Recently, with the development of deep learning, the field of lightweight object detection has witnessed significant progress. However, mainstream lightweight detectors ignore the extraction of multi-scale semantic information. In addition, these approaches ignore the relationship between deep semantic features and shallow detail features. To relieve above shortcomings, a Pyramid Pooling Enhanced Multi-scale Network(PPMENet) is proposed and an Efficient Pyramid Pooling Block (EPPB) is designed to extract multi-scale deep semantic information,strengthening the feature expression ability of the model. On the other hand, a Cross Semantic Level Interaction Attention Module (CSIAM) is designed to enhance information interaction between features at different semantic levels. Experimental results on the MS COCO 2017 test set show that PPMENet gets 28.0% average precision, only with 2.16×10⁶ model size and 0.97GFLOPs,and achieves inference speed of 218 frame/s. Compared with other methods, PPMENet realizes a good balance between detection accuracy and model execution efficiency.

For the optimal feature subset selection and model parameter optimization in ultra-wideband non-line-of-sight(NLOS) recognition，a new NLOS recognition method based on the cross-validation recursive feature elimination algorithm of Light Gradient Boosting Machine(LightGBM) and Optuna parameter tuning is proposed. First，six important features，including the difference between the first path signal and the total received signal power，and the maximum noise，are selected as the optimal feature subset using the recursive feature elimination and cross-validation algorithm. Then，Optuna is used to optimize the hyperparameters of LightGBM model. Line-of-sight and non-line-of-sight feature data is collected，and the Support Vector Machine，Extreme Gradient Boosting algorithm，and parameter-optimized LightGBM model are trained and tested. The results demonstrate that the selected features exhibit excellent discriminative ability，with the optimized LightGBM model achieving a recognition accuracy of 95.28% .

In order to reduce complexity of construction of indoor positioning fingerprint database and improve the positioning accuracy, an indoor fingerprint positioning algorithm based on matrix completion under the 5G ultra-dense network is proposed. In the offline database construction stage,the algorithm first uses the K-nearest Neighbor(KNN) interpolation method to complete the matrix of part of the fingerprint database to construct a complete database. Secondly,the sparse auto-encoder is used to extract the sparse features of the fingerprint database, and the high-dimensional received signal strength indication (RSSI) signal is reduced. In the online fingerprint matching stage,the weighted KNN algorithm is used to estimate the coordinates of the point to be located. After experimental simulation,the average relative error of the algorithm to reconstruct the fingerprint database is 0.31% . Compared with that of the traditional KNN fingerprint matching algorithm,the average error is reduced by 24.41% .

In order to further improve the accuracy of human action recognition and fully explore the spatiotemporal features of action sequences, a graph convolution action recognition method based on spatiotemporal feature fusion and attention mechanism is proposed. The spatial attention map convolution is used to refine the topology to capture the correlation features of the joints under different motion types,and the time convolution structure is extended by the time domain multi-scale convolution module to capture the multi-scale time features. A multi-level feature fusion module is constructed,which takes the initial feature and the convolution output feature of the time-domain multiscale graph as the module input,and uses a two-branch structure to obtain the global and local channel features respectively. On this basis,a limb attention mechanism is proposed to divide the human topological structure and calculate the attention weights in the channel dimension respectively to enhance the model's ability to pay attention to local action features. The experimental results show that the recognition accuracy is 93.0% and 96.9% in CS and CV evaluation mode of NTU RGB+D data set,and 89.8% and 91.1% in X-Sub and X-Set evaluation mode of NTU RGB+D 120 data set,respectively. The recognition accuracy is higher than that of ST-GCN,CTR-GCN and other models.

It's difficult to implement maintainability design and evaluation effectively in the early stages of development of space TT&C ground system. To solve this problem,a maintainability design and evaluation system is constructed based on virtual reality technology,a maintainability design and evaluation workflow is developed,and a comprehensive evaluation criteria for maintainability design is proposed. In the virtual maintenance scenario,virtual maintenance prototypes and virtual maintenance resources are used to verify the overall process,and immersive simulation is conducted to verify the accessibility,visibility,and human body comfort of the maintenance of components of a vehicle mounted space TT&C ground system. Thus the quantitative comprehensive evaluation of virtual maintainability design is achieved. The results show that this method reduces costs by 68% , shortens duration by 61% compared with traditional method. It may serve as a practical reference for visualizing,quantifying carrying out maintainability design and evaluation in the early stages of development of space TT&C ground system.

