Abstract
This paper discusses the use of Capon's minimum-variance method (MVM) and Amplitude and Phase EStimation (APES) spectral-estimation algorithms to synthetic aperture radar rangeazimuth focusing. The rationale of the algorithms is discussed. An implementation of a Capon or APES processing chain is explained, and processing parameters such as chip-image size, resampling factor, and diagonal loading are discussed. For multichannel cases, a joint-processing approach is presented. A set of Monte Carlo simulations are described and used to benchmark Capon- and APES-based processing against conventional matched-filter-based approaches. Both methods improve the resolution and reduce sidelobes. APES yields generally better estimates of amplitude and phase than Capon but with worse resolution. Results with RADARSAT-2 quad-polarization data over Barcelona are used to qualitatively study the real-life performance of these algorithms.
Original language | English |
---|---|
Article number | 5411775 |
Pages (from-to) | 2388-2402 |
Number of pages | 15 |
Journal | IEEE Transactions on Geoscience and Remote Sensing |
Volume | 48 |
Issue number | 5 |
DOIs | |
Publication status | Published - May 2010 |
Externally published | Yes |
Keywords
- 2-D spectral analysis
- Adaptive processing
- Synthetic aperture radar