In recent years, as a crucial and fundamental task in applications such as autonomous driving, mobile robotics, and virtual reality, 3D object detection has received extensive attention from researchers in various fields. It aims to localize and classify objects of interest in 3D space and give the corresponding 3D bounding boxes, including the position, size, and orientation of objects, which provides the basic information for the subsequent understanding and perception of the 3D scene as well as planning and decision-making. Point clouds captured by LiDAR have become the most commonly used input data for 3D object detection due to their accurate 3D information and depth information. In this paper, the 3D object detection methods based on LiDAR point cloud with deep learning are reviewed, the characteristics and processing methods of point cloud are summarized, and several corresponding types of detection methods and multimodal fusion methods of point cloud and image are introduced. At the same time, this paper compares the performance of different methods and discusses the challenges and development trends of 3D object detection based on point cloud in the future.

In order to solve the networking communication problem of the main-subsidiary satellite cluster under discrete and sparse topology and to ensure the demand of real-time interactive remote control telemetry and task information between satellites,this paper proposes a set of communication system architecture and inter-satellite networking protocol stack design scheme according to the hierarchical and clustered spatial topology characteristics. Firstly, in the design of the network layer, the data structure of the network layer, the subnet cluster head selection backup mechanism based on the multi-weight priority cost function and the process of the sub-satellite network access are given respectively. Secondly, in the data link layer, the data is processed hierarchically according to the subcontracted remote control and subcontracted telemetry system. Then, in the design of the physical layer, the spread frequency hopping is used as the communication system of the inter-satellite link, and the time division multiple access method is used to realize the multi-user communication of the satellite cluster, and the specific communication index of the inter-satellite link is given. Finally, the link calculation of inter-satellite communication is carried out. The calculation results show that the physical layer design meets the communication requirements of each satellite.

Satellite resources on board are limited and precious, and ground terminals in real communication scenarios have different geographic distribution and satellite communication service demands, so dynamic on board satellite resource management is needed to design flexible and efficient resource management schemes. In this paper, the use of hopping beam technology can provide flexible resource management at the beam level to solve the problem of unbalanced service demand between ground cells covered by point beams generated by high-throughput satellites. Firstly, in order to solve the problem of co-channel interference in beamhopping satellites, an interference avoidance strategy between beam-hopping clusters based on frequency mode switching is proposed. Further, two cluster configuration strategies, uniform and non-uniform clustering, are proposed. In order to solve the problem of poor equalization effect of uniform clustering and unsuitable for the dynamic change of the ground, combined with the problem of co-channel interference, a non-uniform cluster configuration strategy based on frequency mode switching is proposed for the interference avoidance strategy between beam-hopping clusters. Finally, the interference avoidance strategy is verified and simulated. Besides,the different clustering strategies are verified based on various intelligent optimization algorithms. The simulation results verify the feasibility of the clustered configuration and show that the non-uniform clustering algorithm with no distance limitation has the strongest flow-balancing ability.

X-ray communication technology is a kind of space communication mode using X-ray as a carrier, which has the advantages of a large communication bandwidth, light weight, small volume, low power consumption, and high confidentiality. By designing a multi-target X-ray signal modulation device, the X-ray communication rate can be improved, and the accuracy of X-ray energy spectrum recognition at the signal receiving end can be ensured, so that the advantages of X-ray communication based on energy load can be truly played. In this paper, an X-ray communication demodulation method based on peaking multi-level support vector machine is proposed to accurately identify the X-ray characteristic energy spectrum of multi-target materials. A peaking multilevel support vector machine classifier suitable for four-element communication is designed. The parameter tuning and verification ensure high accuracy and generalization ability. The simulation results show that support vector machine provides an efficient, accurate, and robust signal recognition solution for X-ray communication based on energy load.

To ensure the security of authentication and key distribution in the space information network, this paper proposes a public-key management scheme based on the elliptic curve cryptosystem. With the capacity for network access authentication and key agreement/update, this scheme can achieve various security goals, such as authenticity, anti-replay, and forward/backward security. For interstellar authentication and key agreement, our scheme only requires 3 interactions between the satellites. We provide a security and efficiency analysis for our scheme, and show that the proposed scheme not only satisfies the actual availability in the space information network, but also performs better than other existing public-key management schemes.

Compared to 32QAM technology, 32APSK technology reduces the number of amplitudes and is suitable for a nonlinear channel in a relay satellite communication system. TCM technology combines channel coding and multi-level modulation,doesn't increase the spectrum bandwidth, decreases the transmission power, cuts down energy consumption, lowers the requirements for technical indexes of the power amplifier, and is beneficial for achieving lightweight and miniaturization of satellite payloads. This paper combines 32APSK technology and TCM technology, proposes a kind of 32APSK-TCM technology, discusses details of the 32APSK-TCM technology constellation subset splitting method and constellation point selection method, and analyzes the performance of the 32APSK-TCM technology. This paper simulates the 32APSK-TCM technology and 16APSK technology using the simulation platform of the relay satellite communication system developed by our research team. Simulation results demonstrate that under the condition of an ideal channel, I/Q amplitude phase imbalance, amplitude frequency characteristics, group delay, phase noise, power amplifier saturation point, and nonlinear channel, if the maximization value of the required bit error rate is 1E-6, compared to 16APSK technology, the minimization value of the signal-to-noise ratio of 32APSK-TCM technology saves 13.29 dB, 13.29 dB, 14.84 dB, 15.54 dB, 15.11 dB, 15.77 dB, and 16.37 dB, respectively.

