In the orbital environment, spacecraft face challenges such as scarce fault samples, varying operating conditions, and a strong reliance on accurate models and labeled data in traditional diagnostic methods. This paper systematically reviews transfer learning techniques for spacecraft fault diagnosis, highlights their recent advancements, and outlines future research trends. Transfer learning strategies are categorized into four types: instance-based, feature-based, model-based, and domain-adaptive. The principles,advantages, limitations, and representative applications of each strategy are analyzed, along with key enabling techniques such as importance weighting, adaptive batch normalization, parameter fine-tuning, and adversarial training. The review shows that transfer learning effectively mitigates issues of data insufficiency and distribution shift by enabling knowledge transfer from source to target domains. In particular, multi-source domain adaptation and adversarial domain adaptation significantly improve cross-condition diagnostic performance by enhancing model generalization and robustness. It is concluded that transfer learning provides a promising framework for intelligent spacecraft fault diagnosis. Future research should focus on source-free domain adaptation, multi-modal data fusion, semi-supervised transfer learning, and model interpretability, aiming to support practical deployment in real-world aerospace missions.

This paper provides an overview of the current status and future trends in remote sensing satellite data transmission technology. The article begins by introducing the developmental history of remote sensing satellite data transmission. It then further elaborates on the technologies involved in remote sensing satellite data transmission, detailing key aspects such as data compression,data encryption, data framing, data encoding, data modulation, and data transmission. Subsequently the paper analyzes the challenges faced by remote sensing satellite data transmission, such as bandwidth limitations, signal processing, real-time requirements, and data handling. It also summarizes strategies to address these challenges, such as adopting laser-based data transmission, more complex modulation schemes, and multi-source fusion with computational transmission. Looking ahead, the paper envisions future trends, including inter-satellite laser relay transmission, on-board computational transmission, and the application of computational constellations for intelligent in-orbit fusion in remote sensing satellites. Research indicates that remote sensing satellite data transmission technology will evolve towards higher transmission rates, higher reliability, more convenient data processing, and more intelligent mission workflows.

Existing deep learning-based modulation recognition methods are difficult to adapt to HF channels with significant time-varying characteristics, limiting their application in modulation recognition for High Frequency (HF) signals. In addition, in non-cooperative HF communication scenarios, the training dataset is often difficult to cover all possible modulation types, so that open-set modulation recognition also has important practical significance. This paper proposes a multi-modal feature fusion-based open-set modulation recognition method by combining communication domain knowledge and open-set recognition techniques,which effectively reduces the impact of time-varying HF channels and unknown modulation types on recognition performance. The proposed method first utilizes communication domain knowledge to obtain multimodal features that are robust to channel variations,and then extracts discriminative deep feature representations through multimodal feature fusion and deep feature learning to effectively identify known and unknown modulation types. In addition, the method also generates dummy samples through manifold mixing strategy to assist network training, which can enhance the network's ability to identify unknown types. Experimental results indicate that the proposed method outperforms existing open-set modulation recognition methods. When the channel conditions of training and testing signals are the same, the proposed method improves by over 3% in open-set recognition performance. When the channel condition of testing signals is drastically changed, that is, the channel conditions of the training and testing signals are different,the proposed method improves by over 8% compared to existing method, which exhibits strong robustness to channel variations.

In the field of aerospace tracking, telemetry, and command (TT&C) communications, the capability for precise telemetry, telecommand, and data transmission is a critical technology ensuring reliable spacecraft operations. As the complexity of space missions continues to escalate and the TT&C communication environment becomes increasingly demanding, higher requirements are imposed on the reliability of communication links in TT&C systems. Polar codes, as a short-frame burst coding scheme with high reliability, low complexity, and superior coding gain, yet exhibit high sensitivity to carrier frequency offset errors. Addressing the dual high demands for reliability and synchronization accuracy in TT&C systems, this paper proposes a Polar code-aided frequency offset estimation (PCAFOE) algorithm. Compared with the traditional carrier synchronization algorithm, PCAFOE algorithm is demonstrated with higher estimation accuracy, which is able to effectively improve the carrier synchronization performance of TT&C communication systems, and provides effective technical support for next-generation aerospace TT&C systems.

