has been widely referenced in the technical literature.
Geselowitz:
When did you first become aware of IEEE and its publications? Was this your first IEEE publication? Or?
Proakis:
No, my first publication actually was my master’s thesis at MIT, which was on sequential search for radar systems. I worked in the Electronic Systems Laboratory at MIT at the time. This was during the period 1959-1961. The work on my masters thesis was published as a paper in the IEEE Information Theory Transactions, I think in 1963 or 1964. Then the second paper was actually on the phase estimation problem and that was published around 1964. By 1969, I had a couple of other journal publications from my Ph.D. thesis.
Geselowitz:
And were you already an IEEE member in those days?
Proakis:
I was an IEEE member. First, I was a student member at Cincinnati in the 1950s.
Geselowitz:
Oh, okay. Great.
Proakis:
Yeah, and now I’m trying to think. At that time, it was still the IRE, right? The merger didn’t take place until the early sixties as I recall. Is that about right from your…?
Geselowitz:
Yeah, 1963 was the merger.
Proakis:
Right, okay. I was an IRE member up to the merger. At the Sylvania Laboratory there where several consultants and I got to meet them. One was Irwin Jacobs. Irwin was an Assistant Professor at MIT and he would come into Sylvania from time to time for consulting work, so I got to know him. I learned at the time that he was planning to leave MIT and go to the University of California in San Diego. That was one acquaintance. Another acquaintance, also a consultant was a fellow by the name of Harold Raemer. Harold was the department chairman at Northeastern University, Department of Electrical Engineering, and so I got to know him quite well. At some point, I mentioned to him casually that I had an interest in teaching at some time in my career. Around the spring of ’69, he came to my office and said that there was an opening at Northeastern in Electrical Engineering and would I be interested in applying for it. I said oh sure, I’d be very interested in applying. It turned out that they had offered the position to Lou Franks. Lou Franks actually at that time had two offers. One was from the University of Massachusetts in Amherst and the other one was from Northeastern. He chose to go to UMass and that’s how the opening at Northeastern became available. I applied for it and got it.
Starting in September of 1969 I began my teaching career at Northeastern University. About the same time Seymour Stein decided that he would venture on his own and form his own company. He said John, are you interested in joining us? I said listen Seymour, I just accepted this position at Northeastern. I said I’m going to follow through with that decision , but I’d be happy to work part-time, and so I ended up being a consultant for Seymour Stein. His company was called Stein Associates and I did a lot of consulting work for him throughout the 1970s.
One of those projects that I remember clearly was something that I got involved in for the first time. It was on underwater acoustic communications. The navy was very interested in funding programs that apply novel modulation and coding techniques to the underwater channel. I knew very little about underwater channels at the time but along with other engineers at Stein Associates, we designed a very novel communications system that was based on multitone frequency-shift keyed modulation (FSK), actually multichannel FSK. The underwater acoustic channel has significant multipath components that you have to deal with and so the idea was that if you were sending a bit, let’s say on one of two frequencies, you wanted to make your symbols much longer than the multipath spread of the channel so that you would have very small intersymbol interference compared with the bit signal duration. By removing that part of the tail or the front end of the signal which had multipath interference, you can completely wipe out the intersymbol interference. If you had a certain bandwidth available, what you can do is subdivide that bandwidth into multiple frequency slots and simultaneously transmit different bits on different frequency slots. We married that with coding. Another consultant on the project was Jack Wolf. Jack at the time was at University of California in San Diego and so he proposed some interesting Reed-Solomon type codes which were very, very effective in correcting channel errors. In 1978 we published a very interesting paper in the IEEE Transactions on Information Theory where I was a coauthor with Jack and another couple of people from Stein Associates. . That was my introduction to underwater acoustic channels at the time. About fifteen years later, I returned to the topic of underwater acoustic communications. I’ll say more about it as we continue on.
As I mentioned, I started teaching at Northeastern University in the Fall of 1969. I began with teaching communications courses. Harold Raemer, the department chairman, believed that all faculty should teach not only in their specialty areas, but they also should teach some other undergraduate courses so they get a sense of what the curriculum looks like. I was appointed to teach circuit theory and then linear systems, signal processing, et cetera. So, I became aware of, and really learned a lot about, the undergraduate curriculum, and that was something that really helped later on in my career when I became department chair at Northeastern.
