Computer Science Books

Explosion of the Internet pushes the demand of bandwidth to a critical point. Various technologies on opical fibre system enhance the transmission bandwidth in the worldwide communication system. But, a high transmission bandwidth network reveals its bottleneck on the switching bandwidth - how to switch singal and data in tight time bound in various fashions.

Existing router and switch vendors use network flow probability models to simulate the performance and blocking probability of a switching matrix design. This trial and error approach has no guarantee on the performance and implementation complexity.

Prof. Li breaks a new ground by using algebra calculation in the switching fibric representation. By his 20+ years research, his algebra algorithm has an unique ability in handling non-blocking switching, Time-Space-Time 3-phase switching and concentrator/sorter implementation. And he suggests some interesting implementation oriented design in reducing implementation complexity in a manhatten VLSI layout.

This book covers Prof. Li's lifetime work and his research in switching, probability models and information theory. It worths any information engineers for a deep look, and it's a fanastic reference book for any switching researchers...

Explosion of the Internet pushes the demand of bandwidth to a critical point. Various technologies on opical fibre system enhance the transmission bandwidth in the worldwide communication system. But, a high transmission bandwidth network reveals its bottleneck on the switching bandwidth - how to switch singal and data in tight time bound in various fashions.

Existing router and switch vendors use network flow probability models to simulate the performance and blocking probability of a switching matrix design. This trial and error approach has no guarantee on the performance and implementation complexity.

Prof. Li breaks a new ground by using algebra calculation in the switching fibric representation. By his 20+ years research, his algebra algorithm has an unique ability in handling non-blocking switching, Time-Space-Time 3-phase switching and concentrator/sorter implementation. And he suggests some interesting implementation oriented design in reducing implementation complexity in a manhatten VLSI layout.

This book covers Prof. Li's lifetime work and his research in switching, probability models and information theory. It worths any information engineers for a deep look, and it's a fanastic reference book for any switching researchers.

John Ho :-p

This book is a valuable reference -- a real work of mathematical scholarship concerning problems from elementary number theory, such as primality testing, square roots mod p, quadratic residues, polynomial factoring, and generation of random primes -- algorithms for which efficient solutions are known.

Related algorithms such as the lattice reduction algorithm of Lenstra, Lenstra, and Lovasz, and elliptic curve point counting over finite fields are not covered.

Three outstanding features of this book are:

1) The extensive chapter end notes that provide a comprehensive review of the history and state of the art for each topic addressed in the book. These notes are so detailed that they are like having a mini book within a book. Anyone doing research in the field would do well to own this book for this reason alone.

2) Exhaustive bibliography, all together there are over 1750 bibliographic entries.

3) Applications of the RH and ERH(Riemann Hypothesis and Extended Riemann Hypothesis). I know of no other single reference that covers the consequences of these conjectures being true in terms of primality testing, quadratic non-residue testing, primitive root finding and so on.

The algorithms are presented in pseudo code and practical implementation remarks are reserved for the notes section of each chapter.

Recommended for upper level undergraduates and all the way on up to faculty.

As a bonus the book is a real pleasure to view due to the excellent job done in the layout and typesetting.

I look forward to volume two which will focus on algorithms for intractable problems for which efficient (polynomial time) algorithms are NOT known such as factoring and the discrete log problem.

Bach and Shallit have done a wonderful job of preparing a survey of Number Theoretic Algorithms. After covering the basic mathematical material and complexity theory background, the book plunges in to discuss computation in (Z/(n)) and various algorithms in Finite Fields.

The part of the book that I like best are the last two chapters which deal with prime numbers and algorithms for primality testing. The authors have done an exhaustive survey of this area. Proofs of the correctness of the algorithms are wonderfully concise and lucid. The second volume [not published yet] will discuss problems for which efficient algorithms are currently unknown for example factoring, discrete log etc. The authors also promise coverage of the Adleman, Huang proof that Primes \in ZPP.

Exercises have been chosen carefully, and most of the solutions are available as an appendix (for the others references are given). Finally the bibliography is *huge* with close to 2000 citations. Overall an excellent book for reference and for a one stop introduction to the wonderful area of Algorithmic Number Theory.

This is one of the best IT management books in recent times. The author gets right down to the point with his vivid, direct approach in addressing the problems with the IT co-ordination with the Business objectives of any company, relying on key IT solutions. The book proposes a number of frameworks and solutions in identifying and eliminating the key IT-Business issues that plague many companies. A great read for anyone, even outside the IT field, as a number techniques from this book can be applied to any sector.

