Customer Reviews

Digital Communications (Mcgraw Hill Series in Electrical and Computer Engineering)

5 star:		(10)
4 star:		(6)
3 star:		(7)
2 star:		(6)
1 star:		(1)

 

 

Many of the reviewers criticized the book, largely because of it being difficult to read. However, the real question a potential reader should ask him/herself is whether the book is intended as a reference manual or as a textbook for learning new material. As a textbook, the pace is too fast, there are few examples, and the level is too complicated for the novice reader. However, as a reference manual, for the practicing engineer or researcher, this book encompasses a vast span of material and is extremely useful. The potential buyer must determine whether the book is intended as a textbook or reference manual. As a reference manual the book deserves the full five stars, but as a textbook, two stars are more than enough.

this is a great reference if you already know digital communication, but definitely not something that you read cover to cover. i hated this book when i was first forced to study from it, but on a second glance and a few years of research in communication theory, i like it quite a lot. definitely good for refreshing your memory or learning small things, but not for learning communication theory. you're much better off with the Proakis/Salehi "Communication System Engineering" book.

This is an excellent graduate level treatment of communication theory. This book is not about communication system engineering. Rather, it is about hard core communication theory. The book follows the topical organization established in three previous editions with minor modifications, mostly new added material on channel codes and transmit-diversity through the use of space-time codes. It has the usual first chapters on probability, random process theory, the sampling theorem, and bandpass processes before it launches into the heart of the subject which starts with optimum detection of signals in classical AWGN channels, estimation of signal parameters (viz. frequency, phase, symbol time). Interestingly, the estimation of signal amplitude is not covered even though it is a critical parameter for the demodulation of QAM signals. The book next takes a minor detour to introduce Shannon theory and channel coding for error control before returning to modulation-demodulation. The band-limited channel is taken up next. Signaling waveforms that have either zero or controlled (that means small) intersymbol interference (ISI) are covered, as is the reception of signals passed through band-limited channels by means of maximum-likelihood sequence estimation (MLSE) and various equalization approaches. However the issue of tracking a time-varying channel and the required speed of adaptation for doppler spread channels, such as are encountered in UHF and microwave mobile communication systems, is not addressed. This book is about fundamentals. Higher dimensional signaling, under the guise of multi-channel and multi-carrier communication is nicely introduced, including the FFT multicarrier method used in xDSL systems. A chapter is devoted to introducing direct-sequence and frequency-hop spread-spectrum signaling and code division multiple access (CDMA). The next chapter covers the practical problems of communicating through channels that exhibit fading due to multipath. Spatial diversity receive processing and transmit diversity, aka space time coding are covered, but as mentioned above, fast fading channels are not covered. The last chapter is on multiuser communication but focuses only on channel access methods. The book would be better if the last chapter covered optimal demodulation of signals in channels that are impaired by fading plus AWGN plus cochannel interference (CCI), which would have lead naturally to a discussion of multiuser detection. The book then would have a pedagogic structure leading from the simple "known-signal-in-AWGN" channel through a hierarchy of increasingly difficult channel impairment models. Comparing the fourth edition to the first, which was published in 1983, it is gratifying to see how the book has evolved to stay up with current trends. Minor technical improvements are visible too, such as the elimination of the complementary error function erfc(.) in favor of the complementary cdf of the standard normal distribution Q(.) in error probability formulas. This book will serve today's students of communication theory well, as did its predecessors. Anyone who masters this book will be quite well prepared to move into any digital communication specialty field such as satellite communications, wireline communications, xDSL, mobile wireless communications, 3G, fixed broadband wireless, free-space optical and optical fiber.

Stephen D. Stearns
TRW Electromagnetic Systems Laboratory
Sunnyvale, CA

I was told by one of my friends that this book has been reviewed in this site and obtained five-star comments.

Having used this book as a text book for a full term, I really enjoyed nothing from the book, except for, maybe, its complete coverage on various topics and the depth it touches on (,which is the reason I still give it a three-star evaluation).

The author seemingly knows nothing about the essence of the concepts he tries to develope in his book ( although this in fact should not be true given the fact that he is such a famous person in the area, his writing makes him to look so). He extensively uses complicated mathematics to develope the ideas, without any instructive interpretation of those formulas. Mathematical derivation is certainly necessary in such kind of books, but a conceptual insight into the formulas is indeed more important than the math itself. If it were Ronald Bracewell who wrote the book, it would contain 10 times more insights into the concepts and 10 times less mathematics to confuse readers.

