Abstract：To meet the future industry application requirements, the 6G network needs to build a high-performance, highly flexible, and highly reliable network. Starting from the scenarios and requirements of the 6G distributed network, this study designs a 6G distributed network architecture composed of central network nodes and distributed network nodes, and proposes key technologies such as information interaction at the network node granularity, service continuity assurance, network autonomy, and networking coordination. The research results show that this architecture and key technologies can effectively support the diversified services and flexible deployment of the 6G network, providing theoretical support and practical guidance for the development of 6G networks.

Abstract：The emergence of non-terrestrial networks (NTN) marks a major breakthrough in communication technology. It expands the coverage of terrestrial networks by using satellites and high-altitude platforms, thus providing new solutions for global communications. This paper provides a multi-faceted research of NTN, including the following aspects: Firstly, the basic concepts and importance of NTN are introduced, and the deployment mode, application scenario and working frequency bands of NTN are comprehensively outlined, while different system architectures are described in detail; secondly, the main technical challenges faced by NTN are discussed, and related research results are summarized; finally, the future development of NTN is prospected, and potential technological innovations and development paths are pointed out.

Abstract：With the widespread application of encrypted communications, traditional malicious traffic detection methods based on content analysis have gradually become ineffective. How to efficiently detect malicious behavior in encrypted traffic has become a research focus in the field of network security. This paper proposes a neural network-based encrypted malicious traffic detection method, which realizes the classification of malicious encrypted traffic through a deep learning model. First, the network traffic is preprocessed and key features are extracted, including packet size distribution, time interval, and protocol type. The features are then mapped into a two-dimensional feature map as the input of the deep learning model. A scalable window self-attention mechanism is designed, and the Transfomer neural network model is used to classify feature maps, achieving efficient detection of malicious traffic.Experimental results show that this method performs well in detection accuracy, recall rate, and model robustness, and provides a feasible solution to the problem of malicious behavior detection in encrypted traffic.

Abstract：Network slicing is crucial for enabling networks to meet the diverse service demands of various verticals. To address the issue of dynamic changes in slice types, a Federated Multi-Agent Reinforcement Learning (F-MALML) algorithm with dualtime scale resource allocation is proposed. At the large time scale, a finite memory learning algorithm allocates resources to each base station. At the small time scale, each base station uses F-MALML to dynamically allocate resources to users. A probabilistic learning mechanism is introduced to adjust the allocation strategy based on previous results and the current network state. Simulation results show that the proposed algorithm achieves significant improvements in slice satisfaction for newly added slices and system spectral efficiency compared to the two benchmark algorithms, while demonstrating better stability.

Abstract：This article reviews the latest developments and trends in intelligent processing technology for guidance radar. This article elaborates on the current technological challenges in the field of intelligent processing of guidance radar, starting from multi-source information fusion technology, tracking and filtering methods, intelligent and cognitive technology, and intelligent anti-interference technology. In summarizes the research progress in recent years and looks forward to the future development trend of this comprehension technology. Overall, intelligent tracking and processing technology is developing towards high integration, autonomous intelligence, and multimodal fusion, aiming to enhance the accuracy and reliability of guidance radar, and provide more powerful, flexible, and reliable guidance information for future guidance systems.

Abstract：With the rapid development of machine learning, intelligent models created by machine learning algorithms are gradually becoming the core part of new sensor data analysis. The paper firstly introduces the background of intelligent sensors that support machine learning, briefly describes the construction of intelligent models and the generation, verification and testing process of data sets, then lists the application of intelligent sensors based on machine learning, finally points out the current problems and challenges of intelligent sensors based on machine learning and puts forward feasible solutions to provide valuable academic reference for relevant researchers.

Abstract：To improve the bandwidth and efficiency of power amplifier, a monolithically integrated X-band high-efficiency continuous B/J power amplifier is designed based on the 0.25 μm GaAs pseudomorphic high electron mobility transistor (pHEMT) process. Continuous B/J waveform is achieved by controlling the second harmonic impedance of the output stage. The amplifier consists of two stages, with the driver stage using enhanced transistors for high gain and the output stage using depletion transistors for high efficiency and watt-level output power. The simulation results show that the power amplifier achieves an output power of 29~30.6 dBm in the frequency band of 7.3~12.2 GHz, with a power gain of 17~18.6 dB, a power added efficiency of more than 50%, a peak efficiency of 59%, an input return loss of less than 10 dB, and a chip size of only 2.1 mm × 1.3 mm.

