Abstract：With the rapid upgrades of products such as AI PCs, Wi-Fi routers, and GPON, high-speed wired network is an essential part of network infrastructure, Ethernet PHY chips have to provide higher transmission rates to better meet the growing bandwidth demands in the era of big data. 10G Ethernet technology is bound to become the key evolution direction for the next generation of Ethernet. The development of 10G Ethernet PHY chips need to overcome a series of technical challenges, including high-speed signal integrity and low-power design. This paper focuses on the key issues of signal processing in 10G Ethernet PHY chips, firstly introducing the background of relevant international standards for 10G Ethernet, then discussing the network requirements for 10G Ethernet, and next describing the key technologies for signal processing at the physical layer of 10G Ethernet. Finally, the paper makes a summary and outlooks the further development of high-speed Ethernet.

Abstract：LDPC code is a widely-used high-performance error correction code. In recent years, LDPC decoding based on deep learning and neural networks becomes a research hotspot. Based on the (512,256) LDPC code of the CCSDS standard, this paper firstly studies the traditional decoding algorithms of SP, MS, NMS, and OMS, laying a foundation for the construction of neural networks. Then a data-driven (DD) decoding method is studied which adopts the information with its encoded, modulated and noise-added LDPC code as the training data within a Multi-layer Perceptron (MLP) neural network. In order to solve the problem of high bit error rate (BER) in data-driven method, the Neural Normalized Min-sum (NNMS) decoding in which the NMS algorithm is mapped to the neural network structure is proposed, achieving more excellent BER performance than that of NMS. The BER declines by 85.19% when channel SNR equals to 3.5 dB. Finally, improved training methods to enhance the SNR generalization ability of the NNMS network is studied.

Abstract：As one of the key technologies to realize the vehicle intelligence, the automotive data transmission technology with high speed, high reliability, and low latency, is getting more and more attention. At the same time, the demand for breaking the monopoly of proprietary protocols and enhancing interconnectivity between devices is growing increasingly, driving the development of public standards on automotive data transmission to be a hot topic for many domestic and international standard organizations. This paper firstly introduces the features of the automotive data transmission link. Then, the progress of the mainstream standards is briefly summarized, with a special focus on the high-speed automotive data transmission physical layer interface chips. We conduct a comprehensive comparison and analysis of the main performance indicators and physical layer technologies for these standards.

Abstract：Dynamic Random Access Memory (DRAM) is widely used in modern large-scale computers and ultra high speed communication systems due to its high storage density and cost-effectiveness. This article mainly introduces the development history, key technologies, domestic and international research progress, and future development directions of dynamic DRAM. Firstly, the classification, basic unit structure, and working principle of DRAM are introduced. Secondly, the key performance indicators of DDR SDRAM and the development of dedicated DRAM are introduced in detail. Then, innovative DRAM architectures and technologies that improve DRAM access speed, capacity, and density are introduced, as well as capacitor free storage cell structures, 3D stacked DRAM technology, and Rowhammer security issues and defense mechanisms. Finally, the future development direction of DRAM technology is discussed, and the importance of in-depth research and innovation on existing DRAM technology is emphasized to meet the growing demand for high-speed, low-power, and high reliability storage.

Abstract：This study proposes a crossbar matrix transmission design based on adaptive optimization, employing the Advanced High-performance Bus (AHB) protocol. It introduces innovative mechanisms for adaptive burst transmission adjustment and adaptive priority adjustment. By dynamically adjusting burst transmission length and priority allocation, efficient data flow management is achieved, improving bandwidth utilization and transmission efficiency. Front-end simulation and back-end layout routing were conducted to validate the design. Simulations confirm the method’s superiority under different load conditions, optimizing bus resource allocation, increasing transmission speed, and reducing overall power consumption.

Abstract：In this paper, a system verification architecture for time-sensitive network (TSN) switch chip, which is based on UVM, automatic comparison and coverage-driven verification thinking, is designed. This architecture adopts the method of classifying and pipelining packets, combined with traffic detection, time slot detection and packets automatic comparison scheme, successfully satisfies the implementation of the TSN system verification method, finally ensures the system verification completion. Chip test results meet commercial requirements, once again demonstrates verification architecture completeness.

Abstract：With the explosive growth of data generated by the widespread use of cloud computing, Ethernet, and the Internet of Things (IoT), the demand for wired input/output (I/O) bandwidth in large data center networks has surged rapidly. The development of high-speed Ethernet aligns with the rapid growth of network traffic. However, as the data transmission rate of Ethernet increases, the challenges related to signal integrity in serial links are further amplified. This paper focuses on the challenges faced by Ethernet equalization technologies under high-speed transmission, analyzing various equalization techniques in depth. It explores the working principles and characteristics of different equalizers and discusses their applicability in high-speed transmission environments. The paper provides a reference for the future development and optimization of high-speed Ethernet equalizers, aiming to better meet the growing demand for higher transmission efficiency and lower bit error rates in communications.

