Ahmad Bazzi

Ahmad Bazzi

he/him

Research Scientist

Detection of the Number of Superimposed Signals Using Modified MDL Criterion: A Random Matrix Approach

Published in ICASSP 2016 - IEEE International Conference on Acoustics, Speech and Signal Processing, 2016

This work addresses estimating the number of superimposed signals from noisy observations of an N-antenna array when the number of snapshots L is of order N (so classical asymptotic regimes do not apply). Using random matrix theory, a modified Minimum Description Length (MDL) estimator is proposed that improves detection accuracy in this regime relative to classical MDL.

Recommended citation: A. Bazzi, D. T. M. Slock and L. Meilhac, "Detection of the Number of Superimposed Signals Using Modified MDL Criterion: A Random Matrix Approach," in ICASSP 2016 - IEEE International Conference on Acoustics, Speech and Signal Processing, Shanghai, China, 2016. https://ieeexplore.ieee.org/document/7472547/

Show BibTeX
@article{bazzi2016detection,
  title   = {Detection of the Number of Superimposed Signals Using Modified MDL Criterion: A Random Matrix Approach},
  author  = {Ahmad Bazzi and Dirk T. M. Slock and Lisa Meilhac},
  journal = {ICASSP 2016 - IEEE International Conference on Acoustics, Speech and Signal Processing},
  year    = {2016},
  month   = {mar},
  publisher = {IEEE},
  doi     = {10.1109/ICASSP.2016.7472547},
  url     = {https://doi.org/10.1109/ICASSP.2016.7472547},
}