Abstract：A magnetic isolation IGBT drive architecture with short circuit protection is designed in this paper. The architecture adopts transformer isolation, integrates desaturation detection circuit, Miller clamp circuit and soft turn-off. When IGBT exits the saturation zone due to short circuit fault, soft turn-off is performed on the device. The gate voltage spike is suppressed by Miller clamp circuit. At the same time, the fault signal is fed back to the front controller through the fault feedback channel to realize the fast response to the short-circuit fault. The simulation and measured results show that the architecture has an isolated voltage withstand of 8 kV, desaturation detection and Miller clamp threshold of 9 V and 2 V respectively, desaturation fault response time of 419 ns, fault error time of 311 ns, soft shutdown time of 136 ns. The architecture realizes short-circuit fault protection and fault feedback, and has been applied to a high voltage isolated IGBT driver.

Abstract：This paper presents the design and implementation of a low-offset, high-gain operational amplifier (op-amp). The overall circuit consists of a bandgap reference, oscillator, frequency divider, auxiliary op-amp, and main op-amp. By employing a ping-pong auto-zeroing architecture, the circuit achieves continuous operation while significantly reducing the input offset voltage of the amplifier. Furthermore, this paper proposes a novel and efficient control timing scheme that effectively reduces the settling time during ping-pong switching at a relatively low cost, further minimizing the output glitches of the amplifier. The amplifier chip was designed and fabricated in a 350 nm CMOS process. Measurement results demonstrate that under a 5 V supply voltage, the amplifier consumes 0.65 mA current, achieves a maximum input offset voltage of 3 µV, provides a gain-bandwidth product of 8 MHz, and exhibits a noise spectral density as low as 30 nV/√Hz.

Abstract：The traditional manual opening mode of rotary hidden door handle does not have the intelligent automatic unfoldment and retraction function, so a design scheme of hidden door handle controller is proposed. The controller uses N32G455RBL7 as the main control chip, and listens to the on/off lock, gear and other signals on the CAN bus of the body controller through its on-chip bxCAN (basic extended CAN) module, and drives the supporting door lock execution motor according to the intelligent control requirements, so as to realize the electrified and intelligent control of the hidden door handle. The controller has been tested for a long time on a number of domestic new energy vehicles, and the results show that it has good stability and reliability in practical applications, and can provide a reference for the intelligent control scheme of manual hidden door handles of other structures.

Abstract：Aiming at the problem of false detection and missing detection in multi-scale SAR ship object detection in complex scenes, an improved SAR ship object detection method is proposed in this paper. Firstly, a multi-scale object feature extraction network (MFE-Net) is used to extract feature information to improve the detection capability of multi-scale objects and reduce redundant calculations. Secondly, deformable convolution (DConv) is introduced to improve the detection performance of SAR ships in complex scenarios by adjusting the shape of the convolution kernel adaptively. Finally, an attention mechanism is introduced to suppress background clutter and enhance feature information. The experimental results show that the detection accuracy of the proposed method on SSDD and HRSID data sets reaches 97.9% and 93.1%, respectively, and the overall performance is better than the existing mainstream object detection algorithms.

Abstract：With the construction of Internet of Things in power system, safe and efficient intelligent inspection technology is an important guarantee for the healthy development of the power industry. Intelligent inspection has the characteristics of high communication requirements and complex working environment, so it is difficult to match with wireless technologies. Firstly, it analyzes the characteristics of intelligent inspection service and the requirements for various wireless technologies, and then proposes an adaptation method of intelligent inspection service and wireless technology based on Bayesian Best-Worst Method (BBWM) and improved VIKOR method. Finally, the adaptation simulation experiments of three intelligent inspection services and five wireless technologies are carried out. Experiments show that the proposed method can solve the adaptation problem of intelligent inspection service and wireless technology well.

Abstract：To improve the accuracy of mulberry leaf disease detection and enable convenient and rapid deployment of models on mobile devices, an improved version of the YOLOv8 model, named YOLOv8-Evo, is proposed to address issues such as small lesion spots and complex backgrounds in natural environments. The algorithm introduces a deformable convolution module within the Backbone to capture disease details and shapes more flexibly. Additionally, a Convolutional Block Attention Module (CBAM) is incorporated into the Neck to highlight key features and regions in the image. After validation on a dataset of 18 849 mulberry leaf disease images, the YOLOv8-Evo model demonstrates a 2.4% increase in precision, a 1.5% increase in recall rate, a 1% improvement in mAP50, and a 0.7% improvement in mAP50-95 compared to the YOLOv8s model. These results provide both theoretical support and technical backing for the automation of mulberry leaf disease identification.

Abstract：That the frequency of frequency modulated continuous wave(FMCW) signal varies periodically makes the traditional time-frequency analysis very difficult and crabbed. The phase of the beat signal is analyzed in this paper. The beat signal is proved to be periodical and its spectrum is discrete. The power is distributed to all the harmonics. The weight of each harmonic depends on frequency deviation,distance,modulation period. When the target moves in radial direction, spectrum of beat signal is influenced by Doppler effect and becomes nonperiodic when the Doppler shift is not an integral multiple of original repetition frequency.

