Abstract
Image formation of large scenes is still challenging in medium-earth-orbit (MEO) synthetic aperture radar (SAR) due to the existence of severe 2-D space variance. In this paper, the properties of space variance are analyzed in detail, and then a variable-coefficient fourth-order range model is adopted to model the space-variant range history of every target in a large scene accurately. A method integrating a modified chirp scaling algorithm with joint time-Doppler resampling is proposed to address the range-variant range cell migration, as well as the azimuth-variant frequency-modulation rate and higher order Doppler parameters. The computational burden and alternative implementation approaches are also discussed. Finally, processing of simulated data for MEO SAR with 2-m resolution is presented to validate the proposed algorithm.
Original language | English |
---|---|
Pages (from-to) | 3573-3586 |
Number of pages | 14 |
Journal | IEEE Transactions on Geoscience and Remote Sensing |
Volume | 56 |
Issue number | 6 |
DOIs | |
Publication status | Published - Jun 2018 |
Externally published | Yes |
Keywords
- 2-D space variance
- azimuth resampling
- chirp scaling (CS)
- medium-earth-orbit (MEO) synthetic aperture radar (SAR)
- stripmap mode