The advantages of digital signal processing (DSP) - sustainable accuracy, flexibility and versatility, noise immunity, falling costs, and the possibility of implementing a wide range of complex and intricate functions not realisable with analogue processing - have led to a phenomenal increase in the applications of DSP in the past three decades. DSP is now used in every conceivable field that involves the transmission and storage of information, the acquisition and interpretation of signals, and the control of systems - from household gadgets to nuclear missiles.
As a consequence, good books on DSP techniques are in demand to support a wide range of courses designed for the following groups: practising engineers and technologists who need to acquire new skills for survival in a changing workplace; undergraduate students in electrical and computer engineering departments and other related disciplines; postgraduate students and researchers requiring in-depth DSP knowledge; and DSP-skilled practitioners wishing to update their knowledge.
The core material of a DSP course typically includes representation and analysis of signals in time and transform-domains, linear time-invariant systems, structure and design of digital filters, analogue/digital signal conversions and multi-rate systems, DSP implementation issues and applications.
Sanjit Mitra's book gives a highly readable in-depth treatment of the core topics and would amply meet the needs of the first three groups.
The first chapter successfully sells the DSP concept with an overview of typical DSP operations, applications and advantages. This should provide DSP novices with much-needed motivation to delve into the remaining chapters of the book, where the going can get tough as various key topics are explored skilfully and at length.
An important feature of the book is the large number of clear worked examples in each chapter. Matlab is used throughout the book with adequate listing of program codes. The publisher's website provides free download of the Matlab codes to all readers and a solutions manual for instructors. Highly recommended.
Hussein Baher's book presents the techniques of analogue and DSP in one volume, a useful idea to emphasise the continuing importance of analogue signal processing (ASP). Separate chapters are devoted to in-depth treatment of Fourier series, Fourier transform, Laplace transform and FFT. There is a chapter on ASP systems and another on stochastic signals, the remaining chapters being devoted to core DSP topics.
The treatment of the ASP and DSP topics is fairly traditional and rigorous. The "unifying" approach claimed by the author in the preface would have been innovative and interesting if it were more visible.
Nevertheless, the book is excellent for postgraduates and those new to the field who prefer a mathematically rigorous approach. But the growing number of lecturers saddled with picky "mathophobes" may also adopt the book and tread cautiously through the early chapters in order not to hurt their department's student-retention statistics.
Chi-Tsong Chen's book provides a good introduction to spectral computation and digital filter design. Fourier series, Fourier transform and FFT of analogue and discrete-time signals are lucidly discussed with careful attention to terminology. The second part of the book covers digital filters, beginning with a chapter on linear time-invariant lumped systems where the Laplace and Z-transforms are introduced. It also includes separate chapters on the design of FIR and IIR filters.
The clarity of treatment of a carefully selected set of core DSP topics coupled with a Matlab-integrated approach makes this book a good first course for novice readers. The book provides the firm grounding needed to deal with more advanced topics and applications.
Ifiok Otung is principal lecturer in electronics, University of Glamorgan.
Digital Signal Processing: Spectral Computation and Filter Design. First edition
Author - Chi-Tsong Chen
ISBN - 0 19 513638 1
Publisher - Oxford University Press
Price - £29.99
Pages - 440