NICOP - Adaptive Algorithms for Independent Component/Vector Extraction
Abstract
We plan to develop adaptive algorithms for passive tracking of distant sources of vibrations(sound) or of electro-magnetic waves in a multi-path (acoustic) environment using multiplesensors. The methods will operate as intelligent self-steering radars/sonars. Be"ing initializedin an approximate direction, they ne-tune themselves towards a closest independent source,localize it and separate""s its signal.The algorithms will be developed under a new concept of Independent Component/VectorAnalysis (ICA/IVA), called Indepe""ndent Component/Vector Extraction (ICE/IVE), where thegoal is to extract one source of interest. The source of interest will be sou"ght as an independentcomponent of a convolutive mixture of signals observed by the array of sensors. The goal is todevelop fast ad"aptive methods that are capable to process signals from arrays with hundreds ofsensors, in real-time.1
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 23, 2018
- Source ID
- N629091812040
Entities
People
- Zbynk Koldovsk
Organizations
- Office of Naval Research
- Technical University of Liberec
- United States Navy