A novel channel estimation algorithm based on energy ratio confidence with dynamic dual-threshold decision is proposed to address the issues of missed path detection and performance degradation in existing channel estimation algorithms for dual-antenna space-time block coding (STBC) communication systems within unmanned aerial vehicle (UAV) swarm networking scenarios. These issues arise due to fixed thresholds and insufficient noise handling within the channel response length. Significantly, the proposed algorithm does not require prior knowledge of the exact channel impulse response length. Simulation results demonstrate that the proposed algorithm achieves a performance gain of nearly 2 dB at a bit error rate (BER) on the order of 1E-5 compared to the conventional threshold-based Fourier Transform-Least Squares algorithm. Furthermore, the algorithm effectively suppresses the noise in the estimated channel impulse response, leading to significantly enhanced channel estimation performance for UAV swarm networking.

The problem of output voltage waveform oscillation in the 28 V radiation-resistant P-channel field effect transistor at a resistive load of 10.4 Ω was studied, and it was found that the main reason was the mismatch between the device's own parasitic parameters and the peripheral circuit parameters. To avoid this phenomenon, a circuit improvement solution was proposed, that is, the series resistance in the gate electrode reduces the V_GS change rate and destroys the conditions for oscillation. At the same time, the clamp diode is added to the gate source electrode to ensure that the gate source voltage V_GS is always lower than the rated value, so as to improve the safety of the circuit operation and long-term working reliability.

The sum and difference channels of S-band single-channel measurement and control equipment exhibit phase differences, which may vary with environmental changes, transmission media, and assembly processes. Phase difference correction must be completed before the signal enters the baseband terminal for demodulation, to ensure the correct angular error signal is demodulated and stable target tracking is ultimately achieved. In practical work, the phase difference correction is basically carried out by setting up beacon machines at high positions, which can effectively solve the problem of inconsistent phase difference. However, with the expansion of the distribution area of measurement and control equipment and the shortening of state preparation time, traditional phase correction methods are no longer applicable. This article focuses on the research of S-band single-channel measurement and control equipment within a certain frequency working range. By pre-testing its phase difference and adopting methods such as cable compensation or adding phase shifters, the impact of phase difference on angular error is reduced within a certain error range at the design stage. In the later stage, during the operation of the equipment, the complex workflow of phase correction is eliminated. This article presents the principle and method of phase free calibration for S-band single channel measurement and control equipment,and verifies the effectiveness of the proposed method through experiments.

The Media Access Control (MAC) protocol is a critical component of wireless communication systems. The Statistical Priority-based Multiple Access (SPMA) protocol optimizes channel resource allocation through priority thresholds and a backoff mechanism. However, in specific scenarios, the traditional SPMA protocol exhibits shortcomings in threshold setting and backoff time calculation. This paper proposes an improved protocol that combines dynamic threshold adjustment and a multi-factor backoff mechanism. For dynamic threshold adjustment, the protocol adapts the thresholds in real-time based on the transmission success rate of data packets at each priority level, ensuring alignment with dynamic service demands. Under high-load conditions, a circuit-breaker mechanism is employed to suppress low-priority transmissions. In terms of backoff time calculation, a channel load differential factor is introduced, integrating priority level, traffic proportion, and load variation speed to construct a multi-factor fusion back-off algorithm. Simulation results demonstrate that the improved protocol significantly outperforms the traditional SPMA protocol in network throughput, transmission success rate, and latency performance. Under low-load scenarios, the transmission success rate of low-priority traffic improves by approximately 5%. In high-load scenarios, the circuit breaker mechanism suppresses low-priority transmissions, ensuring that the transmission success rate of high-priority traffic remains above 80%. At the same time, the improved protocol controls the average end-to-end delay of all priority levels within 10 ms, with the highest priority delay being less than 2 ms, effectively meeting the requirements for differentiated service quality.

Nowadays, spacecraft software is becoming increasingly complex and its functions are gradually increasing. If it can be split reasonably in different software, the project can be managed better. However, considering factors such as cost, power consumption, and wiring, it is difficult to use different processors to run different software. Now, multi-core processor is developing rapidly. One processor contains more than one core, so that different software can be run on different cores. Therefore, this paper proposes a design of a boot and monitor software based on multi-core processor which can enable 2 cores to run 2 different software. The ground tests and on-orbit experiments indicated that this scheme can correctly perform boot for different cores and accomplish refactor software for updating. The boot and monitor software based on this design runs well on the orbiting satellites.

