Customer Reviews

Communication Systems

5 star:		(8)
4 star:		(4)
3 star:		(1)
2 star:		(5)
1 star:		(2)

 

 

To the best of my knowledge, this is the first undergraduate book on communication systems to include introductory treatments of time-frequency analysis and cryptography. Three years ago when I took my first communications course, I have used this book as a reference. "Modern digital and analog communication systems", second edition by P. B. Lathi, was the adapted textbook by my professor. I had four books to study the subject from; Lathi, Haykin, Zimer and Tranter, and Leon W. Couch II. Because Haykin was the most popular, I though that it was the one that I should concentrate on. Upon completion of reading each chapter, I had more questions than answers, so I had each time to turn to Lathi and Zimer, which I found much more fluid and accessible. Couch's "Digital and analog communication systems" is simply a piece of garbage. After reading Lathi's "Modern digital and analog communication systems", for analog communications course, B. Sklar for digital communications course, I gained an adequate maturity on the subject such that enabled me to read Proakis's "Digital communications", third edition, which I found to be the most serious and involved book on the subject. I still think that Haykin's "Communication systems", is very much similar to Kant's "Critique of pure reason", since you have to read the same thing over and over again until you reach the point where you start wondering "have I understood it, or is it the repetition that made me memorize the subject?!". Imagine Bertrand Russell or Kurt Godel trying to explain multiplication and division to the first grade students, the situation is analogues, Haykin, world leading communication researcher trying to explain introductory analog communication principles and digital pass band transmission to senior students, he must admit that it isn't easy to imitate R. Hamming's insightful writing style. Chapter 2 starts very well, until section 5, where the author is pointlessly trying to mention uncertainty principle, and yet, he does not. Maybe in next edition, he will include a section on Wigner distributions and somehow manage to mention tensor product. In 2.11 concept of pre-envelope is discussed in very dry manner, maybe in the next edition it will be mentioned that the real and imaginary parts of an analytic function are harmonics and in the footnote complete theory of harmonic analysis will be summarized in 5 lines. Chapter 4, Random Processes doesn't even mention optimal filtering, the subject that arises naturally in the context of signal recovery. In next edition it is expected that chaotic dynamics of the sea clutter will be compressed in a single section. Don't waste your time on chapter 5, read Lathi's (1998) chapters 12 and 13 on the same subject. Section 6.13 unfortunately doesn't include filter banks, pyramid algorithms, quadrature mirror filters and thorough five-line discussion of multi-resolution analysis. Section 7.9 does not include generalization of LMS algorithm to error back propagation, and blind deconvolution is not even mentioned. Section 8.4 on geometric interpretation of signals says nothing about Banach spaces, nor Frames. It's true that section 8.6 includes very nice discussion of coherent detection of signals in noise, but for partially coherent, deferentially coherent and double deferentially coherent communications with waveforms the reader is advised to refer to "Digital communication techniques, signal design and detection", by M. K. Simon and S. M. Hinedi, 1995. Chapter 9 on spread spectrum modulation says no more than 16 lines about synchronization, the subject that is given four pages in B. Sklar (1988). Delay locked loops does not even ring a bell. Don't you agree that synchronization is one of the main issues in the spread spectrum systems? Add to that, that examples of pseudo-random codes given in this book do not exceed periods of 256 bit length, that is less than one microsecond of randomness, while practically attainable periods are of one century duration. Anyway, for a better one chapter treatment of spread spectrum systems refer to Proakis. Section 10.13 "Compression of information" has very nice comprehensive 9 lines on vector quantization. Maybe in next edition "Information-theoretic models" and "Radial basis function networks", will be included as a section in this chapter. Don't waste your time on chapter11 "Error control codes", read B. Sklar's (1988) chapters 5 and 6. Chapter 12, what it is about is left for you to figure out. Let me conclude by saying that "Communication systems" is an undergraduate text written with graduate language. I found Haykin's "Neural networks, a comprehensive foundation", in order of magnitude superior to any other book on the subject. Meanwhile I am using his "Adaptive filter theory" to gain enough knowledge about the subject for my current project on system identification with LMS implemented with FPGA. If you are a beginner in communication systems subject, use your time efficiently by reading Lathi and Zimer on introductory subjects. If you like to go on further, than get B. Sklar, and then Proakis. Don't get deceived by brain damaged readers who claim that it is the best book as an introduction, maybe they say so, because they have failed in the course over and over again till they decided to memorize all formulas and definitions which they found organized in Haykin's. Also note that tons of universities started to adapt Lathi's instead of Haykin's.

First, you really need to have prior exposure to the subject to thoroughly understand some of the topics involved. Some equations are introduced 'ad hoc' and are not explained, even intuitively. If you want to know the 'why' and 'how', look to other books such as Lathi's and Proakis's, which are much better. I found their books to be more systematic, more organized, and simply more logical. Our professors at my university do not seem to like this book, and the students now know why. We switched to Communication Systems Engineering by Proakis, which is more mathematically sound and much clearer. However, you can gain some useful insights from Haykin's book, so I give it two stars instead of one because everything in Haykin's book cannot necessarily be found elsewhere, so it is of use at times.