Looking back, in the seventies there was a very severe downturn in the economy. Around 1972, 1973 Northeastern began to see lower enrollments and their program at the time was heavily dependent on tuition. The administration made some drastic reductions in staff. They also imposed a quota on tenure, which was no more than 60 percent of the faculty in a college could have tenure. In the college of engineering, there were only two tenure slots available from 1972 to 1975. When I came up for tenure in 1975, it was questionable as to whether or not there would be a slot available. Fortunately, things turned around fairly quickly. I did get my tenure in 1975 and I continued on beyond that.
In the communications area during the 1970s, a major development was Andy Viterbi’s algorithm for decoding convolutional codes. Viterbi’s paper was published in 1967. However, it took a few years before communications engineers understood and appreciated the impact of this new algorithm in the field. This is a maximum likelihood algorithm and has applications not only to decoding, but also to channels with intersymbol interference. Dave Forney, who you’ve interviewed, was involved in building modems at a company called Codex Corporation. Dave recognized the importance of the Viterbi algorithm in our field and in 1972 published an outstanding paper on its major applications, including the problem of suppressing intersymbol interference. I was following these developments, and I was focused on using the algorithm to combat intersymbol interference on fading channels. In 1975, Andy Viterbi had undertaken the task of editing a special publication of a book on advances in communications and he asked me to write a chapter in that book on the application of different algorithms for intersymbol inference cancellation. I contributed a nice tutorial treatment on the performance of the Lucky algorithm, the Widrow/Hoff algorithm, and the Viterbi maximum likelihood algorithm. That was one of the highlights of my work during the 1970s.
The other major development in the communications field during the 1970s was the work of Ungerboeck, who demonstrated for the first time that coding applied to band-limited channels such as telephone channels provided significant performance improvement. His scheme is called trellis-coded modulation. This scheme opened up another avenue for research into equalization algorithms for channels in which trellis coding is used.
In my consulting work during the late 1970s, I got involved in another interesting area which I knew very little about, that is magnetic recording channels. The large computer company in the Boston area at that time was Digital Equipment Corporation, and they had a research lab in Westborough. I was contacted by a fellow by the name of Roy Gustafson who was heading a group that was just beginning to look at the application of equalization algorithms for magnetic recording channels. The problem is that in the readback channel you have intersymbol interference and so the same kind of techniques that we developed for telephone channels also had application to magnetic recording channels. At that time, I helped design a large simulation program that simulated the readback process of the magnetic recording channel. So, we were able to simulate different types of signals and actually look at how they performed using different types of algorithms in the decoding of the read back signals. This was a very interesting area for research, so several of my graduate students at Northeastern worked in this area toward their PhD degree.
In 1982, I had my first initiation into academic administrative work. I was appointed as Associate Dean for the Graduate Engineering Program at Northeastern University. The former dean had retired and the Dean of the College of Engineering asked me to take over the position.. This was a critical time in engineering education because the electrical engineering industry throughout the country was having difficulty getting enough qualified graduates from engineering schools to fill the positions that they had. It was the beginning of a tremendous expansion in the computer industry and in VLSI design and applications. Northeastern University, of course, was very heavily involved throughout its history in continuing education. We had the suburban Burlington campus, where I did some evening teaching, which provided part-time graduate courses for engineers in industry. Northeastern also had four TV channels that were authorized by the FCC to be used for TV in the instructional television frequency band. So, during the time that I served as associate dean, we had a lot of interest from companies who wanted faculty to come into their individual plants and teach courses. We just didn’t have the faculty resources to satisfy the demand. We decided to use the TV channels for graduate instruction. I don’t want to take credit for inventing something that I did not do, because Stanford University had pioneered the system of using instructional television to reach out in Silicon Valley to various industries that needed continuing education courses for their engineers. I thought this was a great opportunity for Northeastern to do the same.
Geselowitz:
Okay.
Proakis:
Anyway, we proposed that to local industry. The industry person that was the champion of this approach was Ray Stata from Analog Devices. Analog Devices had hired many of our graduates and also co-op students from Northeastern, so Ray knew us well and as I said, he held several meetings with local industry people, basically drumming up support for this kind of continuing engineering education. So, in 1984 we began our instruction television system in which we installed receivers at various plants throughout the Rte128 area, and all the way out to Rte. 495. . We put our system into operation at that time and it served us well for many, many years until more recently when streaming video basically has taken over. That was the beginning of another venture from Northeastern which worked out quite well and I was involved in supporting it.