A data compression collaborative algorithm that integrates transmission optimization and delay control was proposed to address the data dimension explosion and complexity increase caused by the large-scale grid connection of distributed photovoltaic systems. Firstly, based on the theory of decision boundary sensitivity, an optimization framework for minimizing total delay was constructed, and an enhanced inter-class boundary preservation algorithm (EIPB) was proposed to reduce the amount of transmitted data by dynamically maintaining key instances of decision boundaries. Secondly, the traditional distance instance selection method was improved by proposing an enhanced instance selection algorithm (EIS) based on K-dimensional tree (KD-Tree) spatial partitioning, which utilized nearest neighbor search acceleration technology to enhance instance selection efficiency. Then, a dynamic error allocation sector lossless compression algorithm (DEASC) was proposed, which achieved collaborative optimization of compression efficiency and fidelity through adaptive slope constraints and multi-stage entropy encoding. Experimental verification showed that the EIPB-EIS joint algorithm improved the average compression ratio to 7.8 compared to traditional methods, reduced the root mean square error of reconstruction percentage to 0.51%, and reduced transmission delay by 62.7%, effectively solving the problem of efficient transmission and accurate reconstruction of high-dimensional photovoltaic data.

To address the multiple requirements of low cost, low power consumption, and high accuracy (sub-nanosecond level)for time-frequency systems in large-scale low Earth orbit (LEO) navigation constellations, a precision frequency control method based on micro-step adjustment is proposed for oven-controlled crystal oscillators (OCXOs), targeting the mitigation of long-term frequency drift. A quantization noise model was established, and a 48-bit direct digital synthesizer (DDS) was employed to achieve high-resolution real-time frequency correction at 5×10^-13 resolution. A constraint framework and frequency adjustment procedure compatible with rapid precise position services were constructed. Experimental results demonstrate that during 24-hour continuous operation, a clock bias accuracy (RMS) of 0.24 ns was achieved. Compared to free-running state, a long-term frequency stability was significantly improved by three orders of magnitude, with a slight degradation of 25% in short-term stability (1 s-20 s). This approach provides a cost-effective, high-precision, and stable time-frequency reference solution for LEO constellation construction.

With the development of electronic technology and motor PWM control technology, the electric servo system has gradually changed from analog system to digital system. Closed-loop control cycle is a key parameter in digital servo system, which is restricted by system maneuverability, hardware resources, power supply capacity, function performance and so on. At present,there are few literatures on the closed-loop cycle of servo system at home and abroad. This paper takes servo system of electric actuator as the research object. Firstly, the composition, working principle and signal transmission link of electric actuator servo system are analyzed in detail. Then the mathematical modeling of the servo system of the electric actuator is carried out, and the P-D control law is designed. Then the servo system simulation model is established based on Simulink to analyze the influence of different closed-loop control cycles on the servo system performance. Finally, an experimental platform is built to verify the effects of different closed-loop cycles on servo system performance. The simulation results show that the step response speed decreases with the increase of the closed-loop period. Appropriately increasing the closed-loop period can reduce the transient current to a certain extent,and avoid the risk of computer reset caused by large transient current pulling down the system voltage. The research in this paper has certain guiding significance for servo system and the platform system interacting with them.

This paper conducts an engineering feasibility study on the dielectric absorption coefficient test requirements stipulated by relevant fixed capacitor standards. The study includes equivalent circuit analysis and physical verification, achieving satisfactory results: confirming the reasonable feasibility of the single-branch model, which effectively simplifies circuit analysis. The residual voltage and current measured after capacitor charge/discharge cycles exhibit distinct peak characteristics, clearly reflecting the presence and variation pattern of the absorption effect. Analysis of the peak conditions readily reveals the necessity for test capacitors to have identical capacitance values, facilitating comparative grading. Expressions for the dielectric absorption coefficient include voltage-based and current-based types. High-input-impedance test instruments must be matched to the source impedance.

With the swift advancement of radar jamming techniques, the variety of active jamming types and the diversity of jamming strategies have surged, urging for accurate identification of jamming types. Conventional active jamming identification methods lack efficiency and universality. Meanwhile, current deep learning-based approaches are encumbered by large-scale parameters and the need for extensive data, which significantly limit their practical applications. To enhance recognition capabilities under conditions with limited parameters and data, a lightweight few-shot radar active jamming identification method based on multi-modality fusion is proposed. Lightweight fusion is achieved by leveraging the temporal locality of time-frequency features and the high-resolution range profile features. Additionally, few-shot classification performance is improved through exploiting metric learning and feature retrieval techniques. Experiments conducted on both simulated and measured datasets demonstrate the superior performance of the proposed method under a variety of conditions.

