The research activities of the team SC are divided into two areas:
- First area : Statistical Signal Processing
- Second area : Signal for communications
First area : Statistical Signal Processing
The group is working on classical topics such as parametric modelling, estimation, detection and classification. Some original activities on simulation methods have also been conducted during the recent years.
These activities are detailled in what follows:
- Analysis and modelling of stationary and non stationary signals : parametric models, evolutionary models, non-uniform sampling, neural networks and Bayesian estimation
- Detection and classification : Bayesian methods, detection by using time-frequency and time scale representations, classification by using Monte Carlo Markov chain methods
- simulation methods : these methods have been used to solve various signal processing problems on estimation, detection and classification. The group is working on off-line methods based on the Gibbs sampler and the Metropolis-Hastings algorithm. Some works on sequential methods based on Kalman filtering and particle filtering are also studied and applied to global positioning systems.
- Compression: some activities are conducted on the compression of biomedical signals and hyperspectral images.
Second area : Signal for communications
Signal processing methods are developed in the context of mobile and satellite communications. These methods are applied to multiple access, modeling and identification of communication channels, synchronization, coding and equalization.
More specifically, new activities have been recently studied in these directions :
- Iterative processes (Turbo codes, CPM...).
- Multiple access (spread spectrum, time-varying filters).
- Identification and equalization of non-linear channels with neural networks or nonlinear devices (Volterra series, non-linear models)
- Multicarrier systems (OFDM, multicarrier-CDMA, ...)
- Digital receivers (algorithms, architecture, multi-user techniques)
- Performance computation in real channels (effects of non-linearities, phase noise, receiver imperfections, ...)
- Error modelling in various systems such as DVB-RCS
- Multi-layer optimization: Cross-layer, classification of modulations.
This section doesn't contain any article.