Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 40837 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Bastani, Mohammad Hassan
- Abstract:
- Although up to now different processing algorithms have been proposed for Synthetic Aperture Radar (SAR) raw data, all of them suffer from one common problem and that is huge amount of data to be processed. So because of current system limitations, efficient compression algorithms for processing, saving, or transmitting data are needed. Up to now many algorithms have been proposed for SAR raw data compression, but each of them has some defects that should be payed attention to. The most important reason of these defects is the special characteristics of SAR images. With the aid of “Compressed Sensing (CS)”, the new field which has emerged recently, a special characteristic of the scene reflection function called “sparsity” can be used to reduce the sampling rate of the received signal. So from the beginning of the process, there are fewer samples to be dealt with but nonlinear reconstruction algorithms should be used to recover the reflection function. Also with this technique, the amount of data to be saved or transmitted is reduced. In this work, based on CS theory and according to the previous related works, a number of new algorithms have been proposed to reduce the amount of SAR raw data. Based on the imaging dimension, these algorithms categorize into two groups: one dimensional and two dimensional imaging. One dimensional imaging algorithm is on the basis of Born hypothesis i.e. the received signal is the convolution of the transmitted pulse and the reflection function. Of two dimensional imaging algorithms, one is based on old SAR processing algorithms but with better efficiency. The other two algorithms use new processing methods. One of them is not sensitive to the radar pulse but the other is. However in the Fourier domain, both of them provide good inspection of how the reflection function is reconstructed. Although in these algorithms the number of available samples is much fewer, their performance is much better. These algorithms can be used for scenes which have reflection function changing with squint angle and/or radiation frequency as well as scenes with moving targets.
- Keywords:
- Synthetic Aperture Radar (SAR) ; Mutual Coherence of Measurement Matrix ; Scene Reflection Function Reconstruction ; Radar Resolution ; Compressive Sensing ; Sparse Signal Processing
Digital Object List
-
محتواي پايان نامه
- view
Bookmark