I have to tell you a funny story. We had a problem with getting some faculty to teach on the TV system; particularly, some of the old timers were very concerned about job security. We recorded the courses in case some of the people working in industry were on travel and they needed to have a backup. So, we would send the video recordings to their plants and they could watch them. The faculty were concerned that once we recorded their course on video that we would not need them anymore. We could just run the video courses over and over again for year after year. We assured them that that was not our intent. In addition, we provided a fringe benefit. We said that every time you teach a course on television, and this was a course being taught to our full-time students in class at Northeastern at the same time that it was transmitted out, the university will set aside a thousand dollars for professional development. You can use this money for travel to conferences and for support of graduate students in your research. That made it more palatable and they liked that idea. That program was a great success.
The chairman of the electrical engineering department had stepped down and that happened in 1983. I was at that time running the graduate program. There was a nationwide search for a person to fill the position. It was ultimately offered to a fellow from Harvard University who really wanted the position, but then he began to look at all the fringe benefits he had at Harvard and finally decided to stay at Harvard. The position then was not filled in 1983 and there was a second search that started in the fall of that year. At that point in time, some of my colleagues suggested that I apply for it and I did. I was offered the position and starting in the fall of 1984 I became department chair.
At that time, we had a tremendous shortage of people teaching in the computer engineering area. Also, at that time, Northeastern established a separate computer science department and the program in electrical engineering was expanded to include computer engineering. We had to hire people in computer engineering and there was a tremendous shortage of faculty because industry was paying much higher salaries for qualified candidates. At the same time, the enrollments were very, very high in electrical engineering. In fact, about half the college of engineering students were in electrical and computer engineering. In any case, we began to expand the size of the faculty in 1984.
During this time period, there was a movement in the Boston area to try to compete with Silicon Valley in the VLSI design and computer engineering area. Several industry leaders, including Ray Stata and George Kariotis, lobbied the Commonwealth of Massachusetts to establish a fab facility in Massachusetts, which would serve the educational needs of universities in the state. This effort resulted in the creation of the Massachusetts Technology Center, in Westborough. Major computer companies in the state, such as Digital Equipment Corporation, donated computer workstations and VLSI design tools to the engineering schools in the state. These donations allowed us to teach VLSI design courses to our students. The circuit designs were then sent out to the fab center in Westborough for fabrication. These circuits would then come back to the universities and to the students for testing. That was another major activity during the time that I began the department chairmanship. We managed to hire several faculty to teach VLSI design and circuit fabrication.
Geselowitz:
Before you move on, at this point you're known. You mentioned your deanship and now we’re into your chairmanship. You're known as an educator, you won the IEEE Education Award. Hadn't you already written a textbook by this time?
Proakis:
Well, yes, I did.
Geselowitz:
I don’t know if you want to say something about that.
Proakis:
Yes. I started writing the first book in 1981. Of course, I had written class notes for my students taking communications courses. Book publishers came to the university periodically and knocked on my door asking if I had anything interesting to publish. In 1981, a fellow from McGraw-Hill came by and visited me and I said I have some class notes in the communications area. He suggested I think about writing a book. I started working seriously at it. At that time there was no word processing capability so everything had to be done on a typewriter. Fortunately, I had an excellent typist in my wife, who was my “personal secretary” as well as the mother of our two children. She was an excellent technical typist, and she typed the one thousand-page manuscript, which went out to McGraw-Hill and was published in 1983. It was entitled Digital Communications. That was my first textbook. It did extremely well.
Let me mention a couple of other textbooks that were available at the time that my book was published. There was a book by Lucky, Salz, and Weldon, published by Bell Labs, that was a very basic textbook on telecommunications focusing a lot on wireline channels. There was also the book by Wozencraft and Jacobs, which was just an outstanding book that was published back in the late sixties. This book served as a textbook for many undergraduate and graduate courses at universities. However, the Wozencraft and Jacobs book was not focused on adaptive equalization issues and fading channels. My book was really unique at the time and it fulfilled a need. So, it was used at many universities. A Second Edition came six years later in 1989, Third Edition came in ’95, and the Fourth Edition in 2000. It’s now in its Fifth Edition, published in 2008. For that I have a coauthor, my Northeastern colleague Masoud Salehi. Masoud is also my coauthor on an undergraduate book on communications. The latter editions of Digital Communications have been translated into half a dozen languages including Russian, Japanese, Chinese, and Korean. So, the various editions of my digital communications book have been used around the world and ultimately led to the IEEE Education Medal that I received in 2014.
