Abstract:To promote the application of fifth-generation mobile communication technology(5G) in the construction of high-speed rail dedicated networks, this paper takes the Xiamenbei Railway Station in the 2.1 GHz frequency band as the research scenario, studying the significant multipath spatial characteristics brought about by objects such as columns, tracks, and platforms in the scenario. Through Ray Tracing(RT) simulation, multidimensional multipath data from the transmitter to the receiver in the high-speed railway station scenario is obtained; based on the K-means clustering algorithm, the multipath signals are clustered, and relevant cluster parameters are extracted and analyzed to complete the construction of the Cluster Delay Line(CDL) model for the high-speed railway station scenario. This provides spatial domain information of multipath for different polarization combinations of transmitters and receivers in the high-speed railway station scenario, complementing the channel model of high-speed railway station scenario for 5G-Railway(5G-R), and serving the construction of dedicated mobile communication systems for railways using 5G.

Abstract:Urban rail transit is one of the primary systems for urban transportation capacity. With the increasing demand for the development of smart cities in recent years, the design, research, and optimization of smart rail transit systems have become the research direction and focus for many scholars. Smart rail transit systems require trains to have intelligent computing power to meet a variety of intelligent service needs. Due to the numerous limitations of on-board equipment in trains, it is not practical to deploy high-performance computing devices on them, hence the need to introduce other devices to provide computational support. This study, aimed at the special scenarios of smart rail transit systems, designs a cloud-edge collaborative computing architecture for intelligent tasks based on 5G and edge intelligence. The resource allocation process in this architecture is mathematically modeled and transformed into an optimization problem that minimizes task latency. To solve this optimization problem, this paper employs a discrete stochastic approximation algorithm to minimize the total processing delay of tasks in the smart rail transit system. Simulation results indicate that the algorithm can effectively reduce the processing delay of intelligent tasks in smart rail transit systems.

Abstract:Urban rail transit plays a significant role in alleviating urban traffic congestion, and the coordinated control of multiple urban rail vehicles has been a research hotspot in recent years. The multi-vehicle coordinated computing task is limited by communication, leading to issues such as poor resource allocation balance, slow system response to environmental changes, and limited cooperative operation capabilities. The integration of 5G communication and Mobile Edge Computing(MEC) can effectively improve the real-time and accuracy of task processing, enhancing the overall system performance. This paper designs an autonomous coordinated computing architecture for urban rail vehicle operation control systems based on 5G and MEC. According to the characteristics of multi-vehicle coordinated control tasks, the problem of edge server selection in multi-vehicle coordinated computing offloading is modeled as a Multi-Armed Bandit(MAB) learning model, and a solution based on the Upper Confidence Bound(UCB) algorithm is proposed to minimize the overall energy consumption and latency of the urban rail vehicle multi-vehicle coordinated control system. Simulation results show that the proposed algorithm model has significant performance advantages in terms of average reward, best selection probability, average execution latency, and weighted total cost.

Abstract:To assist in determining the base station spacing for the new generation of mobile communication systems for 5G-Railway(5G-R) in complex railway environments, a link-level transmission rate simulation technology is proposed. This technology relies on various empirical propagation models to obtain the path loss of electromagnetic waves in different scenarios along the railway line, thereby assessing the Reference Signal Received Power(RSRP) of the wireless link. Based on the obtained RSRP data, link-level transmission rates are calculated by using the Monte Carlo simulation method. Simulation results indicate that in open-field scenarios, base station spacing can be set between 2 000 meters to 3 000 meters, while in urban and complex terrain environments, the spacing should be reduced to maintain performance. This technique accurately reflects the signal transmission characteristics of 5G-R systems across various scenarios, including the changes in transmission rate and RSRP with different base station spacings. And it provides a scientific basis for future base station deployment planning in 5G-R systems.

Abstract:In response to the difficulty in detecting defects in high-speed rail clip springs caused by complex lighting environments, an improved Faster Region Convolutional Neural Networks(R-CNN)-based defect detection method for clip springs is proposed. By extracting defect feature maps through multi-layer convolutional neural networks, the network's attention to defect features is enhanced, and the impact of interference from complex lighting environments is reduced. A region proposal network is designed to generate candidate regions, and based on these regions, pooling is performed to extract the corresponding specific defect locations in the feature maps. The fully connected layers of the region proposal network are employed to calculate the specific categories and precise locations of defects, yielding the final detection results. The proposed algorithm can fully suppress the interference of lighting environments, significantly enhance the representation ability of defect features, simplify the image pre-processing stage, and reduce the requirements for the quality of the original image. Experimental results show that the proposed algorithm can effectively detect defects in high-speed rail clip springs, and compared to existing algorithms, it has a higher accuracy, stronger robustness, and significantly improved computational efficiency.

