Statisical and Adaptive Signal Processi

The book is well written, and it is a great reference book for anyone interested in Statistical signal processing and MIMO systems.

This book gives a brief overview of fundamentals of Digital Signal Processing and Stochastic methods, before graduating to the core topics, namely Signal Modeling and Parameter EstimationNon-parametric estimation, Optimal filter design and structurs, RLS, LMS and Adaptive Filters.Though high on content, the topical organization of the book leaves a lot of room for improvement. A logical sequence of topics to be studied by an advanced level DSP student is recommended as follows - 5. Linear Signal Models, 9. Signal Modeling and Parametric Spectral Estimation, 6. Optimum Linear Filters, 7. Algorithms and Structures for Optimum Linear Filters, 10. Adaptive Filters, 8. Least Squares Filtering and Prediction, 11. Array Processing. You may have to keep skipping advanced topics towards the end of a chapter, only to come back later after having gone through related background material in other chapters. In this respect, this volume is indeed inconvenient.However, the authors have more than made up for all its faults with the depth of content, and also the breadth. Assuming that this book is meant for an advanced reader, it is very much self contained, from the ground up, barring a few minor low-level details, which the authors have assumed prior knowledge of.Chapters 11 and 12 essentially deal with very specialized applications for Radar Engineers and people dealing with esoteric math involving Signal Processing techniques - the case in point are the topics on Blind Deconvolution and Unsupervised Adaptive Filtering.The authors have also provided some rudimentary background information on Holor algebra (matrix and vector algebra esp.)I would recommend the reader to keep a more basic text on Mathematical methods for Signal Processing as a cross reference while using this book. A case in point is Mathematical Methods and Algorithms for Signal Processing, by Todd K. Moon and Wynn C. Sterling.

I had the chance to read most of the text, and selectively attack some of the problems. Here are my opinions:Pros 1. Some of the chapters, including the first few introduction ones, are very well written, and enjoyable to read; 2. Cover a great deal of technical contents; 3. surely echos the title "statistical and adaptive".Cons 1. some chapters are a little messy, including the most important ones such as chapter 6 and 10. Much better approaches could be used; 2. excersizes are not so well-designed. Not many really innovative/inspiring problems are given. Need more diverse Difficulty levels. Generally a bit easy. 3. references are not very supportive. If you find one section or discussion interesting, you would like the book to thread a few summit journal papers in this issue to refer. It has not done an outstanding work in this sense.Over all, an above average text/reference book, good if taught by a good teacher. However, need to bring into a lot insights and better examples/excersizes to make the book a true classic. However, the cons mentioned above plus the text itself gives people impression that the authors are either not yet true masters of the material, or they hurried up finishing the book. But they do have a lot knowledge.

I believe this book is destined to become the "classic" graduate text used to teach statistical and adaptive digital signal processing.If you enjoyed the introductory text "Digital Signal Processing" by Proakis and Manolakis (Prentice-Hall 1996), I think you will enjoy this book by Manolakis, Ingle, and Kogon. It is written in a similar style, with an introduction to each chapter previewing the material to be covered, a logical development of the material including examples, and a conclusion summarizing the high points of the material covered.At chapter's end, there is a set of well thought out exercises ranging from easy to difficult. There are no answers to the problems in the back of the book, but there are enough examples in each chapter that one should be able to tackle most of the exercises. Some of the exercises require MATLAB. The authors have written some custom MATLAB functions which are available from the publisher as an e-Mail attachment.I would say this book is written at the graduate level and requires knowledge of several disciplines: 1) DSP- At the level of Proakis + Manolakis intro text (cited above). 2) Linear Algebra- Cramer's rule, LDU factorization, eigenvalues, eigenvectors, Hermitian and Unitary matrices, etc. 3) Statistics- Random variables, averages, variances, estimators, sampling distributions, auto- and cross- correlations. I had no previous knowledge of stochastic processes, and was able to pick up enough from Chapter 3 to get through the rest of the book.This book is, above all, a mathematical text written for engineers. It describes the theory and equations underlying statistical filters.There is a lot of meat in each section. I typically had to read each section an average of 3 times for it to sink in.With help from Figure 1.2.8 of the book, it covers the following material:Chapter 1- Introduction to applications of spectral estimation, signal modeling, adaptive filtering, and array processing.Chapter 2- Review of discrete-time signal processing.Chapter 3- Review of random vectors and signals: properties, linear transformations, and estimation.Chapter 4- Random signal models with rational system functions (AR, MA, ARMA, ARIMA).Chapter 5- Nonparametric spectral estimation.Chapter 6- Optimum filters and predictors -- matched filters (including Wiener) and eigenfilters.Chapter 7- Algorithms and structures for optimum filtering (including algorithms of Levinson, Levinson-Durbin, Schur, Kalman, ...)Chapter 8- Least-squares filtering and prediction (normal equations, orthogonalization, SVD).Chapter 9- Signal modeling and parametric spectrum estimation.Chapter 10- Adaptive filters: Design, performance, implementation, and applications (includes steepest descent, LMS, NLMS, CRLS, QR-RLS, fast RLS, fast Kalman, RLS lattice-ladder, ...).Chapter 11- Array processing: theory, algorithms, and applications.Chapter 12- Higher order statistics, blind deconvolution and equalization, frac