Research On Array Error Estimation Of Distributed Spaceborne SAR-GMTI Radar Systems

A new kind of small satellites group formation technology has been emphasized on in the middle of the 1990s', which was formation-flying satellites, also named distributed satellites or constellation of formation-flying satellites. This constellation is composed by several small satellites or micro-satellites usually. In this implementation, several small satellites fly in a formation and operate as a single "virtual satellite". Through the reasonable path design, the constellation can move in the given orbit reliably on the condition that it need consume very little fuel. With the rapid development of the small satellite technology, the small satellite has many advantages over a traditional large satellite such as a light of weight, a small of size, a short of development cycle, a low of price, a flexible of launch, a good survivability, a high of reliability and so on. What is more, the constellation can perform multiple space mission functions, such as high precision SAR imageing, ground moving target indication and interferometric SAR, and its performance is much better than that of a single large satellite. However, the constellation of small satellites is faced with several challenging problems at the same time. These problems include that high sparseness of the array, range or Doppler ambiguities, array measurement errors and so on. These problems should be carefully considered to ensure that the advantages of constellation small satellites are carried out.In this doctoral dissertation, ground moving target detection and array error estimation approaches are studied for distributed small satellite systems. The main returns of this doctoral dissertation are listed as follows.1. Some main factors that affect the performance of ground moving target detection for distributed spaceborne SAR are analyzed in detail. The analyzing result shows that the GMTI performance using the traditional methods such as DPCA and ATI will be affected seriously. A new approach on moving target detection and relocation is proposed based on multi-channel and multi-pixel adaptive signal processing in image domain. This method makes use of the steering vector of moving target; as a result, the detection and velocity estimation of the ground moving target can be implemented jointly in this method and need not add satellite. Furthermore, this method has a good robustness to the above factors that affect the GMTI performance, such as image registration error, clutter decorrelation and so on.2. The existence of steering vector error caused by the array errors will affect the performance of GMTI and lead to the error of moving target velocity estimation. Accordingly, the array error estimation and calibration method for the linear array along track is studied and a new error estimation method is proposed. In this method, the echo signal of a very visible point on the ground is used to estimate the along-track errors of satellites and other baseline error components including gain and phase errors are estimated combined with self-calibration method. This method can obtain a high estimation precision and need not many times complex matrix inversion and iterative operation.3. On the basis of the above error estimation method, the affections of the array errors on the performance of SAR imaging and GMTI are analyzed respectively. Many simulation results show that different kinds of errors have different affect on different applications such as two-dimensional SAR imaging and GMTI. After the errors are compensated using the above error estimation method, a high-precision SAR image and high performance GMTI result can be obtained.4. The futural spaceborne SAR is required to detect the ground moving target and to generate three-dimensional image of the fixed scene. For this kind of three-dimensional array, the performance of GMTI will be affected by the height of topography, so that the elevation of topography shoud be estimated and compensated. High-precision baseline parameters are necessary to get the precise height information of the ground. A new baseline estimation method for InSAR is proposed. This method is based on subspace projection and in which the information of some points on the ground is used. This method can obtain a explicit solution of baseline errors and need a little amount of computation to gain a high estimation precision. It has a higher robustness to the unwrapping phase and can restrain the noise very well.5. On the basis of the above method, a more robust ground moving target detection and velocity estimation method is proposed. In this method, the multi-channel and multi-pixel joint data is equaled to a simple array model. The real steering vector of the moving target on the condition that there is image registration error can be estimated through the space projection approach. And the estimation precisions of cross-track velocity and real azimuth are much higher.6. The blind speed and speed ambiguity are two very important and difficult problems in GMTI. For the blind speed, the author proposes that it can be eliminated by optimizing the baselines of the small satellites constellation. But in practical, for the number of satellites is very limited, the performance of this approach is restricted seriously. Therefore, another elimination blind speed method is proposed through optimizing the baselines under some constraint conditions. For the speed ambiguity, the solving method is analyzed in theory. And when the number of satellites is very few, another feasible solving method is proposed aimed at the ground moving target detection method described in this paper.