Statistical Signal Processing

Code:

EL-DSD4001

Acronym:

EL-DSD4001

Keywords
Classification	Keyword
OFICIAL	Electronics and Digital Systems

Instance: 2020/2021 - 2S

Active?	No
Responsible unit:	Department of Electrical and Computer Engineering
Course/CS Responsible:	Master's Degree in Physical Engineering

Cycles of Study/Courses

Acronym	No. of Students	Study Plan	Curricular Years	Credits UCN	Credits ECTS	Contact hours	Total Time
MI:EF	0	study plan from 2017/18	4	-	6	56	162

Teaching language

Suitable for English-speaking students

Objectives

Gain knowledge in:

1. Signal processing for systems that are subject to uncertainties, modeled by random variables.
2. Stocastic systems modeling.
3. State estimation in stocastic systems.
4. Estimation using spectral analysis.

Learning outcomes and competences

This course will enable students to use techniques and technologies of statistical signal processing in such areas as multimedia signal classification, recognition, interpretation, annotation and recommendation, as well as in other areas involving parameter estimation and machine learning, namely control, communications, and biomedicine. The methodology adopted in the course also fosters the deepening in innovation skills in these areas.

Working method

Presencial

Program

1. Signals and systems in discrete time.
a. Fourier analysis
b. Z transform
c. Random signals

2. Basic Signal Modeling
a. Least-Squares method
b. Methods of Padé, Prony and Shanks

3. The Levinson recursion
a. Cholesky decomposition
b. Toeplitz maxtrix inversion
c. Levinson recursion

4. Lattice Filters (FIR, IIR)
a. FIR and IIR lattice filters
b. Lattice methods for all-pole signal modeling
c. Stochastic modeling

5. Kalman Filters
a. Algorithm
b. Modeling
c. Implementation

5. Spectrum Estimation
a. Non-parametric methods (periodogram, Welch method)
b. Minimum-variance spectrum estimation (MLE)
c. Maximum entropy method
d. Frequency estimation using eigen-analysis (MUSIC, ESPRIT)
e. Principal components spectrum estimation

Mandatory literature

Alan V. Oppenheim; Discrete-time signal processing. ISBN: 0-13-083443-2
Monson H. Hayes; Statistical digital signal processing and modeling. ISBN: 0-471-59431-8

Teaching methods and learning activities

This curricular unit will involve theory presentation of the main topics, discussion/resolution of illustrative problems, some of which in the form of mini-tests that will be graded (25%), practical assignments involving Matlab Programming that will be graded (25%), and a final exam (50%)

Software

Matlab

Evaluation Type

Distributed evaluation with final exam

Assessment Components

Designation	Weight (%)
Exame	50,00
Teste	25,00
Trabalho escrito	25,00
Total:	100,00

Amount of time allocated to each course unit

Designation	Time (hours)
Estudo autónomo	90,00
Frequência das aulas	52,00
Trabalho escrito	20,00
Total:	162,00

Eligibility for exams

Frequency of this course is obtained through the satisfaction of both the following conditions:
1. Participation in at least 75% of the classes.
2. Obtention 25% or more in the classification of the distributed evaluation component.

Calculation formula of final grade

The final classification results from the weighted sum of:
25% Mini-tests subject to evaluation;
25% Practical assignments;
50% Final exam.