The fast moving target indication and imaging have been discussed in stripmap SAR. Firstly the echo model of a moving target with fast cross-track velocity has been established, and the Doppler centroid ambiguity problem has been analyzed. It is concluded that the location of azimuth spectra cannot reflect the actual value of Doppler centroid, while the RWM is not affected by the Doppler centroid ambiguity.
Based on the echo analysis, a Doppler centroid estimation algorithm based on curve fitting, a multiple target indication and extraction method, and a fast moving target imaging algorithm based on Hough transform and third-order PFT have been proposed.
The Doppler centroid estimation algorithm is a modification of the energy balancing algorithm. By eliminating the strong targets in the scene, the homogeneity of the scene is alleviated, and the estimation accuracy is improved.
Consider the multiple target indication with different motion parameters, the indication algorithm uses two steps to detect the targets and extract their spectra. The indication and extraction are the foundation of GMTIm, and is highly suitable for practical applications.
The fast moving target imaging algorithm uses Hough transform to estimate the slope of the RWM to derive the cross-track velocity and actual Doppler centroid, and uses the third-order PFT to estimate the azimuth parameters and refocus the target. The advantages of this algorithm include: first, the Doppler centroid ambiguity problem is solved; second, the third-order phase error caused by the fast cross-track velocity is corrected; third, the targets that not located at the center of the scene can be accurately relocated by using PFT; fourth, the cross-track acceleration can be correctly estimated.
The algorithms proposed in this chapter are effective for both the fast moving targets and slow moving targets. Both simulations and real data have been used to demonstrate the effectiveness of these algorithms. However, for the targets with slow velocities and weak RCS, these algorithms show poor performances due to the limitations of single-antenna systems.