The other area in which I published extensively, and in which I have several textbooks, is digital signal processing. The pioneering book in DSP was written by Oppenheim and Shafer in 1975. Oppenheim is a good friend of mine, dating back to the mid-sixties. We used his book for graduate courses at Northestern. However, in the early 1980s, his book was too advanced for undergraduate students. During this time, there was a need to begin teaching digital signal processing at the undergraduate level. A new book appeared by Ken Steiglitz from Princeton University at the time. It was a small book, and we used it for a couple of years to teach DSP at the undergraduate level. That was the motivation for me and a coauthor by the name of Manolakis, to publish an undergraduate DSP book, which was published in 1988 by Macmillan. That book just now came out in its Fifth Edition; it's published by Pierson Education. It's very popular for undergraduate courses. Now, it has expanded in coverage so it's suitable for both undergraduate and graduate courses. Of course, DSP is a basic tool in the implantation of communication systems. So, the two areas are closely related. I have also published books on the use of MATLAB in filter design in DSP and in digital communications. In all, I have published about ten different books. Surprisingly, I managed to find time to do that. I have enjoyed writing.
Now, to move on to the nineties. I remember being contacted by a fellow from Woods Hole Oceanographic Institution who was again very interested in underwater acoustic communications. As mentioned previously, I had done the work on underwater acoustic communications for navy applications back in the seventies, but I had not done any work since then. The individual at WHOI was interested in applying adaptive equalization techniques to underwater acoustic channels as a means for getting higher data rates. He was also interested in transmitting wideband acoustic signals and possibly video signals. I was appointed a guest investigator at WHOI and I was funded to support a couple of graduate students who did Ph.D. thesis work at the time. We designed some very powerful adaptive algorithms for equalization of wideband acoustic signals. We tested them at sea with the use of WHOI ships and they were very effective both in shallow water and in relatively deep water. That led to a number of projects that were funded at WHOI and Northeastern by the navy. We did get a patent on this technology for adaptive equalization for high rate underwater acoustic communications which was jointly owned by Northeastern University and WHOI. The two institutions encouraged us to start a company and to commercialize this technology. We did do that and we approached Raytheon and some other companies for funding. They were interested in the applications of course, but they were not interested in giving us funding to start and grow this technology. It turned out that the business opportunities were quite limited. I mean, who is interested in this type of technology? -- the U.S. Navy, obviously, and oceanographers doing underwater measurements. Oil companies to some extent, are interested in being able to turn on and off all kinds of systems that are underwater in deep water oil exploration; but again, the applications were quite limited and there was not sufficient business to support a company. We basically gave up on that aspect of it. Of my students that graduated from Northeastern with Ph.D.s with this background, one is now a professor; while the others went on to work in industry. The field of underwater acoustic communications has expanded so there's now a lot of interesting work being done at several universities that have facilities to do underwater acoustic transmission.
The early 1990s ushered in the era of cellular communications. . In 1989 I was appointed to a committee formed by the Cellular Telecommunication Industry Association, CTIA. This committee consisted of five academics. At that time, there was a big controversy among the members of CTIA as to what type of system they should recommend for multiple access on cellular channels. The committee had no bias toward any of the systems that were being proposed. One of the systems was FDMA and the other was TDMA. TDMA was the one that ultimately became the GSM system for Europe. The new system that was at its infancy at the time was CDMA. Irwin Jacobs and Andy Viterbi formed Qualcomm in 1985 and built a CDMA system for communicating with truck fleets. They had started work on using CDMA for cellular telephone but in 1989 this was not an option for the committee to consider. CTIA just gave us a choice of either TDMA or FDMA to recommend. After six months of study we recommended TDMA and that ultimately led to the first cellular system that was built in the US. Qualcomm started getting into the picture with proposals for IS95, their CDMA system by 1990.
