Logo UAB
2022/2023

Digital Signal Processing

Code: 102687 ECTS Credits: 12
Degree Type Year Semester
2500898 Telecommunication Systems Engineering OB 3 1

Contact

Name:
José A. Lopez Salcedo
Email:
jose.salcedo@uab.cat

Use of Languages

Principal working language:
english (eng)
Some groups entirely in English:
No
Some groups entirely in Catalan:
No
Some groups entirely in Spanish:
No

Teachers

Francisco Jose Fabra Cervellera

Prerequisites

It is recommended to have passed the following courses: Calculus, Algebra, Statistics, Discrete Systems and Signals, Fundamentals of Communications.

Objectives and Contextualisation

Once completed the subject, the student will be able to:

  • Use vector and matrix algebra normally.
  • Operate with numerical series and stochastic processes.
  • Rigorously use different probabilistic tools.
  • Estimate the parameters of a model from the signal samples at its output.
  • Estimate the power spectral density of a random process.
  • Design optimal filters in the MMSE sense and implement them in an efficient manner using iterative/adaptive algorithms.
  • Apply signal processing techniques to situations in real life.

Competences

  • Apply deterministic and stochastic signal processing techniques to the design of communication subsystems and data analysis.
  • Develop personal attitude.
  • Develop personal work habits.
  • Develop thinking habits.
  • Learn new methods and technologies, building on basic technological knowledge, to be able to adapt to new situations.
  • Perform measurements, calculations, estimations, valuations, analyses, studies, reports, task-scheduling and other similar work in the field of telecommunication systems.

Learning Outcomes

  1. Adapt the knowledge and techniques of the digital signal treatment in accordance with the characteristics of communication systems and services as well as fixed or mobile work scenario.
  2. Adapt to unforeseen situations.
  3. Analyse and specify the fundamental parameters of communication subsystems from the point of view of the transmission, reception and digital treatment of signals.
  4. Analyse the advantages and disadvantages of different technological alternatives or the implementation of communication systems from the point of view of digital signal treatment.
  5. Apply adaptive statistical filtering and control theory to the design of dynamic algorithms for the coding, processing and transmission of multimedia information. Apply multichannel signal processing to the design of fixed and mobile antenna grouping based communication systems.
  6. Apply detection and estimation theory to the design of communication receivers.
  7. Apply statistical signal processing to estimate synchronisation parameters in digital communication and radio-navigation receivers.
  8. Autonomously learn new knowledge related with digital signal processing in order to conceive and develop communication systems.
  9. Be able to analyse, encode, process and transmit multimedia information employing analogue and digital signal processing techniques.
  10. Describe the operational principles of radio-navigation, its architecture and the techniques for dealing with its sources of error.
  11. Develop critical thinking and reasoning.
  12. Develop curiosity and creativity.
  13. Develop independent learning strategies.
  14. Develop mathematical models to simulate the behaviour of communication subsystems and to evaluate and predict features.
  15. Develop scientific thinking.
  16. Develop the capacity for analysis and synthesis.
  17. Generate innovative and competitive proposals in professional activity.
  18. Manage available time and resources.
  19. Manage information by critically incorporating the innovations of one's professional field, and analysing future trends.
  20. Propose innovative solutions for problems related with the transmission, reception and the digital treatment of signals.
  21. Work in complex or uncertain surroundings and with limited resources.

Content

1. Introduction

  • Discrete random processes, frequency representation.
  • Fundamentals of matrix algebra.
  • The autocorrelation matrix.

2. Estimation theory

  • Fundamentals of model-based methodology.
  • Classical vs. bayesian estimation.
  • MVU criterion and properties of good estimators.
  • Maximum likelihood estimation.
  • Cramér-Rao lower bound.
  • Suboptimal estimation methods.
  • Applications in communication and positioning systems.

3. Spectral estimation

  • Non-parametric methods.
  • Capon or minimum variance method.
  • Parametric methods.
  • Super-resolution methods.
  • Applications in voice coding and multi-antenna signal processing.

