Blind Time-Frequency Analysis for Source Discrimination in Multisensor Array Processing
Abstract
The project has focused on advancing the theory of time-varying spectrum analysis and its applications to the important area of sensor array processing. We pioneered the development of multi-sensor receivers based on quadratic time-frequency and joint-variable distributions, and have provided the theoretical framework for solving direction finding and blind source separation problems using bilinear transforms. Time-frequency signatures of signal arrivals have been, for the first time, utilized for improved detection, angle-of-arrival estimation, and blind signal recovery. Time-frequency distributions have been elevated from a purely temporal processing tool to a more powerful spatio-temporal processing technique. Our work has shown that time-frequency distributions represent the natural mechanism to process narrowband signals in broadband communication platforms, stationary signals in a dominating nonstationary environment, and spread spectrum communication signals contaminated by powerful jammers.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2001
- Accession Number
- ADA395852
Entities
People
- Moeness Amin
Organizations
- Villanova University