By the mid-nineties, I began to see patent litigation regarding the new cellular telephone technology. One of the first big trials that occurred was when a small company in Philadelphia sued Qualcomm on CDMA technology. I was an expert witness for Qualcomm on this trial. It was held in Philadelphia. I had an interesting sidebar session outside the courtroom in which the judge asked me questions on some of the technology. It turned out that the company that was suing Qualcomm had a document for a CDMA system which was marked Qualcomm proprietary. This document was found in the company’s file cabinet during discovery. Shortly after the date on this document, the company applied and was granted a patent on a CDMA system which was almost a carbon copy of this Qualcomm CDMA system diagram. So, the judge wanted to know what the similarities were between this company’s patent and the Qualcomm document. I started explaining that the Qualcomm document was a block diagram of the CDMA system that showed the functionality of each of the blocks. After I went through about a half a dozen of these blocks and explained to the judge the functionality of each of these blocks in the block diagram, he asked me, who is Mr. Block? This is just a funny story because it shows you that these judges really should not be involved in these highly technical patent cases. The trial began and Irwin Jacobs was listening to the testimony and watching the jury members, who had no technical knowledge of the CDMA technology. As Irwin Jacobs was watching the people in the jury and seeing that some of them were falling asleep when there were some technical discussions on particular issues that we had to go through, Irwin decided this was not a good sign and the outcome of the trial could have gone either way. So, he entered into negotiations into a settlement, and finally settled for $5 million. I worked on many patent infringement cases over a period of about twenty-five years or so. Due to having written textbooks in digital communications and DSP, a lot of the patent technology would cite my textbooks. So, I was contacted frequently for this legal work.
In the Summer of 1997, I stepped down as department chairman and at the end of that year I decided to retire from Northeastern University. I was sixty-two at that time. I still had some consulting work during that time with the navy lab in San Diego. So, the winter of 1998 I convinced my wife to take a leave of absence from her job and we went out to San Diego and spent three months there. I did my work with the navy down in Point Loma, and was working with a fellow by the name of Joe Rice who was funding my work at the university at the time. My wife and I were overwhelmed by how beautiful the weather was during January, February, and March and at the end of the three-month period, I convinced her to retire as well. I had several friends at the University of California in San Diego. Jack Wolf, Larry Milstein and a number of others convinced me that I could become an adjunct professor and do research without having to teach or serve on committees. I took that option and that became my second career. It lasted for about twenty years. Over that period, I had several grants and a number of Ph.D. students. It was really a rewarding experience. I was able to divide my time with about five months of the year in San Diego and seven months back in Boston. During that period of time, I collaborated with Scripps Oceanographic Institution on a major five year grant funded by the navy. The work involved modeling of the underwater acoustic channel and basically continuing some of the earlier work that I had done at WHOI.
Geselowitz:
Are you still doing that work today?
Proakis:
No, in 2018 I came to the conclusion, Mike, that I could no longer compete with some of the younger people in writing proposals for grants. At UCSD, I had three major grants, each of which lasted for five years. They were all DOD research funded by the navy and the army. These grants were called multi-university research initiatives, or MURI grants. Over the past several years, the communications field had changed where today there is much more focus on communication networks, whereas my work was focused on physical layer research. I was over 80-years old and it was time to stop. However, I did continue my book writing. The Fifth Edition of my DSP book was just published by Pearson. Today, I can tell you that I’m now fully retired. The pandemic hit us very hard. I’m a social guy, I have lots of friends, and during this time, my wife and I had to isolate ourselves from everyone.
Geselowitz:
Well, hopefully that’s ending.
Proakis:
It is now. We received our shots in February. The big concern for me was going to the supermarket once a week and having to expose myself to this and making sure that when we visited the grandchildren, we only did it when we were all outside and we were all wearing masks. It made it very, very hard.
Geselowitz:
Oh yes.
Proakis:
Hopefully things will get back to normal, certainly by this Fall.
Geselowitz:
Yes, that’s the hope.
Proakis:
Yes.
Geselowitz:
That’s really a fascinating story. I have just a couple of questions, some general questions and then I’ll let you have the final word. Let me just say, I’m just curious, what does your wife do when she’s not typing your manuscripts?