4. Wiener filtering and adaptive filtering

  • Minimum mean square error (MMSE) estimation.
  • Linear prediction.
  • Steepest descent method.
  • Convergence criteria.
  • Least Mean Square (LMS) method.
  • Applications in noise cancellation.

Methodology

Directed activities:

  • Lectures, which convey the theoretical contents of the course.
  • Exercise classes, where exercises related to the theoretical contents of the course are solved by the lecturer with the participation of the students.
  • Laboratory classes for the application of the contents conveyed during the lectures using Matlab.
  • Written assessment tests.


Autonomous activities

  • Study of the theoretical and practical contents of the course. Preparation of the problem solving. Exam preparation.
  • Practical assignments: complete the laboratory reports and consolidate the knowledge acquired during the laboratory classes.

Annotation: Within the schedule set by the centre or degree programme, 15 minutes of one class will be reserved for students to evaluate their lecturers and their courses or modules through questionnaires.

Activities

Title Hours ECTS Learning Outcomes
Type: Directed      
Exercise classes 15 0.6 1, 4, 3, 5, 7, 6, 8, 10, 15, 16, 14, 11, 20, 9
Laboratory classes 25 1 4, 3, 5, 7, 6, 10, 13, 14, 9
Lectures 60 2.4 4, 3, 5, 7, 6, 10, 15, 16, 12, 14, 9
Type: Supervised      
Tutorials 15 0.6 5, 7, 6, 10, 11, 9
Type: Autonomous      
Prepare laboratory classes 30 1.2 1, 4, 3, 8, 13, 16, 14, 11, 17, 19, 20
Problem solving 40 1.6 1, 4, 3, 5, 7, 6, 8, 10, 15, 16, 14, 11, 20, 9
Study 100 4 1, 4, 3, 5, 7, 6, 8, 10, 15, 13, 16, 14, 11, 19, 20, 9

Assessment

Continuous evaluation

It consists of an exam at half the semester (Exam1) and another one at the end of the semester (Exam2). The average mark of these two exams leads to the continuous evaluation mark for the theory classes:

MarkTheory = [(0.35 x Mark Exam1) + (0.35 x Mark Exam2)] / 0.7

Additionally, the mark for the laboratory classes is given by the weighted average of the mark from the laboratory reports, the mark from the "in-situ" reports and the mark from the laboratory exam, as follows:

MarkLab = [(0.13 x MarkLabReports) + (0.13 x MarkLabReport "in-situ") + (0.04 x MarkLabExam)] / 0.3

With the theory and laboratory marks, the final mark of the course will be computed as:

If (MarkTheory >= 3.5) && (MarkLab >= 3.5)   --> FinalMark = (0.7 x MarkTheory) + (0.3 x MarkLab)

If (MarkTheory < 3.5)  || (MarkLab < 3.5)    --> FinalMark = min{MarkTheory, MarkLab}

Therefore, the student should get a mark at both theory and laboratory classes equal or greater than 3.5 to be eligible to have its marks being averaged.


Re-assessment

Students who have been evaluated of the two exams of the continuous evaluation, and have obtained a final mark of the course equal or greater than 3.5, will be eligible to attend the re-assessment exam.

It should be borne in mind that if a student has the laboratory mark smaller than 3.5, then the final mark of the course will also be such laboratory mark smaller than 3.5 (according to the formula for computing the final mark shown above). Therefore, since the final mark is smaller than 3.5, the student will not be eligible to attend the re-assessment exam.

In contrast, if the final mark of the course is equal toor greater than 3.5, the student will be eligible to attend the re-assesment exam. This exam will be carried out within the period of exams published by the School. In this exam, the student can re-assess the part corresponding to the Exam1, the part corresponding to the Exam2, or both. The mark obtained in each part of the re-assessment exam (mark ExamRA1, mark ExamRA2) supersedes the previous mark that the student had in the corresponding continuous evaluation exam. The re-assessment mark is computed as follows:


MarkTheoryRA = [(0.35 x Mark {Exam1 or ExamRA1}) + (0.35 x Mark {Exam2 or ExamRA2})] / 0.7

