Techniques and Algorithms for Signal Processing

An introductory course on techniques and algorithms for signal processing for students in computer science and engineering, designed to teach students the fundamental techniques of digital signal processing of analog and digital signal, using both software and hardware solutions. Topics covered: A/D and D/A conversion, oversampling, sigma-delta modulation, digital filtering, FFT algorithms and FFT processors, MAC (Multiplying ACcumulating device), signal processors, noise in digital systems, low noise structure, multi-rate processing, decimation and interpolation, adaptive filtering principles, Kalman filtering.

An introductory course on techniques and algorithms for signal processing for students in computer science and engineering, designed to teach students the fundamental techniques of digital signal processing of analog and digital signal, using both software and hardware solutions. Topics covered: A/D and D/A conversion, oversampling, sigma-delta modulation, digital filtering, FFT algorithms and FFT processors, MAC (Multiplying ACcumulating device), signal processors, noise in digital systems, low noise structure, multi-rate processing, decimation and interpolation, adaptive filtering principles, Kalman filtering.


Lecture contents
Digital systems. Representation of signals, codes, representation errors, floating-point arithmetic, noise model of representation errors, noise spectrum (3h).
Analog-digital & digital-analog converters. A/D converter characteristic, conversion errors, error spectrum and probability distribution function, converter types: flash, successive approximation, parallel/serial converters anti-aliasin filtering. D/A converters, conversion spectrum, 0 and 1 order interpolation (6h).

Digital filters. SOI & NOI filter structure, transfer characteristic, group delay, zeros and pools, filter stability, Characteristic approximation: Butterworth, Chebyshew, Elliptic. Linear phase SOI filters. Digital noise in filters, effects of coefficient finite representation (5h).

Multi-rate data processing. Over-sampling, decimation, decimation filters, comb filters, sigma-delta modulators, sigma-delta A/D converters, noise propagation in multi-rate processing. Up-sampling, interpolation and interpolation filtering (5h).

DSP hardware: PLD implementation of DSP algorithms, Multiplying-accumulating devices, Parallel and serial implementation, pipelining, Harvard architecture, DSP processors, integrated digital filters (5h).

FFT: FFT algorithms, windowing, numerical complexity, noises in FFT, FFT processors, Implementation of FFT algorithms, fast filtering, spectrum estimation, covariance function estimation, CEPSTRUM, echo cancellation (3h).

Advance DSP techniques: Adaptive filtering, high-resolution spectrum analysis, Kalman filtering (3h).






Laboratory contents
Laboratory exercises:

1. A/D conversion


2. DSP Filters


3. Noise in DSP Filters


4. Multi-rate processing


5. Sigma-delta conversion


6. FFT analysis


7. DSP processor programming



Projects contents
One computer-based, DSP-based or PLD-based individual projects are required to be completed, individual choice of subject.

1. Richard G. Lyons, Understanding Digital Signal Processing


2. Vinay K. Ingle, John G. Proakis, Digital Signal Processing Using MATLAB


3. Jayant Datta, DSP Filter Cookbook.


4. Steven W. Smith,The Scientist and Engineer`s Guide to Digital Signal Processing, www.dspguide.com