Offshore ship object detection is a very challenging task and has received widespread attention from scholars and experts. Detectors based on Convolutional Neural Networks (CNN) and attention mechanisms have made significant progress in offshore ship object detection. However, the problem of false detection in the detection process is caused by the apparent similarity and background interference of ship targets. In order to solve this problem, this paper proposes a detection head module for fine-grained appearance discrimination implemented with Faster RCNN. This module includes a category fine-grained branch and an efficient full-dimensional dynamic convolution localization branch. The category fine-grained branch mines and utilizes category fine-grained identification features through global feature modeling and flexible perception range. The efficient omni-dimensional dynamic convolution positioning branch distinguishes objects and backgrounds through the efficient and flexible perception of ship boundary information, thereby reducing false and missed detections. Through experimental verification on the offshore ship public dataset Seaships7000, the proposed algorithm reduces false detections and missed detections and improves detector performance.

Hyperspectral target detection is crucial in Earth observation for both military and civilian applications. However,complex backgrounds and the scarcity of target samples pose challenges in hyperspectral image analysis. In this paper, we first employ the CEM coarse detection method to extract background data. Subsequently, a novel knowledge distillation model, namely KDTGAN (implemented through Transformer-GAN), is introduced. The generator of this teacher model adopts the structure of a Transformer encoder and combines it with a multi-scale data fusion approach to accurately learn the background distribution, which in turn enables target detection by reconstructing the background information. To overcome the challenge of unstable GAN training,especially the scarcity of pure background data, we propose a new loss algorithm to reduce the negative impact of suspicious target samples on model performance. To reduce the computational burden of the model, we introduce knowledge distillation and design a new distillation loss to constrain the student model to lighten the model while improving the student model's detection accuracy. The experimental results show that KDTGAN performs better than current detection methods with higher detection accuracy and robustness.

With the rapid development of remote sensing technology, the intelligent decoding of weak targets in optical remote sensing images has become one of the research hotspots in remote sensing information processing. The feature targets of remote sensing images are often characterized by small scale, many types, a large number, fast moving speed of some key small targets, and are easily affected by the complex background environment and noise, which makes it a great challenge extract information from weak targets in remote sensing images. Early research on weak target segmentation, detection, and tracking algorithms in intelligent interpretation algorithms mostly relied on template matching and a priori knowledge, and such algorithms need to consume a lot of resources, arithmetic, and expert knowledge costs, and there were problems of large computational volume and poor generalization ability. In recent years, with the rapid development of deep learning and other artificial intelligence technologies, the information of weak targets can be accurately obtained in massive remote sensing data, and the features of weak targets can be quickly extracted by combining deep learning algorithms to provide efficient and accurate decoding information. This paper summarizes the research progress of intelligent interpretation algorithms for weak targets in remote sensing images, including weak target segmentation, detection, and tracking algorithms based on traditional image processing methods, as well as typical related algorithms based on deep learning. By analyzing the advantages and limitations of these methods, it is of great significance to improve the information acquisition ability of relevant targets, enhance the situational awareness level of observation, and future applications.

As a difficulty and focus in the field of radar imaging, radar forward-looking imaging has broad application prospects in automatic driving, navigation, precision guidance and so on. The traditional forward-looking imaging algorithm is limited by the width of the antenna aperture and cannot achieve high-resolution imaging. In this paper, CNN ( Convolutional Neural Networks )neural network and LSTM ( Long Short-Term Memory ) neural network are combined to realize the prediction of azimuth in forward-looking imaging. Firstly, the convolution-like model of the scanning forward-looking imaging signal and its ill-posedness are introduced. The echo signal is preprocessed by pulse compression and range migration correction, and input into the CNN-LSTM neural network to perform azimuth estimation by range unit. The simulation results show that the algorithm can effectively improve the azimuth resolution of forward-looking imaging and realize the super-resolution of forward-looking imaging.