The final mark of the course is then computed as follows:

If (MarkTheoryRA >= 3.5) && (MarkLab >= 3.5) --> FinalMark =  (0.7 x MarkTheoryRA) + (0.3 x MarkLab)

If (MarkTheoryRA < 3.5)   --> FinalMark = MarkTheoryRA

Laboratory rules

  • The assistance to all laboratory classes is compulsory.
  • Laboratory exercises will be carried out individually.
  • The report of each laboratory session must be delivered one week after the end of the last laboratory session of that experiment. Any delay in the delivery of the report will be penalised in the corresponding mark.
  • The laboratory professor may request the students to deliver a preliminary version of the laboratory report when finishing the session ("in situ" report), in order to evaluate the progress of the student during the lab session.



Repeating students

If they had passed the laboratory part of the course in previous years, they will not need to do the laboratory classes again. By default, the marks that they obtained inthat part will be maintained. If a repeating student would like to repeat the laboratory classes this year, then he/she will need to informthe professor in charge of the course.

 

Consideration of "Not assessable"

Students not performing any examination (neither the two exams of the continuous evaluation nor the re-assessment exam) will be marked as "Not assessable".



Additional considerations

Without prejudice to other disciplinary measures that may deem necessary in accordance to the academic regulation, any irregularity committed by the student that may alter the mark of an assessment activity will lead this activity to be marked with zero points.  For instance, copying or letting copy a laboratory report or any another assessment activity will involve to fail it with a mark equal to zero. Furthermore, such activity will not be able to be re-assessed during the same academic course. If this activity has a minimum mark associated to it, then the subject will be graded as failed.

If the student commits irregularities in various assessment activities of the course, the final mark of the course will be 0 in virtue of point 10, article 116, of the Academic Regulation.

Assessment Activities

Title Weighting Hours ECTS Learning Outcomes
Exam 1 35% 2.5 0.1 1, 4, 3, 5, 7, 6, 8, 10, 16, 14, 19, 20, 9
Exam 2 35% 2.5 0.1 1, 2, 4, 3, 5, 7, 6, 8, 10, 16, 14, 19, 20, 9
Laboratory: "in-situ" reports 10% 4 0.16 4, 8, 15, 13, 16, 12, 14, 11, 17, 19, 20, 21
Laboratory: exam of practice knowledge 5% 1 0.04 1, 4, 3
Laboratory: practice reports 15% 5 0.2 1, 4, 3, 5, 6, 8, 15, 13, 16, 12, 11, 18, 20, 9

Bibliography

Basic:

  • S. M. Kay, Fundamentals of statistical signal processing. Estimation theory, vol. I, Prentice-Hall, 1993.
  • M. H. Hayes, Statistical digital signal processing and modeling, John Wiley and Sons, 1996.
  • P. Stoica and R. Moses, Spectral analysis of signals, Prentice-Hall, 2005.
  • S. Haykin, Adaptive filter theory, Pearson, 2013.


Supplementary:

  • D. G. Manolakis, V. K. Ingle, S. M. Kogen, Statistical and adaptive signal processing: spectral estimation, signal modeling, adaptive filtering and array processing, Artech-House, 2005.
  • S. M. Kay, Fundamentals of statistical signal processing. Practical algorithm development, vol. III, Pearson, 2013.
  • S. Lawrence Marple, Digital spectral analysis, Dover Publications, 2019.
  • B. Widrow and S. D. Stearns, Adaptive signal processing, Prentice-Hall, 2985.
  • A. Hjorungnes, Complex-Valued Matrix Derivatives: With Applications in Signal Processing and Communications, Cambridge University Press, 2011.
  • Fundamentals:
    • S. M. Kay, Intuitive probability and random processes using Matlab, Springer, 2006.
    • V. K. Ingle and J. G. ProakisManolakis, Digital signal processing using Matlab, Cengage Learning, 2012.

Software

  • The laboratory sessions will be carried out using the software MATLAB.
  • The laboratory reports are encouraged to be written using LaTeX, for instance, using the online editor Overleaf.