Study On STAP Of Nonhomogeneous Clutter Suppression For Airborne Array Radar

The space-time adaptive processing (STAP) is an important technique for ground moving target detection (GMTD) against strong clutter and interference in airborne radar system, and the performance of STAP is determined by the clutter plus noise covariance matrix estimation. When the clutter samples occur heterogeneous distribution, then the clutter suppression performance declines greatly due to the covariance matrix estimate inaccuracy. The motive of this thesis is that makes STAP obtain anticipant performance by the clutter within heterogeneity. The main work can be summarized as follows:1. In chapter 2, the clutter of different antenna array configuration for airborne radar is modeling. Then the relationship between the clutter distribution characteristic and the configuration is analysised. The principles of the STAP algorithms are described. The clutter occurs heterogeneous distribution on the two sides; the one is the clutter power heterogeneity, the other is antenna array configuration heterogeneity.2. In chapter 3, the Direct Data Domain (DDD) based methods are studied by the clutter within heavy power heterogeneity. As is known that the DDD method obtains enough samples from one range gate by sub-aperture smoothing operation, however, gets the space-time aperture lost. Due to the Multistage Wiener Filter (MWF) does not need covariance matrix estimation and inverse, a new DDD method that hybrid MWF is proposed. This algorithm can obtain better performance than conventional DDD method within low aperture loss. The results of the experiments on the MCARM data show the validity of the method. Usually, the target information spreads due to the range compression by using the window function. To deal with this problem, a robust DDD method is proposed. The algorithm can make full use of the additional information on range and therefore has a deeper notch against isolated interference. Moreover, it is robust to the steering vector with gain and phase uncertain errors. The experiments carried on the MCARM data shows validity of the algorithm3. In chapter 4, the clutter heterogeneity by the antenna array configuration without range ambiguity is studied. To deal with the problem that the moving target detection performance decreases, due to that the spatial angle frequency shifting algorithm enhances the side lobe of the reference range clutter spectrum. Aiming at this problem, a new spatial angle frequency shifting spectrum compensation for forward looking radar is proposed. The algorithm compensates the clutter Doppler frequency of each range cell before using spatial angle frequency shifting. The results of experiment show that the improved algorithm can obtain better performance than spatial angle frequency shifting, and without performance loss under radar clutter with some uncertain errors. To deal with the disadvantage of heavy computational load for derivative based updating (DBU) method, an approach named reduced dimensional DBU algorithm is proposed. The algorithm using matrix block inverse theorem to get revised matrix, then use the reduced dimensional space time adaptive processing(STAP)method of 3-DT. This method can decrease the computational load, reduce the required number of the independent identically distributed (IID) samples, and reach to the original optimal DBU method performance.The results of experiment show that the proposed method still can obtain good performance near the optimal processing when some uncertain errors exist.4. In chapter 5, the clutter heterogeneity by the antenna array configuration with range ambiguity is studied. Due to the linear array has not elevation degree, the short-range clutter cannot be suppressed directly when the range ambiguity exists. By using the pre-information of the clutter spectrum of the non-SLAR, a short-range clutter suppression method for non-SLAR under uniform linear array is proposed. This method estimates the elevation angles of the ambiguous short range gates, and then eliminates short range clutter by space time interpolation while adding the moving target protection in the method. This method can achieve considerable performance of short range clutter suppression. Simulation results show the validity of the method. Especially, the direction-Doppler curves are symmetry for FLAR. Based on this characteristic, a simple and efficient algorithm to suppress range ambiguous clutter for FLAR is proposed. The method uses the data after Doppler filter project to the space of the spatial steering vector which belongs to the middle of the range number. The method reduces the computation complexity and can obtain better performance. The computer simulation results show the valid of this method.5. In chapter 6, the robust nonhomogeneous clutter suppression by the clutter with uncertain errors is studied. A robust elevation prefiltering method is proposed. This method can suppress the ambiguous short-range clutter by adding multi-elevation angles constraint, which results in a deep and wide beam notch in the elevation angle corresponding to the short range. Moreover, the method offers potential robustness to the platform height errors. The residual range-independent clutter can be suppressed by two-dimensional azimuth-Doppler STAP methods. The simulation results show the validity of the method.6. In chapter 7, the clutter suppression for bistatic airborne radar is studied. In factual environments, the vertical heights of the transmitter aircraft and the receiver aircraft are always different, as a result, the vertical baseline appears between the two platforms. This paper analyses the influence of the vertical baseline on bistatic airborne radar. Due to the clutter has range dependence for bistatic airborne radar, and which causes conventional space time adaptive processing (STAP) methods to get low GMTD performance. To deal with this problem, two new compensating methods based on derivative based updating (DBU) are proposed: using received distance in the direction of the main lobe of beam pattern; using receive elevation angle cosine in the direction of the main lobe of the beam pattern. The results of the bistatic radar data based on computer simulation show the two methods can mitigate the clutter range dependence and improve the clutter suppression performance.