Proakis:
She started doing technical typing work when she got out of high school. She worked at Hanscom Air Base and there was a very famous German-born scientist by the name of Hans Plendl, who had done some outstanding work on ionospheric propagation. She was his secretary and she had to learn to do technical typing. She was an outstanding typist; very fast and very, very good, very accurate, so when she typed my book it was a thousand-page manuscript. At that time there was no word processing, so every time I needed to change a paragraph, the whole page needed to be retyped, so she actually did about 3,000 pages of typing. At that time, she said this is the end. From now on, if you need a typist find somebody else. That was the end of my wife’s career as a technical typist. She wanted to devote all her time in raising our two children. Fortunately, my sister, who is a high school teacher, taught business subjects and she wanted to introduce her high school students to technical typing, because many of her students went onto become secretaries at various companies. So, she started teaching them technical typing. By that time there was a program called TechWriter that she started using. It was a Latex software package. She then became the person that did all of my other textbook typing. I did pay her for her work.
Geselowitz:
Did your wife stay in education or what did she do?
Proakis:
A few years after she graduated from UMass Boston, she was a school teacher for a while in the Lexington school system, but at some point in the early eighties, she was the last one in and the first one out when they started laying off people so she ended up getting a full-time position with a company called DC Heath, which was a Raytheon book publishing company. In 1995, DC Heath was bought out by another book publishing company called Houghton Mifflin. She stayed with this company until her early retirement in 1999.
Geselowitz:
Oh, interesting. And how many children and grandchildren do you have?
Proakis:
We have two children a boy and a girl. My son shares with me my birthday. June 10th which is today.
Geselowitz:
Oh, happy birthday. You didn’t tell me that on top.
Proakis:
Thank you very much, Mike. I thought at first that I would decline doing it on my birthday, but then I decided I don’t have anything else to do. My birthday party is tomorrow night at my sister’s house. My son shares my birthday. He called me just before I got on here to wish me a happy birthday. He's forty-nine years old. My daughter is two years younger. My son actually studied civil engineering at Northeastern University. He then went to the Maxwell School at Syracuse University for a degree in Public Administration and to MIT where he earned a degree in City Planning. He is now the director of City Planning group in the city of Somerville, Massachusetts. This is a very interesting job for him. He and his wife have one daughter.
Geselowitz:
Oh, I’m sure.
Proakis:
My daughter was very good in mathematics, but she did not want to become an engineer. She went to Wesleyan University in Connecticut, a small liberal arts school. They had a scientific program, so she studied Earth Science. The funny story that goes with that is that when you try to give your children some guidance, they won’t take it from you, they’ll take it from somebody else. When she was in elementary school she was chosen to take part in this math Olympiad program. She was the only girl among several boys in this math Olympiad and I said to her “since you're very good in math, when you go to Wesleyan make sure you take a lot of math courses. Whatever you decide you want to do as an undergraduate, make sure you take a lot of math courses.” After her first semester at Wesleyan, she told me that her advisor told her she didn’t need a lot of math. I said “what do you mean?” She said that he advised her to only take linear algebra. I said, “my God”. She had started calculus in high school. I told her to take some more calculus courses. No, she wouldn’t listen to me. Anyway, she received her degree with Phi Beta Kappa Honors and a double major which included music.
She wanted to work for a couple of years before going to graduate school. She interviewed and was offered a position at Camp, Dresser, and McKee, which is a large civil and environmental engineering firm in Boston. After a couple of months on the job, she told me that all the interesting work in this company was being done by engineers. When I asked her what the meant, she told me she wanted to study environmental engineering in graduate school. I said great. She worked there for two years, Mike, and after that she applied to MIT, to University of Texas in Austin, and to Berkeley. She was accepted in all three. Berkeley offered her a full two-year scholarship in their civil and environmental engineering department. She ended up going there. She had to take all of the math courses that she should have had as an undergraduate. Her advisor at Berkeley was so impressed with her knowledge of music that he invited her to become part of his little band – which she did. She received the masters degree in civil and environmental engineering at Berkeley and returned to Boston to work for Camp, Dresser, and McKee. She is now working for the City of Melrose in Massachusetts, where she is the head of the Department of Public Works.
Geselowitz:
Wow. Great.
Proakis:
So, she is doing engineering.
Geselowitz:
Right. Engineering management.
