Study On Algorithm Of Inverse Synthetic Aperture Radar Imaging Based On Compressive Sensing

Inverse synthetic aperture radar(ISAR) imagery plays an important role especially in military applications such as target identification, recognition, and classification. In these applications, a critical requirement of the ISAR image is to achieve sharp resolution in both range and cross-range domains. At present, although some methods can achieve high resolution profile, these methods cannot provide good profile under the conditions that the radar data sampling rate is small and observation data are lost partially. However, as a new signal processing technique, compressive sensing(CS) can effectively resolve the problem. Therefore, how to apply CS in the field of radar imaging and obtain higher resolution ISAR imagery has the important and significance research value.This paper has analyzed three methods and it is revealed that the algorithm based on smoothed L0 norm(SL0) can achieve the higher accuracy and the faster convergence rate, which is about two to three orders of magnitude faster than another algorithms. This algorithm is to approximate the discontinuous function by a suitable continuous one, and then analyze it for solution. However, this function is highly non-smooth, and it is convex and contains a lot of local minima. The original initial value is the minimum L2 norm solution, and it is a rough solution. For this problem, this paper has proposed a modified smoothed L0 norm(MSL0) method in Chapter 4. The experimental results show that this method can give a better sparse solution. Moreover, in this paper, we have analyzed the sparseness of ISAR imagery and convert the original imaging model into a problem of signal reconstruction in the framework of high-resolution inverse synthetic aperture radar imaging with the proposed method. The experimental results show that the proposed approach can obtain high-quality image under the scenarios where the data sampling rate is small and data are lost partially. Based on multiple measurement vector(MMV) method and weighted optimization idea, this paper has proposed a Re-weighted L2,0 norm(RWL2,0) method in Chapter 5. A lot of experimental results show that this method can obtain a better reconstruction performance than another algorithms. In order to deal with the whole imaging domain, we have established a new radar echo signal analysis model and applied the proposed method under this model. The experimental results show that the proposed approach has the lower reconstruction error, and the higher signal-to-noise ratio of reconstruction and the better imaging effect.