This paper introduces an application of QC-LDPC codes to telemetry system. At the same time, a joint design method of code length and telemetry frame length is proposed. Spectrum resources, channel qualities, and hardware capabilities are comprehensively considered in this design method, so that the optimal scheme can be designed. By using the RU encoding algorithm and the LLR BP decoding algorithm, a hardware implementation method based on FPGA and AD936X architecture is proposed. The coding gain of the designed system is simulated according to the technical route. Combining the theoretical analysis and simulation results, it can be seen that of the advantages of low latency and high gain, the proposed telemetry system design method has certain application value in future telemetry systems. In summary, the proposed method solves the problem of high coding delay in traditional telemetry system, and brings about 8 dB coding gain at the cost of minimal delay.

A CBAM-GRU classification model based on the combination of Convolutional Attention Mechanism Module(CBAM) and Gated Recurrent Unit (GRU) network is investigated for automatic modulation identification in non-cooperative communication systems. The pre-processed time-domain amplitude, phase and I/Q values of the signal are combined and converted into a matrix of input sample values, which are entered into the network for signal classification and identification. Simulations are conducted using the RadioML2016. 10a radio dataset, and the CBAM-GRU model are compared with the Convolutional Neural Network (CNN), Long Short-Term Memory network (LSTM), GRU, and Convolutional Long Deep Neural Network (CLDNN). The results indicates that the classification accuracy of the CBAM- GRU model reaches 92.79%, showing improvements of 8.52%, 1.84%, 1.75%, and 8. 61% over the comparison models respectively. Compared to traditional CNN or LSTM models, the CBAM-GRU model is more effective in capturing spatio-temporal features of sig- nals, thereby enhancing recognition accuracy.

In order to find a carrier-to-noise ratio estimation method with low computation and strong practicality, this paper draws on the experience of the classical variance summing method and proposes an improved algorithm. The distribution of the absolute value of the I-branch accumulative value is deduced when stably tracking GPS signals, and an expression for estimating the carrier-to-noise ratio is given based on its statistical characteristics. The Q-branch accumulative value is not necessary in this algorithm. From the aspects of estimation accuracy and computation, the improved algorithm is verified by simulation and actual measurement data analysis. The results show that the overall performance of the improved algorithm is equivalent, with an error of less than 0.4 dBHz, the stability is better than that of the classical variance summing method and narrow to wide power ratio method, and the computation is reduced by 50% compared to that the classical algorithm. The method has certain practicality in some receivers with limited hardware resources.

The Phase Gradient Autofocus (PGA) algorithm is widely used to compensate for phase errors in Synthetic Aperture Radar (SAR) images. In the processing flow of the PGA algorithm, the two-step operation of selecting points and adding windows has a significant impact on algorithm performance. Traditional PGA algorithms often suffer from poor point selection quality or incorrect window width estimation, leading to poor focusing effect and slower convergence speed. This article proposes a strong point selection method based on the maximum energy signal-to-noise ratio criterion and an adaptive window width estimation method. Using the dimensions of energy and signal-to-noise ratio, ideal isolated strong scattering points are selected from image data, and two traditional windowing methods are combined and improved to adaptively estimate the window width, achieving improved algorithm stability and convergence speed. The simulation and experimental data processing results confirm the effectiveness of the algorithm proposed in this paper.

Digital channelization technology is often used for broadband electromagnetic signal reception. When its analysis filter banks and comprehensive filter banks have precise reconstruction characteristics,accurate reconstruction of the received signal can be achieved. For the electromagnetic spectrum recognition problem in the field of complex electromagnetic environment perception,it is necessary to accurately restore the electromagnetic signal received by the receiver and perform spectrum recognition based on the accurately restored signal. The article proposes an intelligent electromagnetic spectrum recognition technology based on precise reconstruction of digital channelization. Firstly, a digital channelized receiver structure that can achieve precise signal reconstruction is constructed. Then,wavelet analysis is used to construct a time-frequency waterfall diagram of the signal,and artificial intelligence processing is performed based on this graph to achieve electromagnetic spectrum recognition. Finally,simulation results are provided. The simulation results in the article demonstrate the correctness and effectiveness of the method.

Image fusion model based on autoencoder network gets more attention because it does not need to design fusion rules manually. However, most autoencoder-based fusion networks use two-stream CNNs with the same structure as the encoder,which are unable to extract global features due to the local receptive field of convolutional operations and lack the ability to extract unique features from infrared and visible images. A novel autoencoder-based image fusion network which consist of encoder module, fusion module and decoder module is constructed in this paper. In the encoder module, the CNN and Transformer are combined to capture the local and global feature of the source images simultaneously. In addition, novel contrast and gradient enhancement feature extraction blocks are designed respectively for infrared and visible images to maintain the information specific to each source images. The feature images obtained by encoder module are concatenated by the fusion module and input to the decoder module to obtain the fused image. Experimental results on three datasets show that the proposed network can better preserve both the clear target and detailed information of infrared and visible images respectively, and outperforms some state-of-the-art methods in both subjective and objective evaluation. Meanwhile, the fused image obtained by the proposed network can acquire the highest mean average precision in the target detection which proves that image fusion is beneficial for downstream tasks.

There are more and more applications of change detection methods based on deep learning in high-resolution remote sensing images. However, downsampling and cropping strategies deployed to fit the GPU (Graphic Processing Unit) memory constraints on processing large-size remote sensing images often result in incomplete semantic information and loss of fine details. In this paper, a collaborative supervised network based on feature pyramids is proposed to enable the network to learn local and overall features from cropped and downsampled image blocks. In addition, a feature-sharing mechanism is introduced to fuse global features and local features. We evaluated the network on the LEVIR-CD (a remote sensing change detection dataset) and S2Looking (a building change detection dataset) by comparing it with some representative change detection networks. The comparison experiments show that the proposed network performs better in multiscale change detection, with a 2.69% improvement in precision on LEVIR-CD, and 6.83% and 2.68% improvement in precision and recall on the S2Looking dataset, respectively.