Space‐time adaptive processing (STAP) methods can provide good clutter suppression potential in airborne radar systems. However, the performance of these methods is limited by the training samples' support in practical applications. To address this issue, a deep learning framework for STAP is developed. First, the clutter space‐time data and their exact clutter covariance matrices (CCMs) are simultaneously modelled via simulation, in which various non‐ideal factors such as aircraft crabbing, array errors, and internal clutter motion with all possible levels in practice are all considered. Then, a multi‐layer two‐dimensional convolutional neural network (CNN) is developed. In this CNN, low‐resolution angle‐Doppler profiles estimated by a few training samples are used for the input and the high‐resolution counterpart obtained by the exact CCMs are utilized for the labels. Once trained, the CNN can be used to predict the high‐resolution angle‐Doppler profile using a few measured data in near real time. The high‐resolution clutter spectrum can be further calculated using the space‐time steering dictionary and the above obtained profile. Finally, the CCM of the measured data can be constructed and the space‐time weight vector can also be achieved. Compared with recently developed sparsity‐based STAP methods, the performance of the proposed method is better and the computational load of it is far fewer, and therefore more suitable for real‐world implementation. The simulation results have demonstrated the superiority of the proposed method in both clutter suppression performance and computation efficiency.

Because the maximum radiation direction of the end-fire array antenna points to the axis of the array and the small element spacing required to suppress the grating lobes, the mutual coupling effect between the arrays is more serious than that of the broadside array. The article first established the end-fire array antenna pattern model and clutter signal model with the effect of mutual coupling. The clutter signal model includes two cases, ideal mutual coupling effect and mutual coupling error. On this basis, we take a ten-element uniform linear array using CST Microwave Studio to validate our model. Then we analyze the clutter distribution characteristics of the end-fire array airborne radar and the effect of mismatch on the spatial steering vector. Finally, an adaptive mutual coupling compensation method based on covariance matrix reconstruction is proposed. In general, clutter suppression methods with end-fire array take no account of the mutual coupling error changing over time, but the proposed method first adaptively estimates the mutual coupling error based on the airborne radar echo data and uses the estimated normalized impedance matrix to compensate the spatial steering vector, then combines the system parameters to construct the clutter covariance matrix. The simulations indicate that this method accurately estimate the impedance matrix, and it eliminates the clutter power reduction and spatial steering vector mismatch caused by mutual coupling error. Moreover, it also solves the serious non-stationary clutter problem of the end-fire array airborne radar.

Colored-loading factor optimization is one of the key technologies in knowledge-aided space-time adaptive processing (KA-STAP) field. In this paper, a novel colored-loading factor optimization method based on pre-whitening performance evaluation in direct data domain (Direct Pre-Whitening, DPW) is proposed. As for the proposed method, firstly, the prior knowledge is used for channel pair processing to realize the effective estimation of the space-time covariance matrix tapers. Secondly, the possible target signal is eliminated by orthogonal projection, and the echo samples of different range cell to be detected are obtained by space-time subaperture smoothing. Finally, the optimization function is constructed based on the pre-whitening in direct data domain, and the optimal colored-loading factor is obtained by searching. On the one hand, DPW method improves the accuracy of prior information by using the estimated space-time covariance matrix tapers; on the other hand, the colored-loading factors of different range cells are obtained respectively, thus it effectively solves the problem that the priori covariance matrix may have differences in range cells. Computer simulation results verify the effectiveness of the proposed method.

Clutter suppression for non‐side‐looking airborne radar operating at high pulse repetition frequency is an intractable problem because the clutter statistical properties vary with the range, especially in the near range. In addition, the near‐range clutter is always coupled with the far‐range clutter. Elevation prefiltering is recognised as an effective method to suppress the range‐dependent near‐range clutter, but the elevation training samples required must be independent and identically distributed (IID) and not contaminated by the far‐range clutter. Here, an elevation–compensation–prefiltering space–time adaptive processing (STAP) algorithm is developed. In the proposed algorithm, the elevation training samples are selected from the elevation snapshots corresponding to the first pulse, which are not contaminated by the far‐range clutter. An elevation compensation method is presented to alleviate the range dependence, and then there are adequate IID training samples to estimate the elevation covariance matrix. The performance of the near‐range clutter suppression is improved. As for the residual range‐independent far‐range clutter, it can be suppressed by the cascaded azimuth–Doppler STAP. Numerical examples demonstrate the effectiveness of the proposed algorithm.

Due to the range ambiguity effect and the complex terrain environment, the remote weak target of interest for non-sidelooking airborne radar is usually superimposed with the nonstationary and heterogeneous short-range strong clutter, so it is difficult for the traditional space-time adaptive processing (STAP) methods to achieve effective moving target detection. Therefore, the detection of the remote weak moving target in the heterogeneous and nonstationary clutter environment is one of the difficulties encountered by airborne radar. Through the analysis of airborne radar clutter characteristics, we have found that the short-range nonstationary clutter and the long-range ambiguous clutter do not overlap in Doppler domain, so the homogeneous training samples can be effectively selected through the digital terrain database. On this basis, the paper establishes the refined grid dot level clutter signal model of airborne radar and proposes a 3D-STAP method based on the digital terrain database, namely the DTD-3D-STAP method. The method first accurately registers the airborne radar echo with the grid dots of the digital terrain database; next, the training samples are selected along the equal Doppler line based on prior knowledge; then the training samples can be compensated through power compensation; and finally, the short-range nonstationary clutter is suppressed through 3D-STAP technology. On the one hand, the proposed method ensures the homogeneity of the clutter spectrum and power of the training samples, and the sample can be detected through prior knowledge and power compensation. On the other hand, 3D-STAP technology is used to effectively suppress nonstationary clutter. Computer simulation and experimental results verify the effectiveness of the proposed method.

Bistatic airborne radar has excellent performance due to the separate transmitter and receiver, but its clutter distribution is usually non‐stationary, which causes the performance of the traditional Space time adaptive processing (STAP) method to fail. Herein, firstly a bistatic airborne radar clutter signal model established, followed by an analysis of its clutter distribution characteristics in detail, and finally a bistatic airborne radar space–time compensation STAP method based on adaptive segmentation is proposed. The method first uses the distribution characteristics of the peak spectral centre in the first pulse data of the airborne radar to adaptively segment the clutter data, including extremely non‐stationary, severely non‐stationary, and non‐stationary clutter regions. Then the clutter data are compensated based on the estimation value of the space–time peak spectrum centre. Finally, each range section is processed by using STAP, and the detected target Doppler frequency is adaptively moved. Simulation results verify the effectiveness of the proposed method.