Search

To address the issues of ineffective task offloading decisions caused by multitasking and resource constraints in vehicular networks, the Quasi-Newton method deep reinforcement learning dual-phase online offloading (QNRLO) algorithm was proposed. The algorithm was designed by initially incorporating batch normalization techniques to optimize the training process of deep neural networks. Subsequently, optimization was performed using the Quasi-Newton method to effectively approximate the optimal solution. Through this dual-stage optimization, performance was significantly enhanced under conditions of multitasking and dynamic wireless channels, improving computational efficiency. By introducing Lagrange multipliers and a reconstructed dual function, the non-convex optimization problem was transformed into a convex optimization problem of the dual function, ensuring the global optimality of the algorithm. Additionally, system transmission time allocation in the vehicular network model was considered, enhancing the practicality of the algorithm. Compared to existing algorithms, the proposed algorithm improves the convergence and stability of task offloading significantly, addresses task offloading issues in vehicular networks effectively, and offers high practicality and reliability.

There are many ways to generate adversarial samples for synthetic aperture radar (SAR) images at present, but some problems such as large amount of perturbation of adversarial samples, unstable training, and unguaranteed quality of adversarial samples still exist. To solve the above problems, a SAR image adversarial sample generation model was proposed. The model was based on the AdvGAN model architecture. Firstly, according to the characteristics of the SAR images, an adaptive threshold segmentation method based on the enhanced Lee filter OTSU was designed. The mask extraction module composed of equal modules, this method produced a smaller amount of disturbance, and the structural similarity (SSIM) with the original sample reached that more than 0.997. Secondly, the improved relativistic average GAN (RaGAN) loss was introduced into AdvGAN, and the relative mean discriminator was used to make the discriminator rely on both real data and generated data during training, which improved the stability of training and the attack effect. Experiments were compared with related methods on the MSTAR dataset. Experiments show that the attack success rate of SAR image adversarial samples generated by this method is increased by 10%～15% than that of traditional methods when attacking defense models.

In order to find an effective method to improve the corrosion resistance properties of the candidate structural materials for lead-alloy cooled fast reactors, the 316L steel was treated by high density single elec-tropulsing. And then the interface corrosion characteristics in stagnant lead-bismuth eutectic (LBE) at 823 K for 1000 h were examined. The results show that electropulsing treatment (EPT) has a significant influence on microcosmic appearance and corrosion depth of the corroded specimens. All the specimens were subjected to dissolution corrosion, but EPT with a current density of 1828 A mm-2 can improve obviously the corrosion resistance and reduce the extent of local corrosion of 316L in LBE. In addition, the corrosion depth of the 1828 A mm-2 EPT specimen decreases by 38%, compared with that of the non-EPT specimens.

With the advancement of technology, signal modulation types are becoming increasingly diverse and complex. The phenomenon of signal time–frequency overlap during transmission poses significant challenges for the classification and recognition of mixed signals, including poor recognition capabilities and low generality. This paper presents a recognition model for the fine-grained analysis of mixed signal characteristics, proposing a Geometry Coordinate Attention mechanism and introducing a low-rank bilinear pooling module to more effectively extract signal features for classification. The model employs a residual neural network as its backbone architecture and utilizes the Geometry Coordinate Attention mechanism for time–frequency weighted analysis based on information geometry theory. This analysis targets multiple-scale features within the architecture, producing time–frequency weighted features of the signal. These weighted features are further analyzed through a low-rank bilinear pooling module, combined with the backbone features, to achieve fine-grained feature fusion. This results in a fused feature vector for mixed signal classification. Experiments were conducted on a simulated dataset comprising 39,600 mixed-signal time–frequency plots. The model was benchmarked against a baseline using a residual neural network. The experimental outcomes demonstrated an improvement of 9% in the exact match ratio and 5% in the Hamming score. These results indicate that the proposed model significantly enhances the recognition capability and generalizability of mixed signal classification. [ABSTRACT FROM AUTHOR]