Ship target detection using radar is an important application in military and civilian fields. For the polarization estimation of scattering waves in the underdetermined case, i.e., the number of scattering waves from ships is larger than the number of sensors, this paper proposes two estimation methods with different measurement models. 1) For the single-vector-sensor model, this paper proposes the polarization-invariant ESPRIT-based method. This method can estimate the polarization of signals containing target echo, interference, and noise, which can cure the problem that the accuracy of existing method is poor under low interference signal ratio. 2) For the multi-vector-sensor model, this paper proposes an improved ESPRIT method based on the spatial-invariant and time-invariant simultaneously, which can increase the degree of freedom without increasing hardware cost. As for another problem of multi-vector-sensor, i.e., almost all existing methods assume that the number of scattering waves is known, this paper introduces the polarization spectrum for the first time, which can estimate the polarization parameters when the number of scattering waves is unknown. Finally, we analyze the two proposed ESPRIT-based methods comparing with some existing methods through Monte Carlo simulation, which results demonstrate the efficience of the proposed methods.

In order to improve radar's ability to detect and track targets under countermeasures, it is of great significance to analyze anti-interference. The traditional polarization anti-interference has a poor effect on the variable polarization interference with the same frequency and main lobe as that of radar, since it needs to utilize the interference signals without target for asynchronous estimation. Aiming at this problem, this paper puts forward a Synchronous Method (SM) to suppress the variable polarization interference for the first time. Specifically, for the horizontal/vertical dual-receiving polarization radar, we first analyze its polarization invariant in the virtual variable polarization reception, and propose the Golden Section Method to estimate the interference polarization with a fewer calculation complexity. Then, the Cosine Symmetry Method is proposed by the variable polarization observation of coherent accumulation, which replaces the Least Square Method to estimate the target scattering matrix. Subsequently, the Generalized Squint Projection Transformation is proposed to realize the coherent accumulation of target signals under the variable polarization interference, which can effectively resolve the problem of amplitude loss and noise influence in coherent accumulation. Finally,baesd on these three steps, SM is constructed and its effectiveness is verified by simulation, which shows a potential engineering application value.

Compared with the multisite radar, the monostatic radar has the disadvantages of a single viewing angle and less available information. In the presence of main-lobe jamming, the monostatic radar has the problems of poor anti-jamming effect and difficulty in achieving target detection. To solve the above problems, the radar signal processing method and data fusion method are combined in this paper. A main-lobe jamming suppression method for phased array (PA) netted radar based on the maximum signal-to-noise ratio blind source separation (MSNR-BSS) is proposed. Firstly, the received signals of a monostatic PA radar are separated into target signals and jamming signals by the MSNR-BSS method. Secondly, channel accumulation, pulse compression, and constant false-alarm rate (CFAR) target detection are performed on the separated signal channels to obtain the target radial distance. Finally, the data fusion theory of three global orientations is employed to determine the target spatial position. Simulation experiments evaluate the performance of main-lobe jamming suppression and verify the effectiveness of the proposed method. In the presence of noise amplitude modulation (NAM) jamming or smeared spectrum (SMSP) jamming (JNR=100dB), the target positioning accuracy of the proposed method is 6.613m or 6.596m. And the peak sidelobe level ratios (PSLR) in the two jamming environments are approximately equal to - 8dB.

Due to that the polarization scattering matrixes of single reflection, double reflection and single diffraction at different angles are ignored, the polarization scattering matrix of trihedral corner reflector is usually assumed to be [1 0; 0 1], which resulting in inaccuracy by utilizing its polarization to anti-jamming. Therefore, this paper comprehensively analyzes the polarization scattering matrix of trihedral corner reflector based on high-frequency approximation. Firstly, we seperately analyze various reflection components based on the PO-SBR model, so as to cure the deficiency that the classification discussion methods are susceptible to changes in target size and shape. Then, the polarization scattering matrixes of single reflection, double reflection, triple reflection and single diffraction are rigorously deduced, and demonstrated by combining amplitude/phase information. In this process, we successfully integrate the six kinds of double reflection by the transforming of polarization base. Subsequently, the polarization scattering matrixes in two-dimensional space are extended into three-dimensional space, which are more suitable for the polarization characteristics of trihedral corner reflector. Finally, we compare the model in this paper with the MoM of Feko software, and obviously accelerate the model in this paper through CUDA. The conclusion of this paper is significant for the polarization analysis of trihedral corner reflector to anti-jamming.

Data, in deep learning (DL), are crucial to detect ships in synthetic aperture radar (SAR) images. However, SAR image annotation limitations hinder DL-based SAR ship detection. A novel data-selection method and teacher–student model are proposed in this paper to effectively leverage sparse labeled data and improve SAR ship detection performance, based on the semi-supervised oriented object-detection (SOOD) framework. More specifically, we firstly propose a SAR data-scoring method based on fuzzy comprehensive evaluation (FCE), and discuss the relationship between the score distribution of labeled data and detection performance. A refined data selector (RDS) is then designed to adaptively obtain reasonable data for model training without any labeling information. Lastly, a Gaussian Wasserstein distance (GWD) and an orientation-angle deviation weighting (ODW) loss are introduced to mitigate the impact of strong scattering points on bounding box regression and dynamically adjusting the consistency of pseudo-label prediction pairs during the model training process, respectively. The experiments results on four open datasets have demonstrated that our proposed method can achieve better SAR ship detection performances on low-proportion labeled datasets, compared to some existing methods. Therefore, our proposed method can effectively and efficiently reduce the burden of SAR ship data labeling and improve detection capacities as much as possible.

The modeling and analysis of sea clutter are of great significance in radar target detection studies in marine environments. Sea clutter typically exhibits non-Gaussian characteristics and spatiotemporal correlations, posing challenges for modeling, especially when generating simulation data of continuous correlated non-Gaussian random processes. This paper proposes a novel method for sea clutter modeling. First, feature description functions are constructed to individually characterize the amplitude, temporal, and spatial correlations of sea clutter, allowing for an accurate depiction of its characteristics with fewer parameters. Subsequently, simulation data are generated based on these feature description functions, satisfying the amplitude distribution, temporal correlation, and spatial correlation characteristics of sea clutter. Additionally, complex signal forms are introduced in the underlying signal processing to generate texture and speckle components of sea clutter, enhancing the alignment of simulation data with actual data. Through comparison with measured sea clutter data, the proposed method has been shown to accurately simulate complex sea clutter with real-world characteristics.