Space-air-ground access integrating, integrated sensing and communication (ISAC) will be the key capabilities in 6G. This paper studies an SC-FDE ISAC signal framing method in UAV networking to meet the requirement of communication-aided sensing based on the communication signals. Additionally, the high-precision ranging methods for UAV clusters with 4 and 2 samples per symbol are proposed to improve the cooperative localization performance in the UAV cluster network. After extensive simulations, we found that these two ranging methods could achieve a performance of RMSE=0.1 m at SNR=10 dB. Furthermore, the 2 samples per symbol ranging method outperforms the 4 samples method, and gets closer to the CRB theoretical performance bound.

Trellis coding modulation technology can improve the signal-to-noise ratio, reduce the transmission power, and lower the bit error rate without changing the bandwidth of the information transmission spectrum. The realization method is to divide the constellation map to form a subset and gradually increase the minimum Euclidean distance between the signal points in the constellation map. In this paper, the demodulation performance of 8PSK and 16QAM-TCM modulation techniques is simulated on the simulation platform of a relay satellite communication system developed by the project team. The simulation curve shows the relationship between the required bit error rate and signal-to-noise ratio. The simulation results demonstrate: ① under the channel conditions of an ideal channel, I/Q amplitude and phase imbalance, amplitude and frequency characteristics, group delay, phase noise, power amplifier saturation, if the required BER is 1E-7, compared to 8PSK modulation technology, the signal-to-noise ratio of 16QAM-TCM technology saves 8.85 dB, 9.04 dB, 8.45 dB, 10.2 dB, 8.5 dB, and 14.6 dB, respectively; ② under the channel conditions of non-linear, correspondingly along with increasing of the signal-to-noise ratio, the BER of 8PSK modulation technology fluctuates around the order of 1E-3; ③ if the required BER is 1E-7, compared with the ideal channel SNR simulation result, the non-linear channel SNR simulation result of 16QAM-TCM technology loses 4.8 dB.

As the military electronic system demands more and more the real-time and certainty of the network， FC-AE-1553,as a command response network protocol based on fibre channel with strong real-time and high certainty, has been more and more widely used in the fields of data transmission and flight control in avionics environment. This paper introduces the basic characteristics of FC-AE-1553 bus, then takes the launch vehicle telemetry system as the background, constructs a simple star network topology, and uses FC-AE-1553 bus protocol as the communication carrier to verify the applicability of FC-AE-1553 bus technology in the launch vehicle telemetry system.

Due to the inconsistency in symbol rate and count between the hopped signals, especially with few symbols in low-speed hops, it is difficult to extract clock errors. To address the clock synchronization challenge of frequency hopping signals under high dynamics, a method based on frequency offset estimation for clock offset feedback adjustment is proposed. By estimating the frequency and clock offset through synchronization sequences and adjusting the clock offset and clock tracking through feedback methods, high-precision clock synchronization is achieved. Simulation results show that this method is suitable for ultra-high dyna-mics with flight speeds up to 7.9 km/s and accelerations of 0.2 km/s², exhibiting superior timing synchronization performance. The timing accuracy meets the requirements for high-speed frequency hopping signal demodulation, with demodulation loss less than 0.1 dB.

In the signal demodulation process of the satellite coherent optical communication system, the relative motion of the satellite will cause the received signal light to produce a Doppler frequency shift of GHz magnitude. The long-distance transmission of the signal light leads to an extremely low optical signal-to-noise ratio. The traditional method cannot compensate for the large-scale Doppler frequency shift at a low signal-to-noise ratio, which seriously affects the ability of the communication system. In view of the above problems, this paper proposes a two-stage frequency acquisition algorithm, which includes two stages: automatic frequency sweeping and frequency locking control. In the automatic frequency scanning stage, the frequency difference is reduced to the order of 100 MHz by the local oscillator's automatic frequency scanning. In the frequency locking control stage, the frequency difference is further reduced to the MHz level at a low signal-to-noise ratio through high-precision local oscillator frequency control and FFT transformation. The simulation results show that the algorithm can compensate for the Doppler frequency difference in the 10 GHz range of the large dynamic range under the 2 dB signal-to-noise ratio, and meet the demodulation requirements of the satellite cohe-rent optical communication.

