Robust and Sparse M-Estimation of DOA

A set of MATLAB codes for direction-of-arrival (DOA) estimation, beamforming.

Features

The codes provide:

-Beamforming based on general loss functions

-Four loss functions: 1. Maximum-Likelihood (ML) loss for the circularly symmetric complex Gaussian distribution, Gauss loss / 2. ML loss for the complex multivariate t-distribution (MVT) / 3. Huber loss / 4. Tyler loss

Citation

-C. F. Mecklenbräuker, P. Gerstoft, E. Ollila, and Y. Park, “Robust and Sparse M-Estimation of DOA,” Signal Process. 220 (2024), 109461, pp. 1-10. (ISSN 0165-1684. doi:10.1016/j.sigpro.2024.109461, Available online Mar. 7, 2024.)
-C. F. Mecklenbräuker, P. Gerstoft, and E. Ollila, “DOA M-estimation using sparse bayesian learning,” in Proc. IEEE ICASSP (2022), pp. 4933–4937.

A numerically efficient SBL implementation is available. [CODE]
-P. Gerstoft, C. F. Mecklenbräuker, A. Xenaki, and S. Nannuru, “Multi-snapshot sparse Bayesian learning for DOA,” IEEE Signal Process. Lett. 23(10) (2016).

Updates

Version 1.0: (03/13/2024 by Y. Park)

Contact

Christoph F. Mecklenbräuker
Institute of Telecommunications/TU Wien
cfm@tuwien.ac.at

Esa Ollila
Dept. of Signal Processing and Acoustics, Aalto University, Finland
esa.ollila@aalto.fi

Peter Gerstoft & Yongsung Park
MPL/SIO/UCSD
gerstoft@ucsd.edu
yongsungpark@ucsd.edu

Noiselab