Customer Reviews

Detection, Estimation, and Modulation Theory. Part I: Detection, Estimation, and Linear Modulation Theory (Part 1)

5 star:		(3)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

I have just taken this course from Dr. Van Trees at GMU. Est&Det organized many concepts taught in other graduate engineering courses into a coherent philosophy. The result is not only a rich understanding of estimation and detection, but also random processes, Wiener filtering, Kalman filtering, radar and communications theory etc.

The course was taught directly from the text with little outside material. Very little has become obsolete in the 30+ years since it was written.

The strong positives of this book are the philosophical organization, clear concise writing, and incredibly well conceived homework problems.

The only negative of the book is that there are many proofs done in great detail. This provides the necessary foundation for the material, but also makes it easy for the student to lose track of the bigger picture.

Dr. Van Trees tends to try to drive home the higher level concepts while glossing over many of the details when he is lecturing. The exercise problems then force the student to give the necessary attention to pertinent details. In my opinion, this is an excellent approach to teaching the material.

Overall, this course was as good as any I've every taken. The text is as important and useful as any other I have.

Van Trees, Part I (together with Wozencraft/Jacobs' Principles of Communication Engineering, and Gallager's Information Theory) is a must read to establish a solid background in detection/estimation theory and form connections to applications such as communications engineering and information theory.

Although most recent graduate education uses Kay's book (which is also a remarkable book), there are still a lot of details in which Van Trees, Part I excels. Especially, the exercise problems are actually lectures by themselves, and first time reader is encouraged at least to look at selected problems listed at the end of the book. Note that there is a solution manual floating around for these selected problems.

A good comparison between Kay and Van Trees, and their complementary nature, can be established how they treat the description of the Cramer-Rao bound, Kay emphasizes the recent developments and derivations (mostly of arithmetic and bookkeeping nature, results from post 1968 papers), whereas Van Trees goes leaps and bounds and discusses other bounds which apply when Cramer-Rao does not. I appreciate having both books as a result.

It is interesting to note that after almost 20 years using Van Trees in a couple of courses, I can still navigate my way through the book with ease since it well organized and methodical.

RECOMMENDATION: BEST BUY.

But dont stop here, and buy Wozencraft/Jacobs and Gallager as well.

The title of Van Trees' "Detection, Estimation, and Modulation" theory essentially covers the topics of Volumes I & II of the series, with Volume II covering Modulation--specifically analog modulation which has been overtaken by digital techniques.

Volume I--the text being reviewed--however, is necessary background for anyone wishing to understand communication receivers, or radar or sonar, or the related necessary signal processing topics. Yes, there are other books on these topics such as those by Stephen Kay. However, having taken a one day course from Dr. Kay himself, I can say that nothing beats the content of Van Trees' classic work.

First take a course in Detection & Estimation using Volume I of Van Trees. Then proceed to more advanced and specialized courses in related areas.

Chapters 4 & 5, covering nearly 230 pages, are the heart and soul of Volume I. However, I even found Chapter 3 on Random Processes useful, including a description of Gaussian Processes (including jointly Gaussian processes) that I have not seen elsewhere. If the reader or student masters Chapter 4 on Detection of Signals (in noise or with unwanted parameters), and Chapter 5 on Estimation of Continuous Waveforms, that individual will be well-positioned to do well in follow on courses in Digital Communications, Radar & Sonar, or applications of Adaptive Signal Processing.