Aiming at the problem of how to obtain reverberation samples and estimate their covariance matrix in the space-time adaptive processing(STAP) of sonar system, a new space-time adaptive processing method is proposed based on sparse reconstruction of reverberation in this paper. Firstly, according to the space-time distribution characteristics of reverberation received by moving platform sonar, a space-time steering dictionary for sparse reconstruction of reverberation is designed along the relation curve between Doppler frequency shift and incident cone angle cosine of the reverberation unit. Then, a reverberation sample in the rangecell under test (RUT) is reconstructed with high precision by sparse decomposition of signals obtained from the sonar array in the space-time steering dictionary. Finally, based on the prior information of reverberation probability distribution model, a sufficient number of reverberation samples are generated to meet the requirement of performance loss index on reverberation sample size in the space-time adaptive processing, so as to correctly obtain estimation of the covariance matrix of reverberation. This method can reconstruct the reverberation samples and estimate the reverberation covariance matrix directly from the data in RUT without relying on the auxiliary data from units adjacent to the RUT. Therefore, it is not only suitable for the environment with constant reverberation statistical characteristics, but also suitable for the environment with varying statistical characteristics. Simulation results of sonar forward-looking array and side-looking array indicate that the improvement factor of the proposed method is about 10dB lower than the traditional space-time adaptive processing method. So this new STAP method has good anti-reverberation performance.

In the field of underwater acoustics, the signal-to-noise ratio (SNR) is generally low, and the underwater environment is complex and variable, making target azimuth estimation highly challenging. Traditional model-based subspace methods exhibit significant performance degradation when dealing with coherent sources, low SNR, and small snapshot data. To overcome these limitations, an improved model based on SubspaceNet, called PConv-GAM Residual SubspaceNet (PGR-SubspaceNet), is proposed. This model embeds the global attention mechanism (GAM) into residual blocks that fuse PConv convolution, making it possible to capture richer cross-channel and positional information. This enhancement helps the model learn signal features in complex underwater conditions. Simulation results demonstrate that the underwater target azimuth estimation method based on PGR-SubspaceNet exhibits lower root mean square periodic error (RMSPE) values when handling different numbers of narrowband coherent sources. Under low SNR and limited snapshot conditions, its RMSPE values are significantly better than those of traditional methods and SubspaceNet-based enhanced subspace methods. PGR-SubspaceNet extracts more features, further improving the accuracy of direction-of-arrival estimation. Preliminary experiments in a pool validate the effectiveness and feasibility of the underwater target azimuth estimation method based on PGR-SubspaceNet.

Biases of initial direction estimation and focusing frequency selection affect the final focusing effect and may even cause algorithm failure in determining the focusing matrix in the coherent signal–subspace method. An optimized sonar broadband focused beamforming algorithm is proposed to address these defects. Initially, the robust Capon beamforming algorithm was used to correct the focusing matrix, and the broadband signals were then focused on the optimal focusing frequency by the corrected focusing matrix such that the wideband beamforming was transformed into a narrowband problem. Finally, the focused narrowband signals were beamformed by the second-order cone programming algorithm. Computer simulation results and water pool experiments verified that the proposed algorithm provides a good performance.

In general, the azimuth estimation in array signal processing is derived under the assumption of uniform white noise, whose covariance matrix is a scaled identity matrix. However, in practice, the noise can be nonuniform with an arbitrary unknown diagonal covariance matrix. In this paper, the estimation of the noise covariance matrix is formulated into a solution to the semidefinite optimization problem which can obtain a more accurate sensor noise covariance matrix. In the proposed algorithm, the estimated nonuniform noise is subtracted from the sample covariance matrix. The simulation results show that the proposed algorithm can significantly improve the performance of the sparse spectrum fitting algorithm in nonuniform noise case, while the classic SpSF algorithm is used under uniform white noise assumption. The water pool experiments show that there are indeed significant differences in the noise covariance of the sensors.

Unmanned Aerial Vehicles (UAVs) have become the preferred tool for terrain exploration due to their aerial perspective and maneuverability. Communication security between parties is ensured through the Authentication and Key Agreement (AKA) protocol. Traditional terrain exploration data processing relies on large-scale data centers, which suffer from inconvenient setup and insufficient real-time capabilities. Existing AKA protocols are susceptible to physical attacks and machine learning attacks, and they also suffer from high overhead and single-point-of-failure issues. To address these challenges, we propose a three factors authentication and key agreement protocol for UAVs-assisted terrain exploration. Specifically, we design an elliptic curve cryptography-based AKA framework to avoid single-point-of-failure issues at ground control stations, ensuring the security of the entire AKA process. Using portable terminals as fog nodes to serve as data transit stations, addressing the issues of inconvenient data center setup and insufficient real-time performance. During the login phase, three-factor authentication combining passwords, biometric features, and smart cards ensures the legitimacy of user identities. The physical properties of physically unclonable function (PUF) are utilized to resist physical attacks. Use reconfigurable PUF to update responses and resist machine learning attacks. We use the ROR model, informal security analysis, and ProVerif tool to rigorously verify our scheme. The results indicate the protocol can resist various attacks. Finally, compared to existing protocols, U2D and U2U achieve computational efficiency improvements of 56.9% and 33.4%, respectively, while reducing communication overhead by 38.1% and 28.0%, and lowering energy consumption by 24.5% and 39.2%, simultaneously, they against physical attacks and machine learning attacks, significantly enhancing security.

Relative positioning is a key technology that needs to be addressed for unmanned aerial vehicles (UAVs) to achieve flight mission involving autonomous aerial refueling, cluster formation and cooperative control. To address the shortcomings of the least squares (LS)-based multi-source information fusion method, such as poor outlier-tolerance, the idea of outlier-tolerance is used to improve the LS method. A novel loss function is proposed by replacing the parabolic function with a piecewise function, and a multi-source information outlier-tolerant relative positioning method based on the novel loss function is established. The simulation results show that the established method has a good outlier tolerance ability, which can avoid the adverse effects of outliers and ensure the reliability of the calculation results without significantly affecting the accuracy of the relative positioning.