Manuscript Due

Oct. 15, 2019 (closed)


Most real-life signals are of nonstationary nature, and their separate analysis in the time or frequency domain is often limited, not offering a comprehensive insight into the signal’s time changing spectral properties. Also, standard frequency-domain methods perform spectral analysis without spatial/temporal localization of signal features. Hence, in order to efficiently extract the signal’s useful information (such as number of components, complexity, instantaneous frequencies, instantaneous amplitudes, etc.), joint time–frequency distributions should be applied. Extracting information from nonstationary signals in the time-frequency domain is a crucial step in designing computer-aid decision-support systems based on signal features used for data analysis, machine learning, automatic classification systems, etc.

However, time-frequency signal analysis faces numerous challenges, such as design of high-resolution time-frequency distributions, reduction of cross terms in quadratic time-frequency representations, development of methods robust to noise, challenges in hardware realizations of time-frequency representations, blind source separation, processing sparse signals acquired using compressive sensing, to name a few.
This special issue is an attempt to render a comprehensive venue for recent progress in solving these challenges both in terms of proposing novel methods with their fundamental mathematical background, as well as applying the existing time-frequency techniques to a wide range of practical applications (i.e. biomedical signal processing, audio, radar, sonar, seismology and many more).

Lead Guest Editor

  • Jonatan Lerga, University of Rijeka, Croatia

Guest Editors

  • Nicoletta Saulig, Juraj Dobrila University of Pula, Croatia
  • Milos Dakovic, University of Montenegro, Montenegro
  • Cornel Ioana, Grenoble Institute of Technology/GIPSA-lab, France
  • Danilo Orlando, Faculty of Engineering, Italy
EURASIP Journal on Advances in Signal Processing