Proakis:
Yes, engineering management. She’s forty-seven years old. She and her husband have two children.
Geselowitz:
Terrific. I have one obligatory question and there's another question partly mine. The obligatory question is, besides publications, what, if any, was your involvement with IEEE over your career?
Proakis:
Well, in the Boston area, I had organized and taught courses to raise money for the IEEE section and of course, at the national level I did some associate editor work for The Information Theory and The Communication Transactions, so I did both of those. I gave talks at various IEEE functions. I was one of the people that were nominated to do lecture tours and I did a lecture tour in Mexico, in Puerto Rico, and one in the US.
Geselowitz:
You, yourself were a student member way back in the day. Was there a strong student branch at Northeastern?
Proakis:
Oh absolutely. We had a very strong branch of IEEE in electrical engineering, as well as a branch in HKN, the honor society. HKN did all of the tutoring. A lot of the students, when they need some help in courses, they hesitate to go to the faculty member to ask questions. They’re shy or they're afraid they might seem stupid, so they prefer to ask their classmates. So HKN ran a tutoring session for all the undergraduate courses and they still do that at Northeastern.
Geselowitz:
Great. Great. My last question John with you is, and this is just sort of a philosophical historical question I guess, but I’m sitting here listening to you and you’re talking about all of this activity in the Boston area continuing that you're involved all these companies, you named Ray Stata and all these folks, the tech center, but sort of the public narrative is that yes, Route 128 was a pioneer back in the sixties, but that as Silicon Valley grew it eclipsed all other technological activity. Now actually you ended up on the west coast part-time, but I’m wondering if you have any thoughts about Silicon Valley versus Boston as tech centers, and if there's like a competition or if Silicon Valley is overrated. How do you feel about that?
Proakis:
I don’t think Silicon Valley is overrated. I would say that during the eighties when the computer industry was booming that we actually had a mini-Silicon Valley kind of operations here in the 128 and 495 area. There certainly was a serious attempt to build up local industry and improve the educational infrastructure in this area, especially at places like Northeastern. I never really considered the Boston area as second rate in any way. I believe that we had first rate companies, we still have first rate companies and the technology here is just as good as it is out in Silicon Valley. It just turned out that Silicon Valley is much larger and attracted a much larger clientele. Maybe it’s the weather, I’m not sure. Of course, California is much larger than Massachusetts and it has a number of fine universities. In the Boston area we have Boston University, Northeastern, Tufts, and of course, MIT and Harvard, which are comparable to Stanford and Berkeley. In any case, I think that the density of the population is much greater in Silicon Valley and so the industry then grew much, much larger. I think in terms of quality and certainly in terms of living conditions, I think the Boston area is just as nice.
Geselowitz:
I lived in Boston area for eighteen years.
Proakis:
Yes.
Geselowitz:
Love it. Okay. That was an amazing discussion. I just want to leave the last word with you. Anything I didn’t ask that you’d like to say or that you’d like to say of not being asked.
Proakis:
Mike, as I look back on my career and I have shared this with many of my friends, I consider myself so fortunate to have lived throughout this period. As I look back, Claude Shannon’s pioneering accomplishments provided the motivation for all of the work that has been done in telecommunications over the last sixty years to seventy years. I have lived through this period on a day-to-day basis and experienced all of the major developments that have occurred in my field. It has been nice to be a part of all that. My contributions were certainly not as great as some of the people that I know and worked with, but when I look at Irwin Jacobs who’s a friend, Andy Viterbi who’s a friend, Dave Forney, these are all brilliant people, they accomplished so much. I had a kind of a nice life just following some of these new developments and most important, writing about some of these great developments so that they can be used to educate other people. The book writing, I consider as my number one accomplishment, the educational part of it.
Geselowitz:
Right and even despite all of your amazing technical achievements that’s why you were recognized within IEEE with the education award. Thank you so much for all you do and have done and thank you for the interview. I’m going to turn off the recording if I can figure out how to do that.
Photo gallery
Receiving the 2014 IEEE James H. Mulligan, Jr. Education Medal
IEEE lecture tour to Guadalajara, Mexico
IEEE lecture tour to Guadalajara, Mexico
IEEE lecture tour to Guadalajara, Mexico
IEEE lecture tour to Guadalajara, Mexico