This paper focuses on the spoofing and anti-spoofing technology and protection of GNSS receiver. Firstly, the signal generation and spoofing implementation strategies are introduced, and several aspects such as spoofing threat implementation difficulty and protection difficulty are compared and analyzed. Secondly, an analysis of anti-spoofing technologies is discussed from three aspects: signal system, receiver signal processing technology and auxiliary information verification technology. A comparative analysis of anti-spoofing technologies is provided from the aspects of spoofing adaptability, implementation requirements, implementation difficulty, and effectiveness. Based on these, the anti-spoofing framework of GNSS receiver is proposed to deal with different spoofing attacks, which has significant reference for evaluating the spoofing procetion ability of GNSS receivers

Common-view time transfer involves exchanging the common-view data of two locations to complete long-distance and high-precision time comparison. This paper proposes a method for compressing and transmitting common-view data based on Beidou short message for areas without ground communication network. The method compresses the common-view data and uses the short message function of Beidou to transmit the compressed data in real time. A comparison test has been conducted in an area without a ground communication network. The test results demonstrate that the common-view data compression method designed in this paper has a significant compression effect and can achieve real-time common-view time transfer without a ground network through Beidou short message.

The development of stealth technology and high-speed aircraft has brought serious challenges to radar target detection, and the long time accumulation technology is an effective method to improve the signal-to-noise ratio (SNR) in the field of signal processing. For high speed and high maneuvering targets, signal accumulation over a long period of time will cause the problem of moving across the range unit and moving across the Doppler unit. The "two-migration" problem will seriously affect the accumulation gain, and the traditional MTD (Moving Target Detection) method has been unable to accumulate the energy of the signal. In order to solve the "two-migration" problem and improve radar detection performance of high-speed targets, many methods have emerged in recent years. Based on the target motion model, this paper will analyze long-term accumulation into three categories:long-term accumulation under uniform radial motion, long-term accumulation under accelerated radial motion, and long-term accumulation under complex radial high-order motion. This paper summarizes and presents the "two-migration" problems and solutions under the corresponding model, aiming to provide reference for subsequent research.

Missile-borne bistatic forward-looking synthetic aperture radar (MBFL-SAR) can achieve full-range, forward-looking, two-dimensional high-resolution imaging during the diving phase, enabling collaborative combat between missiles and improving guidance accuracy. In response to the issue that the trajectory of the transmitter directly affects the resolution of bistatic imaging, the two-dimensional resolution characteristics of the diving phase in the MBFL-SAR are analyzed. Combined with the motion characteristics of the missile, under the constraints of two-dimensional resolution and missile kinematics, the three-dimensional acceleration variations of the missile are used as optimization variables, and an objective function related to the angle between the resolutions in the scene is constructed. An improved genetic algorithm is employed for the trajectory optimization of the transmitter. Compared with the existing trajectory design method based on a linear attenuation model, the trajectory optimization method proposed in this paper can better meet the requirements of high-resolution imaging.

Regarding the issue of detection equipment being within the range of spaceborne SAR sidelobe, which causes the signal to be lost in the strong noise background, a method based on data fusion between multi-platform receivers is proposed. Without knowing the signal form, the weak signals submerged in noise can be detected, and the signal can be classified and accurately estimated. Firstly, the reference receiver and other receivers are cross-correlated to obtain the peak information, and the delay position between the signals and the reference signal is obtained according to the position of the peak information, in order to perform delay calibration. Secondly, each receiver performs coarse step FrFT filtering, records peak information for precise estimation, and restores the original signal based on the peak angle and the inverse FrFT. Finally, it is determined whether there is a signal. If the signal is achieved, a new signal will be formed by the fusion of power ratio of multi-platform receivers' original signal. The rotation angle ranges are limited based on the peak information of multiple stations, and the precise step FrFT is used to estimate the chirp rate and central frequency. The joint cross-correlation spectrum analysis is used to realize the accumulation of signal energy, and the left and right boundaries in the signal persistence state are found by using the method of continuously minimizing the boundary valley. The bandwidth and central frequency are accurately estimated, and then calculate the pulse width. The simulation results show that this method can accurately estimate the parameters of the time-frequency domain of Chirp in the background of Gaussian white noise and colored noise with low noise.