Abstract:A Noise Source(NS) is a device that generates controllable noise(with controllable frequency and power)，and it is an important tool in the development process of devices. Terahertz noise sources have important applications in terahertz device noise figure testing, calibration of space-borne microwave radiometers, imaging, and spectral analysis. This article systematically reviews the development and research status of Terahertz(THz) noise sources at home and abroad, analyzes the technical characteristics of terahertz noise sources based on thermodynamics, electronics, and photonics, and looks forward to their future development directions.

Abstract:The existing passive metasurfaces have relatively limited functions. To increase the flexibility of controlling terahertz waves, a polarization-multiplexed reflective metasurface is proposed. This metasurface is capable of various functions such as beam deflection, dual-focus focusing, focused vortex, and far-field imaging. At a frequency of 1 THz, by separately incident x-polarized and y-polarized waves, the metasurface successfully generated two deflected beams and achieved dual-focus focusing on the focal plane. In addition, the metasurface also realized focused vortex beams with a topological charge of l=+2 and far-field imaging under the conditions of x-polarized and y-polarized wave incidence, respectively. The designed metasurface demonstrates the ability to flexibly manipulate terahertz waves under different polarization wave incidences, showing potential application prospects in the field of terahertz communication.

Abstract:The sensor optimization deployment is a multi-objective optimization problem involving sensor coverage effectiveness, frequency conflict probability, and resource utilization. The existing sensor optimization deployment methods mostly adopt weighted approaches to transform multiple optimization objectives into a single objective problem for resolution, which not only relies on prior knowledge but also leads to the loss of diversity in optimal solutions. To address these issues, a Collaborative evolution Multi-Objective Particle Swarm Optimization(CoMOPSO) algorithm is proposed. It designs a collaborative evolution framework that guarantees the convergence of high-dimensional problems through the convergence of the population, and rapidly approaches the Pareto optimal frontier. The diverse population uses the ??-dominance method to ensure the integrity and diversity of the global and local optimal solution sets. A fast non-dominated sorting and elite individual preservation strategy is employed to enhance the quality of solutions. Experimental results demonstrate that, for the sensor optimization deployment problem, the proposed method outperforms traditional optimization algorithms in terms of Inverted Generational Distance(IGD) and M3* indicators, exhibiting better convergence and diversity and effectively improving the performance of sensor networks.

Abstract:Dual-satellite Time-Difference of Arrival(TDoA) and Frequency-Difference of Arrival (FDoA) positionings utilize just two satellites to locate the emitter, bearing lower cost and difficulty than multi-satellite positioning, and better real-time performance than single-satellite positioning system, more suitable for practical applications. In order to solve the nonlinear optimization problem in the process of solving the TDoA/FDoA equation, an Improved Particle Swarm Optimization(IPSO) algorithm is put forward. Dual-satellite TDoA/FDoA location systems use Newton method with high precision, but there exists unsolvable blind areas of initial value of iteration. To address this issue, an cascaded localization method of IPSO and Newton iteration is proposed, in which the IPSO algorithm gives coarse localization result with rapid speed and reliable convergence, and this coarse result is used as the initial value of Newton iteration, so as to reduce the positioning error and avoid non-convergence. By analyzing the simulation results, the proposed algorithm not only increases the success rate of effective initial point selection by 48.15% compared to the Newton method, but also cuts the computation and improves the computational speed.

Abstract:The conventional classification and identification of cruise missiles, air-to-ground missiles, and guided bombs rely heavily on the kinematic characteristics of the targets, often using parameters such as the target's altitude and speed as the basis for classification judgment. With the advancement of missile weapon development technology, single-dimensional kinematic characteristics can no longer meet the demand for refined identification. To address this issue, a classification and identification method for missile targets has been designed, which integrates the kinematic characteristics and Radar Cross-Section(RCS) series characteristics of the targets. By employing a multi-feature hierarchical identification approach, it achieves the classification and identification of the three types of missile targets. Simulation test results indicate that the classification accuracy of this method for the three types of missile targets is better than 95%, validating the effectiveness of the method.

Abstract:Absorption boundary condition is an important condition for solving electromagnetic computing problem by Finite Difference Time Domain(FDTD) algorithm. First of all, four absorption boundary methods are selected, and the principles of Convolution Perfectly Matched Layer(CPML), Surface Impedance Boundary Conditions(SIBC) and the absorbers are derived respectively; FDTD algorithm is programmed and calculated by Matlab software. Then, the four methods are applied to two microstrip structures, the low-pass filter and the branch coupler, and the reflection coefficient is calculated. By changing the thickness of the air layer and the iteration step size, the four methods can get the correct results. Finally, the results of the two models are compared and analyzed in terms of computing time, memory usage and percentage of error. Compared to other absorption methods, SIBC has the least memory usage, higher computational efficiency, and an error percentage of 1.63%.