How many organizations waste resources focussing on projects that are not critical for the success of an orgnization while pushing back projects that could generate new revenue or streamline existing processes substantially ? The author proposes a scientific and formal way for an organization to model it's organizations needs and to create a landscape of all IT requirements within a company so that the entire organization can be looked at wholistically.This process enables high level management to get a complete vision of an organization's needs instead of relying on lower management to filter information.The process will not guarantee the right decision will be made but it will offer a more complete picture for decision makers.

This was a very informative book and gave lots of insight into the hugh effort that must have gone into this age-old quest. The focus at the personal level gives a nice picture of what the individuals must have gone through. Makes you think "Hey, I could have been there..." Very enjoyable.

Morton Grosser came to Shafter and to our homes while we were trying to build the first man-powered plane. He wrote a book that centered on the people that were working, thinking and arguing towards the goal. If you want a feel of what it was like to do something for the first time ever, read this book.

I used this book for a recent class in Analog Filter Design. This book is an excellent reference and learning tool for filter design. Overall This is the best textbook I have used.

Working with numerous analog circuits in the day to day workplace, I found that this book was very helpful. A nicely written and understandable reference.

It is assumed that the reader of this book has had a first course in circuit analysis and is familiar with Laplace transforms. Because the book is intended for undergraduates, algebra is used to explain the mathematical workings of the various filters rather than calculus if at all possible. The book concentrates on inductorless filters. Such filters are required for voice and data communications systems where the size and weight of inductors make their use not very feasible. An advantage of this approach is that the usefulness of the op amp can be stressed. Chapter 2 introduces the op amp in analog operations: addition, subtraction, multiplication, and integration. Chapters 3 and 4 constitute a review of sinusoidal steady-state topics recast to provide an introduction to first-order filters. Chapter five features the universal biquad, and through it the standard forms of filter response: lowpass, bandpass, highpass, bandstop, and allpass.

In the chapters that follow, the functions of a filter are studied in combination with the frequency-response approximation used by filter designers: Butterworth, Chebyshev, Bessel-Thompson, inverse Chebyshev, and elliptic. Chapter 14 treats the synthesis of doubly terminated passive ladders. This, together with Chapter 11, constitute an introduction to passive filter design. An important conclusion is reached in Chapter 14 in regards to sensitivity - the passive ladder circuit has low sensitivity. The study of simulated ladder circuits is begun in chapters 15 through 17. Simulation of the passive ladder is accomplished in three ways: by introducing new elements which make it possible to exclude the inductor, through the simulation of the block diagram representation of the ladder, and through the simulation of resistors using switched capacitors. All of these methods are discussed.

In spite of its academic origins this really is a very helpful design book. It has the circuits side by side with the math, and it also has design rules in numbered steps with very good illustrations and examples throughout. I highly recommend it for the library of the practicing engineer as well as the student.

(Reviewed by a practicing electrical engineer) This book is an excellent introductory text about analog filter design. It was written for use in teaching a junior-senior level undergraduate class at the University level. The book begins with a review of basic op amp circuits and then progresses to simple one-pole circuit designs. From there, the author builds a step-by-step theoretical background for active filter design, starting with the most fundamental two-pole building block (the biquad circuit), and then showing how to use these to design Chebyshev, Butterworth, and Bessel filter characteristics of any number of poles. Cauer, Inverse Chebyshev, and switched capacitor designs are also covered. Other highly useful chapters cover delay equalization, sensitivity to component variations, as well as a general design approach for choosing appropriate filter characteristics for a design requirement. The book's main focus is on active filter design using op amp's, although there is also one chapter devoted to passive filter design. Design is stressed in the examples and problems from the very beginning pages, as opposed to mere mathematical analysis. The mathematics are fairly straightforward, mostly algebra, and the text is fairly easy to read and follow. Examples illustrate each chapter segment, and each chapter has questions at the end. I would recommend this book for anyone seeking a basic theoretical understanding and design capability for active filters. (Probably other introductory texts would be better for passive filters.) This is not a "handbook" or "filter cookbook"; it actually lays the theoretical and mathematical groundwork for the filters it teaches you to design.

Great book. Must for any one working on ADCs and DACs.

