| TSKS15 |
Detection and Estimation of Signals, 6 ECTS credits.
/Detektion och estimering av signaler/
For:
D
I
Ii
IT
MMAT
SY
Y
|
| |
Prel. scheduled
hours:
Rec. self-study hours: 160
|
| |
Area of Education: Technology
Main field of studies: Electrical Engineering
|
| |
Advancement level
(G1, G2, A): A
|
|
Aim:
After completed course the student should:
- with adequate terminology, in a well-structured manner and logically coherent, be able to describe and conduct simpler calculations that relate to classical and Bayesian estimation and detection theory, specifically the Neyman-Pearson theorem, error probabilities, decision regions, maximum-likelihood, linear and nonlinear models, Fisher information, Cramer-Rao bound, circularly symmetric noise, noise whitening, MMSE and LMMSE, GLRT, model order selection, coherent and non-coherent detection, composite hypothesis testing and nuisance parameters and basis expansions of waveforms in continuous time
- be able to describe, apply and implement in a conventional programming language, and show engineering understanding of, the theory and models used in the course
|
|
Prerequisites: (valid for students admitted to programmes within which the course is offered)
Linear algebra, probability theory, and a course similar to Signals, Information and Communications.
Note: Admission requirements for non-programme students usually also include admission requirements for the programme and threshhold requirements for progression within the programme, or corresponding.
|
|
Organisation:
Lectures, problem classes and computer laboratory work. Individual (inclass) oral examination of laboratory work.
|
|
Course contents:
Binary hypothesis tests, Neyman-Pearson theorem, error probability. M-ary detection problems. Bayes cost, minimum probability of error. Nuisance parameters. Classical estimation: Maximum-likelihood. Cramer-Rao bound, Slepian-Bang's formula, efficiency. Linear, vector-valued models with Gaussian noise. Non-linear models. Noise whitening, complex-valued data, Gaussian noise, circularly symmetric noise. Bayesian estimation: MMSE and LMMSE. Composite hypothesis testing: GLRT and Bayesian approach, model selection. Performance calculations, asymptotic properties of estimators. Applications to amplitude and phase estimation, frequency estimation, angle-of-arrival estimation, time-of-arrival estimation, source localization, coherent and non-coherent detection of waveforms.
|
|
Course literature:
S. Kay, Statistical Signal Processing: Estimation Theory and
Statistical Signal Processing: Detection Theory, Prenticeā?Hall.
|
|
Examination: |
|
A written examination
Laboratory work
|
4 ECTS
2 ECTS
|
| |
|
|
Course language is Swedish/English.
Department offering the course: ISY.
Director of Studies: Klas Nordberg
Examiner: Erik G. Larsson
Link to the course homepage at the department
Course Syllabus in Swedish
|
|