Abstract：A novel two-way out-of-phase power divider（PD）is proposed in this paper. The design principle is similar to conventional Wilkinson structure. By setting the length of the two-output transmission lines properly and introducing an extra line at one end of the isolator, the power divider obtains equal divide, out-of-phase, and good isolation in the passband. Besides, the parameters of the structure are deduced in detail using even-odd mode analysis method. Finally, a power divider, centered at 3.8 GHz，is fabricated to demonstrate the principle by planar microstrip line. It is observed from the measurement that the PD has 1 dB fractional bandwidths of 31%, better than -15 dB return loss and 20 dB isolation, respectively.

Abstract：A novel X-band wideband dual-circularly polarized microstrip array antenna is designed to meet the requirements of a satellite X-band communication data forwarding system. The antenna unit adopts microstrip form, the overall energy feed mode is bottom feed, the slot-coupling feed mode is adopted to achieve wide band, and the 3 dB directional coupler is adopted to realize the phase difference between the two signals to meet the dual circularly polarization. The array antenna consists of 24 antenna units arranged in a 4×6 rectangle, each of which is rotated in a group of 4 units. Through the analysis of electromagnetic simulation software, it can be obtained that the array antenna has a gain of 17.9 dBi and an axis ratio of 0.06 dB at the center frequency of 8.2 GHz in the operating frequency band of 8 GHz~8.4 GHz. The antenna achieves the performance indexes such as wide band and double circularly polarization, and has certain guiding significance for the design of X-band wireless communication system.

Abstract：In response to the challenges in designing silicon carbide (SiC) MOSFET driving circuits, such as high difficulty in design, susceptibility to crosstalk, incomplete protective functions, and the issue of full domestic production, a half-bridge driving and protection circuit for silicon carbide MOSFETs has been designed based on domestic chips. This paper focuses on analyzing and summarizing the principles and models of active Miller clamp protection, desaturation protection, and bridge arm interlock protection for silicon carbide MOSFETs. The driving circuit not only isolates the primary and secondary signals but also uses high and low levels of 18 V/-3.3 V to control the upper and lower bridge arm silicon carbide MOSFETs. It integrates functions such as undervoltage lockout, desaturation protection, bridge arm interlock, and active Miller clamp protection. The parameter comparison and function test were carried out with the international advanced level of Wolf Speed's silicon carbide MOSFET driver board CGD1200HB2P-BM2. The experimental results show that the switching parameters of the circuit are similar to CGD1200HB2P-BM2 driver board, which can meet the driving requirements of silicon carbide MOSFET and can trigger the protection function reliably. This circuit has been applied to drive silicon carbide MOSFET.

Abstract：Aiming at the problems that the equipment is difficult to effectively collect underwater magnetic anomaly signal detection, the dynamic range of analog-to-digital converter is limited, resulting in low signal-to-noise ratio, and weak signals are easily submerged by noise. This thesis designs a high-precision magnetic anomaly signal acquisition system based on domestic field-programmable gate array (FPGA). In order to improve the signal-to-noise ratio and dynamic range of the signal, this paper uses the high-precision multi-channel analog-to-digital conversion chip ADS1278 and the Unisplendour Logos series FPGA to collect the weak magnetic anomaly signal output by the three-axis fluxgate after using 2 times and 8 times gain pre-processing respectively, and then uses the Kalman filter to perform secondary correction on the fused data after the average weighted average of the collected signals for preliminary data fusion. The experimental results show that the accuracy of the fused signal is significantly better than that of the single-channel acquisition signal. When using dual-channel AD to acquire weak signals with a frequency of 200 Hz and an amplitude of 500 μV, the signal-to-noise ratio is improved by 26.099 dB, effectively improving the dynamic range of the data acquisition system.

Abstract：Synthetic Aperture Radar (SAR) is an active earth observation system that can observe the earth 24/7. Pulse power supply is an important component of SAR operation. This article proposes a design method for a nationally produced aerospace grade high-efficiency SAR pulse power supply, with an input voltage range of 78 V~110 V, suitable for the requirements of a 100 V wide input bus on board. It has the characteristics of small size, high power density, and low input reflection current ripple. The load pulse repetition frequency covers the range of 500 Hz and above, and adopts nationally produced and aerospace grade devices.

Abstract：Because of the special environment of the satellite-borne power supply, there are not many effective ways to dissipate heat, so the power density and efficiency of the power supply are more demanding. And because of its strict requirements for components, the choice of circuits and components are also correspondingly limited. This paper mainly introduces the design of flyback converter optimized on the basis of the original multi-channel power supply, and analyzes its working principle, then designs a new satellite-borne multiple power supply. Finally, the efficiency of the newly designed multiplex power supply is compared with that of the original multiplex power supply, and the feasibility of the optimized power supply design is verified.