Usually for a book written in such a style, one would think it should be mathematically rigorous (as a trade-off of being boring). However , in this book even the rigorousness is not fulfilled. There are loosey-goosey derivations everwhere. Furhtermore, erroneous illustrations can be found throughout the book (some are even as the title of a section). This wouldn't be a surprise to people who are familiar with his writing style: in one of his other book on DSP, things are even worse.

To make my comments more concrete, I will just mention a few such examples. When he derives the power spectral density for a digitally modulated signal, there are quite a lot confusion he made in terms of the linkage of the psd of a digital signal and that of an analog signal, and even the autocorrelation function was oddly formed without any explanation. When he expains the Nyquist criterion for zero ISI, the Fourier theroy was poorly developed, I would assume Bracewell will just use a few figures, without any math, to do a much better job. In discussing the channel capacity, a erroneous statement is made on the title of a section, saying "achieving the channel capacity with orthorgonal signals", where he messes up the concepts between channel capacity and the Shannon limit (Eb/N0) for an arbitary small transmission error.

So in general, I do not give too much credit to the author, and strongly object using this book as a text book. Nevertheless, you can use it as a reference (since it is rather compelete) to get some rough understanding on various topics, but, keep in mind, do not trust it too much.

I have an earlier version of this book and I also had Dr. Proakis as a professor in graduate school. Dr. Proakis has a nack of making something simple into something complex. His teaching style is similar. I always thought com theory was very difficult until I read books by others that discussed the same topics in a more cogent and lucid style.

I both love and hate this book at the same time! Its great for its wide coverage and unified presentation of material but the organization is rather poor indeed. There is little motivation and the material is presented as a matter of fact collection of topics. The presentation also seems to be out of sequence at times.

Other gripes: The errata is quite large and still not complete. Upon closer inspection, I found that several of the problems require assumptions that are not clearly described in the problem statement. This leads to a frustrating experience at times.

I can think of ways in which the material could be re-organized to make the "flow" much better. Doing this and adding little motivation and connecting material would make this a truly great book. I don't think that much additional effort is required for the re-organization - too bad this hasn't happened in the first 4 editions.

A major plus of the book remains - nothing else comes close in coverage. So it is still a pretty decent book if you can get the intuition and organization from elsewhere.

I agree with the previous reviewer. Without the maths, digital signal analysis is simply impossible. One of the reviewers complained about the proof of the Nyquist theorem, based on the fourier transform. I looked it up; the proof is explicit and clear. Of course, readers who are unfamiliar with fourier transforms may find it difficult to follow the proof - they should first learn some elementary maths before reading this book. I'm glad that Proakis does not treat all the kindergarten stuff in full detail. After reading what this book has to say about equalization, error control, modulation, etc., the road to the professional literature is open.

I took a graduate class at the University of Cincinnati and they used this book (it was round 2 for me as I had Communication Systems with his book that was co-authored by Salehi). Two classes + two Proakis books = brain hemorrhage. I will say that if you are mathematically strong doing well in these courses won't pose a problem. However, the amount of rereading needed to hammer in the concepts is overwhelming. It wouldn't have to be this way if he put in a lot of examples. Also, most of the problems are proofs or very complex and for that reason I noticed that neither of the professors that I had used problems from the book on exams. One exam was forced to become a take home as the results were so terrible and the other was a gram schmidt problem coupled with a small proof on bandpass signals. Some advice is in order. Follow these steps if you want to learn the material:

1.) Take and ace a Random Processes class. This will get you the background you need to blow through chapters 2 and 4 of Digital Communications by Proakis.

2.) Over break or during free time pick up the Communication Systems Engineering book and read through chapter 7 to hammer home chapter 5 of Digital Communications.

3.) Invest in a solution manual if you can find it. This way one can try ALL of the problems in the text and really learn the material.

As a final note, if you can master this course then there is absolutely no graduate class that will stand in your way. Any wireless or DSP class will seem like calc I after it.

My teacher at Georgia Tech said.."this is like the bible of digital communication"..and certainly it is like the bible, confusing and too complex. Theory wise, OK it serves well, but as a reference for practice..get Lee and Messerchmitt!!

This book is an excellent reference a gives a deep idea of many aspects of digital communication. At the beginning (chapter 1) the statistical theory is discussed, then the digitally modulated signals and their power and spectral characteristics are analyzed. Then it goes deeper in advanced matters. I found this book excellent, even if it can be tough to read for people without a good math background or previous knowledge in the field. Anyway if you are interest in looking for particular reference it is really excellent and gives also a brief about where to go for further info at the end of each chapter. Definitely not for undergraduate students looking for a basic idea, but very good for a deeper mathematical approach. It think everyone who works in this field should have a copy.