Lehrver­anstaltung

DSP - Digital Signal Processing


PDF Lehrveranstaltungsverzeichnis English Version: DSP

Version: 2 | Letzte Änderung: 11.09.2019 11:34 | Entwurf: 0 | Status: vom verantwortlichen Dozent freigegeben

Langname Digital Signal Processing
Anerkennende LModule DSP_MaCSN, DSP_MaTIN
Verantwortlich
Prof. Dr. Harald Elders-Boll
Professor Fakultät IME
Niveau Master
Semester im Jahr Wintersemester
Dauer Semester
Stunden im Selbststudium 60
ECTS 5
Dozenten
Prof. Dr. Harald Elders-Boll
Professor Fakultät IME
Voraussetzungen No formal requirements, but students will be expected to be familiar with:
Basic Knowledge of Signals and Systems: Continuous-Time LTI-Systems and Convolution, Fourier-Transform
Basic Knowledge of Probability and Random Variables
Unterrichtssprache englisch
separate Abschlussprüfung Ja
Literatur
John G. Proakis and Dimitris K. Manolakis. Digital Signal Processing (4th Edition). Prentice Hall, 2006.
Alan V. Oppenheim, Ronald W. Schafer. Discrete-Time Signal Processing (3rd Edition). Prentice Hall, 2007.
Vinay Ingle and John Proakis. Digital Signal Processing using MATLAB. Cengage Learning Engineering, 2011.
Abschlussprüfung
Details
In the written exam students shall demonstrate that they are able to solve problems dealing with the design, analysis and implementation of DSP systems in soft and hardware considering computational complexity and hardware resource limitation, by using their thorough understanding of the theoretical concepts, especially frequency domain analysis, and insights gained from the practical implementation of DSP systems in software using Python and on microprocessors, such that they are able to design, select, use and apply actual and future DSP systems for various signal processing application in commercial products.
Mindeststandard
Mindestens 24 der möglichen 50 möglichen Gesamtpunkte aus der Klausur und den zwei Tests während des Semesters.
In der Klausur können maximal 40 Punkte in den zwei Tests während des Semesters können maximal jeweils 5 in der Summe also 10 Punkte erreicht werden.
Prüfungstyp
In the written exam students shall demonstrate that they are able to solve problems dealing with the design, analysis and implementation of DSP systems in soft and hardware considering computational complexity and hardware resource limitation, by using their thorough understanding of the theoretical concepts, especially frequency domain analysis, and insights gained from the practical implementation of DSP systems in software using Python and on microprocessors, such that they are able to design, select, use and apply actual and future DSP systems for various signal processing application in commercial products.

Lernziele

Kenntnisse
Signals, Systems and Digital Signal Processing
Basic Elements of DSP Systems
Classification of Signals
Continuous-Time and Discrete-Time Signals
Deterministic and Random Signals
Even and Odd Signals
Periodic and Aperiodic Signals
Energy and Power of Signals
Some Fundamental Signals
Discrete-Time Linear Time-Invariant Systems
Difference Equations
Discrete-Time Convolution
Unit-Pulse and Impulse Response
Basic Systems Properties: Causality, Stability, Memory

Ideal Sampling and Reconstruction
Ideal Sampling and the Sampling Theorem
Aliasing

Fourier-Transform of Discrete-Time Signals
Eigenfunctions of Discrete-Time LTI Systems
Frequency response of Discrete-Time LTI Systems
The Fourier-Transform of Discrete-Time Signals
Ideal Continuous-Time Filters

The z-Transform
The Two-sided z-Transform
Properties of the z-Transform
The Inverse z-Transform
Analysis of LTI Systems using the z-Transform

Discrete Fourier-Transform
Sampling the DTFT
The DFT and the Inverse DFT
The Fast Fourier Transform
Radix-2 FFT Algorithms
Linear Convolution Using the FFT
Overlap-And-Add

Design of Digital Filters
Design of FIR Filters
Design of IIR Filters

Random Signals
Review of Probablity and Random Variables
Ensemble Averages
Correlation Functions
Stationary and Ergodic Processes
Power Spectral Density
Transmission of Random Signals over LTI Systems