One of the few books in the data converter field. I used this book when I took an advance course on analog circuit design and found this book to be very usefool. This book contain a good amount of information on data converter and cover almost every important areas - i.e specification, testing, High-speed/high accuracy architecture, sample and hold circuit, compartors, noise shaping coding, voltage and current reference sources and also contain a chapter on sigma-delta converters that can lead to more advanced study for IEEE books on oversampled data converters. If you are an analog circuit designer or want to be an analog circuit designer this book will be a good reference for data converter circuit.

Excellent work by excellent authors. Simply the best you can find in this field!

A graduate level book. It talks about the detail model and behavior of induction machine, synchronous machine and etc. The book gave me quite an idea of machinery. It was well written and explained in a detail and clear way. Since it is the first textbook I bought at USA, after reading, I cannot help wonder: that's the reason why American has advance technology, because they have good books. In my coutry, China, the authors likes to write hard, tough things in a small thin book and we can hardly understand them.

With a book as technical as this, it's hard to know who my potential review readers will be. Nevertheless, I will give a very CONCRETE and ACCESSIBLE review. Also you should bear in mind this: the book does have some concrete and accessible aspects before getting into complex issues; moreover -- and maybe this is the most important of all: the simpler, concrete and accessible aspects of this book DO HAVE UTILITY. I'll give a personal example of the utility of the book's simpler material at the end of my review. But now let's proceed in a nice logical order...

The book deals with "combinatorial optimization" problems. These are problems where there are (1) a gigantic number of discrete configurations that are possible, (2) a way of scoring how desirous a configuration is, and (3) ways to change the configuration from the present one. Examples include the scheduling problem of how to assign 20 workers to one job apiece for 20 jobs (with different worker/job pairings having different costs); and, of course, the famous traveling salesman problem -- requiring precisely one visit to each of N cities and a return to the first.

The most easily understood algorithm to solve combinatorial optimization is BLIND RANDOM SEARCH (BRS): generate a random configuration, score it, repeat (always keeping the best score yet encountered and its corresponding configuration saved in memory). You can have stop criteria as you wish -- including an OR'd pair (which I find to be itself a great improvement) -- such as UNTIL (a) score is X good or better OR (b) you've generated N random configurations.

BRS performs relatively poorly. A HUGE improvement is an algorithm called "Iterative Improvement" (II). This algorithm is covered on pages 6 thru 8 of the book. The idea is to take a BRS configuration then do some modest moves around that configuration -- scoring and repeating until you had k failures to improve. The best obtained is that one BRS "point". Generate a new BRS point and compare to the old as usual, but now the II loop probably substantially improved that old BRS score to which you are comparing.

Both BRS and II involve only "downhill" moves. Only a lower score and its companion configuration are kept and a new configuration never becomes the current one if its score is worse. The danger is "getting stuck" in a local minimum as opposed to the global minimum (truly best score). To avoid this danger there is the "probabilistic hill-climbing" algorithm of Metropolis. An improved configuration (one with a better score) still becomes the next current configuration, but you have some probability of taking the next current configuration as being the current contendor even if this contendor configuration has a worse score. The probability is related to the score and a parameter that might be thought of as temperature.

From the probabilistic hill-climbing algorithm of Metropolis, all you need to get to an annealing algorithm is a schedule for appropriately reducing the "temperature" parameter (which controls up-hill acceptance probability) in successive steps. The analogy is freezing a liquid to get its perfectly crystalline line-up of atoms, free of defects. Go too fast and you may get a glass rather than a crystal.

The book's chief aim is how to recommend IN GENERAL, without recourse to your specific problem, a schedule for the "temperature" changes. If this be your aim in considering the book, well it goes without saying you need not consider further: here's your book.

But what about the less technical reader? First of all, the book does gently introduce you to combinatorial optimization, blind random search, Metropolis and annealing. Second, the few pages on Iterative Improvement are EMINENTLY USEFUL in a PRACTICAL sense -- and are a good simple alternative to annealing (my example will be at the end). Third is that the book includes several ancillary extras.

The ancillary extras:

· tutorial on all of matrix mathematics
· tutorial on Markov Chains
· material on probability and conditional probability
· tutorial on Statistics -- esp. w.r.t. the Normal distribution and Central Limit Thm

I'm not saying that the ancillary extras are the best there is for a novice level reader, but most folks would not know of the existence of this material in a book called "The Annealing Algorithm".

