Project Details


The goal of the project is to develop and advance the state-of-the-art in nonstationary signal processing for temporal signals (time-series) and graph signals. In the context of temporal signals, nonstationarity is characterized by the time variation of the statistical properties (and frequency content) of a signal. Accurate processing of nonstationary signals necessitate models that represent signals in the joint time-frequency (T-F) domain. A related problem is that of decomposing a signal into multiple components, leading to the multiscale analysis of nonstationary data. To this end, the project will investigate and develop novel approaches for the T-F analysis and signal decomposition (SD) of time-series data.

For the emerging classes of (time-varying) graph signals, nonstationary signal processing techniques are still in infancy. For instance, fully data-driven techniques for graph signal decomposition and vertex-frequency analysis (counterpart of T-F analysis in graph signals) do not yet exist. To this end, the goal of the project is to investigate and develop data-driven nonstationary signal processing approaches for graph signals and demonstrate their utility on a range of data including biomedical EEG data and industrial vibration data sets.
Effective start/end date01/08/202231/07/2024


  • graph signal processing
  • signal processing


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.