Radar Target Classification by Natural Resonances: System Analysis
Abstract
This thesis examines the system implementation considerations of a resonance based radar target classification system. The basis of the system is the aspect and excitation independent property of electromagnetic scattering from a conducting body. Such a system consists of two components: pole extraction and annihilation filtering. The algorithms investigated here for these purposes are the Cadzow-Solomon pole extraction algorithm and the K-Pulse annihilation filter. Additionally, an aspect-dependent annihilation filter based on an inverse ARMA model is introduced. The procedures are applied to noise polluted synthetic data, as well as scattering data collected for a thin-wire and silver coated 1/72 scale model aircraft.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1990
- Accession Number
- ADA241170
Entities
People
- Peter C. Reddy
Organizations
- Naval Postgraduate School