Talks and Poster Presentations (without Proceedings-Entry):

C. Mecklenbräuker, P. Gerstoft:
"Sparse Bayesian Learning for Wavefields from Sensor Array Data";
Talk: Information Theory and Applications Workshop (ITA 2016), San Diego (CA), USA (invited); 01-31-2016 - 02-05-2016.

English abstract:
We estimate the directions of arrival of plane waves from sensor array data based on sparse Bayesian learning (SBL). Assuming zero-mean circularly symmetric complex normally distributed source amplitudes with unknown hyperparameters (the power levels), the corresponding posterior distribution is derived given the array data observations. To determine the hyperparameters (source and noise power levels), the evidence is maximized. The resulting SBL scheme is discussed in detail and evaluated competitively to compressed sensing ($ell_1$-regularization), the conventional beamformer, and the minimum variance distortion-free response.

sparsity, LASSO, convex optimization, Bayesian learning

Electronic version of the publication:

Related Projects:
Project Head Christoph Mecklenbräuker:
Christian Doppler Lab "Funktechnologien für nachhaltige Mobilität"

Project Head Christoph Mecklenbräuker:
Kompression des rückgemeldeten Kanalzustands für zeitvariante MIMO Kanäle

Created from the Publication Database of the Vienna University of Technology.