See TUMonline

Basics knowledge in Compressive Sampling and reconstruction methods.

ATTENTION: This lecture will be probably given as a block lecture. For planing time schedule we meet on 21.4 at 15:15 in N4410 "Compressive Sampling" (CS) or "Compressed Sensing" is a complete new research field developed rapidly over the last decade and gained a huge impact in almost every area of signal processing, communication, imaging and many other engineering fields. The assumption in CS is that in a very large but finite system, which can be described by n parameters, only few, i.e., k<<n of them determine the signals or states of the system. It is only known that k parameters at most are needed, but not which one. This is called a k-sparse system and it is known nowadays that there exists alternative measuring methods and algorithms which can reconstruct the signals with fewer samples as needed in classical sampling methods. To explain this compressive sampling method we will introduce the student into geometry, approximation theory in finite dimension and convex optimization theory.

Linear algebra, mathematical interest, Signal theory, Mathematics 1-4

Blackboard

oral exam

S. Foucart, H. Rauhut "A mathematical introduction to Compressed Sensing" , Eldar, Y. & Kutyniok, G. "Compressed Sensing: Theory and Applications", R.Baraniuk, E. Candes, Romberg, Davenport - Lecture Notes and Tutorials (on http://dsp.rice.edu/cs) D. Luenberger - "Optimization by vector space methods"