Ph.D. Level Courses
Advanced Topics in Detection and Estimation Theory
This course will treat selected advanced topics in estimation and detection theory. For an outline of specific topics that we plan to cover, see below.
InstructorsProf. Prof. Erik G. Larsson, ISY/Communication Systems, and Dr. <Unknown username: saif>, ISY/Communication Systems
RegistrationThe course is open to students enrolled in a Ph.D. program at Linkoping University/ISY. External participants upon request. If you have interest to participate, please register by sending an email to email@example.com, by March 1, 2011.
PrerequisitesGood knowledge of linear algebra and probability theory. General mathematical maturity. Knowledge of basic discrete-time detection/estimation theory (Chapter 2 in van Trees book, or equivalent) taught for example in our first detection/estimation course.
Tentative course outline
- Nonlinear estimation:
- Array signal processing
- Cramer-Rao bounds under constraints
- Other performance bounds
- Conditional and unconditional Maximum-Likelihood (a.k.a. deterministic and stochastic ML) techniques, concentrated ML
- Techniques for asymptotic analysis of estimators
- Model order selection, detection of number of sources
- Asymptotic analysis of detection algorithms
- Basic spectral estimation
- Periodograms and weighted periodograms
- Basic parametric models and methods
- Direction-of-arrival and time-of-arrival estimation algorithms
- Source localization/positioning systems
- Detection of weak signals and signals with structure
Schedule and reading
- The course will consist of about 9 seminars. Course start is March 10, 2011, at 9.15 am (room: Algoritmen).
- Tentative course schedule is here (this will be continuously updated during the course).
- Reading assignments are here (this will be continuously updated during the course).
- Reading assignments for the last three lectures on Spectral Estimation are here.
- Some lecture notes:
Key concepts in asymptotic analysis (by Prof. Erik G. Larsson);
Applications of asymptotic analysis (by <Unknown username: tvk>);
Notes on CML/UML (by <Unknown username: saif>);
Derivation of Cramer-Rao bound for unconditional maximum likelihood direction of arriaval estimation (by <Unknown username: lindblom>);
Brief note on asymptotic equivalence (by Prof. Erik G. Larsson);
MUSIC and its Asymptotic Distribution (by <Unknown username: mirsad>);
Cramer Rao Lower Bound (by <Unknown username: nqhien>);
A good survey paper on Array Signal Processing Research ;
Notes on Model Selection (by Umut Orguner);
Consistency of model order selection via BIC for the array model (by Niklas Wahlström);
Notes on Covariance Matching Estimation Technique (by Daniel Eriksson);
Notes on Nonparametric Spectral Estimation (by <Unknown username: axell>);
Notes on Parametric Spectral Estimation (by <Unknown username: reza>);
Matlab programs for Parametric Spectral Estimation (by <Unknown username: reza>);
Useful reference books
- H. van Trees, "Detection, estimation and modulation theory: volume I", Wiley.
- H. van Trees, "Detection, estimation and modulation theory: volume IV (optimum array processing)", Wiley.
- S. Kay, "Fundamentals of statistical signal processing: parts I and II", Prentice-Hall.
- P. Stoica and R. Moses, "Spectral analysis of signals," Prentice-Hall.
Credits and ExaminationThe course can give 5 ECTS.
- The focus is on depth reading, presentation and discussion of research papers.
- Participants taking the course for credit are expected to participate actively in all seminars and discussions.
Last updated: 2019 07 29 15:48