Abstract:The function and working principle of the True-Time Delay(TTD) chip has been studied and the chip has been applied in phased array radar. Based on GaAs Pseudomorphic High Electron Mobility Transistor(PHEMT) technology, three broadband millimeter-wave TTD chips are designed and fabricated. The measured results on wafer show that the 6 bit TTD provides 0.446 ps up to 28.125 ps in the frequency range 32 GHz-40 GHz, the TTD phase error for all 64 bit states is -2°~9°, the Insertion Loss(IL) is less than 19 dB, and the 64 states insertion variation is reduced to ±1 dB, the input and output Voltage Standing Wave Ratios(VSWRs) are better than 1.5 on the whole bandwidth; the 4 bit TTD provides 7.142 ps up to 107.148 ps in the frequency range 32~40 GHz, the TTD phase error for all 16 bit states is better than±12°, the IL is less than 12 dB, and the 16 states insertion variation is reduced to ±1 dB, the input and output VSWRs are better than 2.0 on the whole bandwidth; the 3 bit TTD provides 28.57 ps up to 200 ps in the frequency range 32~40 GHz, the TTD phase error for all 8 bit states is -10 °~22 °,the IL is less than 14 dB, and the 8 states insertion variation is reduced to ±1 dB, the input and output VSWRs are better than 1.8 on the whole bandwidth. The TTD chips are characterized with ultra wideband, large time delay and compact size, and mainly applied to wideband active phased array applications.

Abstract:For the problem of symbol rate estimation in non-cooperative communication, an algorithm based on empirical mode decomposition for symbol rate estimation is proposed. The algorithm performs a low-order Empirical Mode Decomposition(EMD) on the instantaneous amplitude of the signal to find the intrinsic mode components that contain information about the symbol rate; then, it estimates the symbol rate through spectral line detection. This algorithm can effectively reduce the impact of interference on estimation performance by separating frequencies, thereby improving estimation accuracy. Simulation results show that the algorithm can obtain an estimate of the symbol rate directly from the frequency band signal under low signal-to-noise ratio conditions.

Abstract:In recent years, deep learning methods have been widely applied in the field of signal processing and have achieved good results. Deep learning methods can automatically acquire useful signal features from massive signal data using neural network models designed by experts, but the manual design of deep neural network models remains a time-consuming and error-prone process. To address this, a method for Automatic Modulation Classification(AMC) based on progressive neural architecture search is proposed. This method can automatically design network structures according to specific modulation classification tasks and obtain the optimal lightweight deep neural network by following a search strategy that maximizes the model performance. Simulation results show that compared to deep learning-based modulation classification methods, the proposed method can achieve optimal modulation classification accuracy without manual design of neural networks, with low parameter volume and floating-point operations, achieving an average recognition accuracy up to 92.82%.

Abstract:Due to the exposure of data caused by open source code, traditional methods cannot block the transmission of attacked data packets, resulting in the inability of data to autonomously defend. Therefore, a network link data tampering autonomous defense system based on Snort is designed. In the hardware part of the system, a packet sniffer is adopted to capture Snort messages, and the layered decoding is integrated into text information through an information decoding module. The integrated text information is sent to the system database through the network for processing high-volume alert data and storing records. In the system software section, a network depth defense model based on Snort is constructed to achieve real-time detection and automatic interception of tampering attacks. Based on the transmission characteristics of information packets in the network, the transmission distance between different nodes is calculated and the location of defense nodes is determined. The data transmission path is derived when the link layer data is subjected to tampering attacks, and an autonomous defense function is constructed for data tampering, therefore the autonomous defense of data is achieved. Using wavelet denoising data processing technology to obtain time-series data, using inverse wavelet transform reconstruction to obtain denoised data, the design of an autonomous defense system for data tampering is completed. According to the experimental results, the system has a high density of secure transmission of network link data, and the maximum success rate of key recovery can reach 98%, demonstrating strong robustness.

Abstract:In response to the issues of low tracking accuracy in video sequences due to factors such as appearance changes, background clutter, and severe occlusions, a novel two-stage adaptive tracking model is proposed. This model includes two phases: target detection and bounding box estimation. In the target detection phase,the model roughly locates the target; in the bounding box estimation phase, the exact position of the target is determined. To address the complexity of video scenes and the challenges of tracking small targets, multi-feature fusion technology is employed to construct a rich target representation. Experimental results show that compared with models such as Simple Online and Realtime Tracking(SORT), Tracktor++, FairMOT, and Transformer, this model demonstrates the best overall performance, effectively balancing the relationship between computational speed and tracking accuracy, and showing good potential for application.