The final bit of ancillary material is Pascal computer code for all the algorithms in the book and a complete program for doing the whole annealing bit on the electronic chip placement combinatorial problem.

MY EXAMPLE OF UTILITY OF ITERATIVE IMPROVEMENT ALGORITHM:

My problem is not combinatorial optimization, but can still use the ideas of iterative improvement since I am solving a deterministic problem (one without any random element) using Monte Carlo methods (using random numbers). My problem: I have the coordinates of the midpoint of a line segment; the line segment's length is also known and is roughly one-fourth the diameter of a circle; the line segment lies the annular area between this circle and a circle with a radius half-a-line-segment bigger radius than that of the original circle; lastly, given the rotation angle of my line segment, I ask this: what are the coordinates (x,y) of the intersection of the line segment and the original circle? (I took steps to check that YES, there was an intersection.) Solving the problem analytically didn't work. (Or at least, I couldn't do it.) I had used a BRS Monte Carlo approach. Then, re-reading this book, it occurred to me to use the book's algorithm (Iterative Improvement) on pages 6-8 (Pascal code page 8). I got a big improvement in lowering the error. Obviously, I had to delete details in this review (like how I even know error in my problem, and if I do know it, why can't I fix it exactly -- hint: circle is the locus of all points equidistant from a given point), but the POINT FOR YOU is that I attained a great improvement in my problem just by using the book's explicit algorithm (Pascal code) for Iterative Improvement.

Finally, the book is nice to read -- both very easy-on-the-eyes typography (unusual for a "math" book) and a good flow to the authors' writing.

The Annealing Algorithm is part of The Kluwer International Series in Engineering and Computer Science, concentrating on VLSI, Computer Architecture And Digital Signal Processing. The Consulting Editor, Jonathan Allen, has put together a concise, academic publication that introduces the conceptually simple procedures of simulated mathematical annealing and mathematical optimization. Research for the book was carried out at the Thomas J Watson Research Center of the IBM Corporation in Yorktown Heights, NY. Chapters include A Preview of The Annealing Algorithm, Preliminaries from Matrix Theory, (Markov) Chains, Chain Statistics, Annealing Chains, Samples from Normal Distributions, Score Densities, The Control Parameter, Finite-Time Behavior of the Annealing Algorithm, The Structure of the State Space, and (Pascal) Implementation Aspects. References and an Index are included. Mathematicians and computer science professionals seeking to learn about the conception of annealing as a combinatorial optimization tool, will find this publication interesting and will appreciate its use as a basis for further research.

This excellent 200 page book describes itself as an "introduction to APL", and it does contain an efficient and orderly introduction to APL functions, operators, syntax, and arrays. By page 30 or so, however, it begins to use APL to solve some serious math problems--integration, Netwon's Method for solving roots of equations, the Power Method for estimating eigenvectors and eigenvalues, as well as common statistical methods. Instruction in APL language is integrated throughout with the exposition of these math problems.

Reiter broaches what is often viewed a multi-faceted task in a manner so as to make functioning in APL (A Programming Language) straightforward, practical, and useful. Basics are taught first, both Monadic and Dyadic use of the elementary operators, later introduction to Del, programming and the advantages of the branch. Mathematical concepts, including the basic linear, and the less familiar higher-order regression techniques are presented along with the APL equivalents of the algebraic technique for simultaneously solving equations.

I appreciated the updated reference for some of the more antiquated terms which may appear in other volumes of earlier vintage. No longer is fuzz tolerance, stack clear, clear stack, or such antiquated terminology present to confuse the reader. No more thorns, here. A simple )RESET sets everything clear.

Those familiar with the Interactive Approach would appreciate that A Rose is a Rose, and the standard for much of what we do, but for the newly initiated into the Magic of APL, Reiter makes all of the initial steps very )CLEAR.

The thema thoughout the excellent text/reference/adjunct to mathematical course is that set theory, Boolian Algebraic processes and APL work hand-in-hand.

After demonstrating basic skills in monadic and dyadic operators, the reader can go to DEL and really begin to gain the experience of just what APL is all about.

Again, )SAVE APL ! with all of its factorials and permutations. Reiter has certainly helped to do so by providing a work which will hopefully introduce a whole new generation to APL.

An excellent work by an erudite writer.

Jerrell L. Driver, Ph.D.
Licensed Psychologist
Health Services Provider Certified
Missouri PY00099
APL