In the last lines of 'Preface', Haykin says: "It is expected that the reader has a knowledge of electronics, circuit theory, and probability theory". In fact he forgot to mention signals & systems, and analog & digital communications, since the book addresses professors, definitely not students. The prose comprises sophisticated ambiguous phrases, as if they were set that way on purpose, although his other books are written in clear, interesting, and intuitive style, like 'Neural Networks', 'Adaptive Filter Theory'. Probably more famous for bibliography than for the material itself, so, one can find there titles such as: 'The Analytical Theory of Heat', J.B.J. Fourier (1878).'Theory of Motion of the Heavenly Bodies', C.F. Gauss (1809). VA Kotelnikov's Theory of Optimum Noise Immunity (1947). 'Adaptive Filter Theory', Haykin (1991). 'Digital Signal Processing', Oppenheim & Schafer, etc , as if the undergraduate student, who barely distinguishes between coherent & noncoherent detection, is expected to go through such subjects. A lot of unrelated topics have been added just to make the book look thicker, like Wavelet Transform, Short Time Fourier Transform & FFT and others, each of which one wouldn't understand from it, unless he/she previously had been exposed to. The bottom line here is that I find this book completely unreadable as a text, but not bad at all as a reference for the topics that one might have had forgotten.

At the start, a lot of what is covered gets a fairly rigorous mathematical treatment, but as you get on through the book, it gets very hand wavy and though there's still lots of maths to do, it's a bit hard to relate it to what Haykin is talking about. Having read the answer book for a lot of the problems, you would think comms engineering is easy, where in fact it's bloody difficult.

This book is excellent for a beginner in this subject. The author gives rigorous mathematical treatment as well as useful insights on difficult concepts. In my opinion, one of the best Communication Systems books ever written ! A must-have for a DSP or a Communications Engineer.

Dr. Haykin's book on Communication Systems is one of the standard textbook in the world. It is the first book from which many EE students start to learn the theory of communications. I have used several books written by Dr. Proakis and Dr. Haykin in my class for many years. Haykin's book features complete (and concise) coverage of fundamental topics in communications, and is mathmatically clear.
However, the high standard required for an excellent writer is not completely achieved in this book.
First, in Example 1.3 (middle of Page 39), the derivation process of the conditional expectation is misleading: the condition $t_d<T-|t_k-t_i|$ is meaningless; in fact, it is the CYCLO-stationary property that leads to Eq. (1.18). This cyclostationary property due to period T has been clearly stated, for example, in the books of (Stark and Woods), and (Gardner). Since Haykin's books were usually error-free (conceptually), it took me a while to check the derivation and figure out there is indeed an error in missing the basic concept of cyclostationarity!
Secondly, in Eq. (1.98) (Page 63), the dummy variable t inside an integral appears in w'(t), which is obviously wrong.
Thirdly, in Eq. (3.20) (Page 191), the absolute value symbol is missing in two terms related to the magnitude |H(f)|. Also, in Figure 3.6 (b) (Page 190), the phase arg[H(f)] is not illustrated correctly.
There are other errors throughout this book (even in the 4th Edition!).
Also, many problems in Haykin's book are quite lengthy and not well-suited for midterm and final exams from an instructor's perspective.
I wonder where to find an error-free comprehensive textbook in the field of telecommunications like Knuth's (hard to read, though) in computer science.
Compared with other textbooks in communications, this book still ranks among the best in terms of its clarity, illustratoins, and comprehensive coverage - if and only if the errors aforementioned can be corrected.

I wish I could give this book 0 stars!
I am using this textbook for an undergraduate communications theory course. The textbook is poorly organized and the writing is difficult to follow! It is difficult to even approach the problems out of the book even after reading the chapters. The text tries to cover a wide range of material while offering little or no insight to the underlying communications principles. Equations after equations with poor explanation is certainly no way to learn communications theory. I deeply regret choosing this textbook for my course. Definitely look to other books if you want to learn some communications theory!!!

This book was my introduction to analog comm. systems. I found it to be clear and expository. I am surprised by the other readers' mediocre reviews and petty diatribes. Admittedly, the book does contain several extraneous topics, but many are listed at the end for those who are interested. I think some derivations are left incomplete intentionally because the reader stands to learn more by verifying the results. I disagree that the book is written poorly. If one pages through a more advanced text, I think it would become quite clear that the level of 'Communication Systems' is introductory and NOT intended for professors and experts.

Concessions: 1) I have enjoyed nearly every book on this subject. 2) My math background is much broader than the average engineering undergraduate.

Although this book did not tell us to learn the Fourier Transform first, the concept introduced in the second chapter is enough. The topics was introduced in a way that the student feels comfortable. However some equations weren't thoroughly developed. Generally, this book is good!

I read this book in my undergraduate course , and it is this book what makes me distinguished from normal students who still try to understand many things in communications theory. I suggest them to study signals & systems , probability and advanced maths first , then they may try to go deep in communications. Communications theory is not a childs game , so you have to be on the level to understand what Haykin says in that book !!