In satellite communication systems operating in dynamic interference environments,the quality of channels and the interference power vary. Limited spectrum resources and complex interference environments pose challenges for anti-interference communication decisions, particularly in terms of resource allocation and service demands. Specifically, the challenge lies in efficiently utilizing resources while avoiding interference frequencies and optimizing power. To address this issue,a deep reinforcement learning-based anti-interference algorithm with multiple reward functions is proposed. The algorithm models the interaction between the transmitter,receiver,and interferer as a Markov decision process. By optimizing the reward function associated with the costs of channel and power switching,it introduces mechanisms for both frequency and power switching,analyzes the interference characteristics in the spectrum of adjacent time slots, and integrates the interference signal features collected during the interaction with channel information to train an anti-interference strategy. This strategy enables joint anti-interference decision-making in both the frequency and power domains. Simulation results demonstrate that the algorithm effectively reduces the probability of interference,accelerates convergence,and optimizes the utilization of power resources.

To address the issue of synthetic aperture radar(SAR) images being susceptible to environmental noise,which leads to a reduction in the signal-to-noise ratio(SNR) at target locations,a method integrating singular value decomposition (SVD) and minimum entropy deconvolution (MED) to enhance the Back Projection(BP ) algorithm is proposed. Initially,the acquired echo signals undergo SVD, and a singular value matrix is obtained. By retaining only the first five singular values and reconstructing the echo signal matrix,an initial noise reduction is achieved. Subsequently,the signals are processed using MED filtering, where filter coefficients are iteratively updated to minimize the entropy of the signal,thereby reducing the kurtosis of the output signal and suppressing noise. A Zero-Phase Filter(ZPF) is then applied to restore any phase delays. Finally, the SAR image is generated using the BP algorithm. Experimental results demonstrate that this method significantly enhances the SNR at target locations,with improvements of 7.9 dB and 9.1 dB for large and small corner reflectors,respectively.

Vehicular edge computing(VEC) converges the computing resources of cloud servers to the edge of the network closer to the vehicle side, allowing vehicles to offload vehicular computing tasks to the network edge servers,thus providing vehicles with low latency,high bandwidth and high reliability services. However,the highly dynamic network topology,strict low-delay constraints,and massive data of vehicular tasks of VEC pose significant challenges for implementing efficient offloading. The digital twin(DT)-driven VEC model can enable real-time monitoring of the state of the VEC network,thus assisting in making sound offloading decisions in the physical world. Firstly, the research progress of edge computing, available vehicles and DT-driven VEC task offloading methods are introduced. Then,the scenario architecture of DT-driven task offloading for VEC is elaborated. Finally,the future research challenges and solutions of DT-driven VEC task offloading methods are discussed,in hope of providing certain guidance for solving the problem of DT-driven VEC task offloading.

In response to challenges in sea clutter modeling within the classical algorithms,including the lack of fitting accuracy due to the inability to satisfy multiple statistical characteristics simultaneously and the limitations in controllably generating accurate class-based results,combining the generative power of U-Net with the potential of complex-valued neural networks to deal with complex nonlinear problems in the electromagnetic domain, a novel approach is proposed. This approach integrates complex-valued network layers and a classifier-free guidance module, establishing an interpretable mapping mechanism for input conditions,resulting in complex-valued guided diffusion model(CVG-DM). This model is centered on the direct utilization of the complex-valued baseband signals from the In-phase and Quadrature(IQ) path of sea clutter, as well as the exploration of the relationship between sea clutter and strong targets in the background. This enables controlled generation of the model under varying conditions of target presence or absence, and assessment based on amplitude distribution, temporal and spatial correlation, nonlinear characteristics,and Doppler spectrum. Simulation experiment validates CVG-DM's capability in realizing sea clutter data augmentation under varying conditions. The simulated clutter can simultaneously take into account above five statistical properties, surpassing the completeness of real number network-based evaluation metrics and further enhancing fidelity.

For the problem that the existing deep learning modulation recognition algorithms are not robust enough and have insufficient generalization ability in complex signal environments，a multi-channel network based on phase parameter estimation and spatial reconstruction(PET-SAMCL) is proposed. First，the input in-phase quadature(IQ) signal is converted by phase parameter estimation and divided into three modules to extract the amplitude-phase feature，IQ combination and branching features of IQ respectively. A spatial reorganization unit(SRU) is added to the feature extraction module to reduce the influence of redundant features. The spatial features are refined and fused by global average pooling and soft attention operations，and the temporal and spatial features are extracted by gated recurrent units(GRU) and bidirectional gated recurrent units(BiGRU) . Ablation study determines the optimal model structure. The model performs well on the RML2016.10a dataset，achieving a maximum recognition accuracy of 93.9% at 14 dB，and the average recognition rate is increased by 7.7% compared with that of other models.