Advanced Sampling Techniques
Quantization and Encoding
Sampling of Bandpass Signals
Sampling of Random Signals
Sample Rate Conversion
Sample Rate Reduction by an Integer Factor
Sample Rate Increase by an Integer Factor
Sample Rate Conversion by a Rational Factor
Oversampling and Noise Shaping

Optimum Linear Filters
Linear Prediction
The Wiener Filter
Orthogonality Principle
FIR Wiener Filter
IIR Wiener Filter

Spectrum Estimation
The Periodogram
Window Functions
Eigenanalysis Algorithms
MUSIC Algorithm
ESPRIT Algorithm

Fertigkeiten
Students understand the fundamentals of discrete-time signals and systems

Students can analyse the frequency content of a given signal using the appropriate Fourier-Transform and methods for spectrum estimation

Analysis of discrete-time LTI Systems
Students can calculate the output signal via convolution
Students can determine the frequency response of a given system
Students can characterize a given system in the frequency domain and in the z-domain

Implementation of discrete-time LTI systems
Students can implement the convolution sum in software
Students can implement different structures for IIR systems in software
Sudents can use the FFT to implement an FIR system

Analyze effects of practical sampling
Quantization noise
Aliasing
Trade-off pros and cons of advanced implementations like noise shaping
Aufwand Präsenzlehre
Typ Präsenzzeit (h/Wo.)
Vorlesung 2
Übungen (ganzer Kurs) 2
Übungen (geteilter Kurs) 0
Tutorium (freiwillig) 0
Besondere Literatur
keine/none
Besondere Voraussetzungen
keine
Begleitmaterial
elektronische Vortragsfolien zur Vorlesung
elektronische Übungsaufgabensammlung mit Lösungen
alte Klausuren und Lösungen
Separate Prüfung
Prüfungstyp
Übungsaufgabe mit fachlich / methodisch eingeschränktem Fokus unter Klausurbedingungen lösen
Details
Zwei semesterbegleitende Tests in Form von Aufgaben, die den bis zum jeweiligen Zeitpunkt in der Vorlesung/Übung behandelten Stoff aufgreifen und so bei Bestehen sicherstellen, dass die Grundlagen zur erfolgreichen Teilnahme an den entsprechenden Praktikumsversuchen und/oder Projekten gegeben ist.
Mindeststandard
Mindestens 2 von maximal 5 erreichbaren Punkten pro Test.

Lernziele

Kenntnisse
Review of Probablity and Random Variables
Moments, Averages and Distribution Functions

Random Signals
Ensemble Averages
Correlation Functions
Stationary and Ergodic Processes
Power Spectral Density
Transmission of Random Signals over LTI Systems

Sampling
Sampling and coding for speech and/or audio signals

Fertigkeiten
Analysis of random variables by means of
Mean and moments
Distribution

Analysis of random signals
Determine whether a given random signal is stationary or not
Analyse whether a random signal contains discrete harmonic components
by using the autocorrelation function
by using the power spectral density

Combatting noise
Remove or suppress high-frequency noise from low-pass signals

Abilty to trade-off different methods for digital coding of speech and audio signals

Determine the quatization noise and the SNR for different sampling schemes
Aufwand Präsenzlehre
Typ Präsenzzeit (h/Wo.)
Praktikum 1
Tutorium (freiwillig) 0
Besondere Literatur
keine/none
Besondere Voraussetzungen
keine
Begleitmaterial
elektronische Beschreibung der Praktikums-Versuche
Separate Prüfung
Prüfungstyp
praxisnahes Szenario bearbeiten (z.B. im Praktikum)
Details
Erfolgreiche Bearbeitung dfer Parktikumsversuche oder Projekte in Kleingruppen von in der Regel zwei Studierenden. Das Bestehen des entsprechenden Tests aus der Vorlesung/Übung ist Zugangsvoraussetzung um am Praktikum teilnehmen zu können.
Mindeststandard
Erfolgreiche Teilnehme an allen Versuchen und/oder erfolgreiche Bearbeitung von kleinen Projekten. Im entsprechenden Test in der Vorlesung/Übung müssen zum Bestehen 2 von 5 möglichen Punkten erreicht werden

© 2022 Technische Hochschule Köln