Registration is closed for the event requested
Non-stationary signal processing in distributed configuration: application to the monitoring of natural environments
#"Non-stationary
#signal
#processing
#in
#distributed
#configuration:
#application
#to
#the
#monitoring
#of
#natural
#environments"
#by
#Drs.
#Cornel
#IOANA
#and
#Gabriel
#VASILE
#Grenoble
#Institute
#Technology/GIPSA-lab
Monitoring natural environments, such as underwater environments, requires powerful tools for signal analysis, mainly due to the uncertainty of the signal’s sources as well as the diversity of the signal’s types. Addressing this first factor – sources are unknown in terms of transmitted signals and positions – requires a distributed network of autonomous sensors with low energy consumption and reduced algorithmic complexity, in order to achieve low installation and maintenance costs.
The second factor – diversity of signal type – is apparently in contradiction with reduced algorithmic complexity: generally, the signals are non-stationary with non-linear time-frequency and/or transient components, and their analysis is naturally driven in the
two-dimensional time-frequency domain.
Our works propose an alternative that will make signal analysis tractable by low complexity algorithms. The key to this alternative is the use of time-frequency-phase continuity that, implemented via 1D processing operators, constitutes an efficient trade-off between signal modeling and algorithmic complexity. Theoretical and practical issues are combined in order to achieve an efficient distributed non-stationary signal processing architecture. Examples from real-life applications – acoustic sensing and monitoring in an underwater environment – will be presented.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
- Contact Event Host
- Alfredo Tan at tan@fdu.edu; Hongya Ge at hongya.ge@njit.edu
- Co-sponsored by School of Computer Sciences and Engineering, FDU