Due to the impact of global climate change, extreme weather events have become more frequent in recent years, with a corresponding increase in both the frequency and intensity of typhoons. As typhoons have extremely strong destructive power, they can have a significant impact on economic and social development, human life and property, as well as maritime activities. Therefore, the real-time tracking and positioning of typhoons is critical for mitigating their adverse impacts. Based on the infrared cloud image data from the multi-channel scanning imaging radiometer aboard the Fengyun-4 geostationary satellite, combined with the tropical cyclone best track dataset, this study uses the YOLOv8 object detection algorithm to achieve automatic identification and rapid positioning of typhoons. The verification results show that the recognition accuracy for typhoons with strong tropical storm intensity and above exceeds 83%, with a precision rate of over 88%. This achievement provides robust data support for maritime activities, maritime transportation, and oceanographic research, effectively improving the accuracy and timeliness of typhoon monitoring and enhancing safety in related fields.

The radial acceleration between the target and the observation platform causes Doppler frequency changes. High-precision estimation of Doppler change rate has broad application in the field of signal processing. To improve the processing gain,firstly, frequency domain accumulation is used within the pulse, the data for long time is diluted into limited spectral line data, then,the estimation of doppler rate of change is transformed into a least squares estimation through spectral correlation processing of interpulse. It is possible for inter pulse accumulation. The processing gain can be greatly improved, and the high estimation accuracy can be easily achieved by intra-pulse accumulation and multi-pulse least squares estimation. The difficulty of accurately estimating the frequency, time of arrival and initial phase is avoided by cleverly utilizing spectral correlation processing. This algorithm is insensitive to intra-pulse modulation and pulse repetition interval, and has strong universality. Finally, the validity, universality and high-precision estimation of the algorithm are verified through simulation in various scenarios.

The adaptive monopulse technology is a common method for phased array radar to achieve target detection and angle estimation. In order to solve the problem that the performance of the traditional adaptive beamforming algorithm is reduced and the monopulse angle estimation results are greatly deviated in the complex environment where the mainlobe and sidelobe interference coexists, a two-stage mainlobe and sidelobe interference suppression method of adaptive monopulse is designed. The first stage of adaptive processing uses the improved GSC structure to complete the suppression of sidelobe interference, and at the same time, the blocking matrix is constructed to protect the target signal and mainlobe interference from being cancelled in the suppression process of sidelobe interference, so as to improve the mainlobe maintenance effect remarkably. The second stage of adaptive processing uses the sum-difference four-channel mainlobe interference suppression to complete the directional suppression of mainlobe interference,that is, while the mainlobe interference in the elevation or azimuth direction is adaptively suppressed, the non-adaptive array pattern of the azimuth or elevation direction is maintained, and then the monopulse ratio maintains undistorted relative to the quiescent monopulse ratio. The simulation results demonstrate that the proposed method can effectively suppress the mainlobe and sidelobe interferences without distortion of monopulse ratio at the same time.

With the development of marine resources and the increasing demand for safety monitoring, the traditional underwater acoustic detection equipment has problems such as large size and difficult deployment. In order to meet the need of high precision underwater acoustic detection, a design method of seismic wave sensor based on micro-electro-mechanical system (MEMS)technology is proposed in this paper. By constructing the second-order dynamic model of the sensitive structure, the fully integrated closed-loop system control is realized by combining the ΣΔ modulator, and the behavior level modeling and non-ideal characteristic simulation optimization are completed by Simulink platform, and the chip integration is realized by 0.5 μm CMOS process. A fully differential precharge amplifier and a correlation double sampling circuit are designed. The test results show that the sensor sensitivity is up to 1.5V/g, the operating bandwidth is 300 Hz, the dynamic range is about 136 dB at 1 Hz bandwidth, the equivalent input noise is as low as 154 ng/√Hz, the nonlinearity is 0.065%, and the bias stability is about 28.2 μg, The FOM value is 204.5 μW·μg/Hz. The high-precision application potential of MEMS technology in the field of underwater acoustic detection is verified, and a miniaturized solution for seismic monitoring and marine environment perception is provided.