Abstract：Due to significant grammatical differences and a scarcity of linguistic resources between Chinese and Vietnamese, the task of Chinese-Vietnamese neural machine translation faces challenges in the accurate translation of nouns. This paper proposes a novel multimodal neural machine translation method that integrates a text-based pre-trained model with a visual-linguistic joint pre-training model. The text-based model captures deep linguistic structures and semantics, while the visual-linguistic joint training model provides visual context related to the text, which helps the model understand and translate nouns more accurately. The two models are combined through a streamlined and efficient mapping network and dynamically integrate multimodal information via a Gumbel gating module to optimize translation outputs. In both Chinese-Vietnamese and Vietnamese-Chinese translation tasks, this method has achieved improvements of 7.13 and 4.27 BLEU points, respectively, compared to the traditional Transformer model.

Abstract：To address the issues of automated sorting and classification of various fruits post-harvest, particularly in the identification of decayed fruits, this paper introduces an enhanced model derived from YOLOv5n, denominated as mobile-YOLO. The initial modification involves replacing the backbone network of YOLOv5n with MobileNetV3 and incorporating depth-wise separable convolutions. This enhancement results in improved computational efficiency and speed compared to the original model, accompanied by an elevation in accuracy．

Abstract：With the rapid development of social networks, people have various virtual identities in social networks. User identity linkage problem that aims to identify various virtual identities of the same natural person is becoming increasingly important. User identity linkage method can unearth some hidden information and form a complete user profile to promote the development of multiple research fields, such as cross-network recommendation, link prediction, information dissemination, etc. Existing user-profile based model and network-structure based user identity linkage model do not consider the influence difference between different users, and the convergence speed is slow. In order to model the influences between users, multi-head attention mechanism is added to network random-walk based user linkage method in this paper. The experimental results show that it can improve the effectiveness of social network user identity linkage method and improve training efficiency.

Abstract：Sonar technology is the main method of the ocean exploration, and underwater acoustic sensor is an important part of sonar equipment. Underwater acoustic sensors are divided into scalar and vector underwater acoustic sensors. The vector sensor can make up for the incomplete sound field information obtained by the scalar sensor, and can accurately measure the sound pressure in water and the vibration velocity to determine the target orientation, etc. This paper mainly studies the frequency response characteristics of velocity vector underwater acoustic sensor, discusses the amplitude model, response ratio and frequency response characteristics of velocity type underwater acoustic sensor under high fidelity and high gain, puts forward the design principles of frequency response, and finally summarizes the application of velocity underwater acoustic sensor. Underwater acoustic sensor is an important means of underwater detection, positioning and communication. The research helps to the development of velocity acoustic sensors with high sensitivity and better low frequency characteristics, so as to greatly improve the operating range of traditional acoustic sensors. The development of far-sound-field sensor helps to improve the detection range of underwater targets such as submarines and UUV, which will fundamentally change the pattern of China's underwater acoustic detection network with important military significance.

Abstract：To address the self-excitation issue in the measurement of capacitive impedance using LCR meters, an LCR instrument has been developed to enhance measurement stability. By analyzing the phase margin of the self-balancing bridge circuit at the measurement front-end, a phase compensation circuit has been employed to resolve the self-excitation problem caused by phase lag. Utilizing direct digital frequency synthesis combined with a phase calibration circuit design, quadrature measurement signals of accurate phase are generated. A switch-phase discriminator lock-in amplifier is used to process weak signals, which are sampled by a 16-bit analog-to-digital converter and input into the microprocessor for vector signal synthesis and output display. Compared to professional instruments, the prototype designed has achieved a relative error of up to 0.5% for resistance measurements and up to 1% for large capacitance measurements.

Abstract：In order to solve the problem of the quality and stability of indoor signal coverage of the fifth generation mobile communication technology (5G), a new indoor 5G signal distribution prediction method is proposed in this paper. A depth residual neural network model based on the full connection layer is adopted. The three-dimensional spatial coordinate information of transmitter and receiver and Reference Signal Receiving Power (RSRP) data of receiver are used as input features, without the need to collect complex environmental features. The experimental results show that the depth residual neural network model can effectively predict the indoor 5G signal distribution with MAE of 0.029 455 and RMSE of 0.041 495 compared with the actual value after normalized training without relying on detailed environmental parameters. This study confirms the effectiveness of the prediction method based on deep residual neural network in indoor 5G signal coverage prediction, which provides scientific basis and technical means for optimizing indoor 5G network deployment and improving user experience, and has important practical application value.

Abstract：As 5G and mobile technologies continuously develop, various types of mobile devices are continuously emerging. Mobile devices have driven the advent of the era of the Internet of Things (IoT). Since CDMA technology appeared, SIM cards can be detached from the mobile devices. Host and communication modules have further decoupled after a phone can fit all kinds of carrier networks. People hope their mobile devices can access any network at any time and any place. It is well known that the rapid advancement of wireless technology is precisely due to its alignment with the trend of division of labor and independent development. In the era of IoT, terminal devices are often characterized by low specifications and large quantities. The continuous emergence of new technologies and frequency bands has resulted in asynchronous development between host devices and communication modules. This combination-based terminal access poses new challenges to mobile security. In this paper, how to utilize digital certificate, security control and other technologies to authenticate terminals in the scenario of communication module being detached from the computing host is discussed.

Abstract：Data filtering algorithms play a crucial role in the field of big data processing. Data filtering algorithms based on regular expression matching technology are suitable for processing large-scale complex data due to their powerful feature expression capabilities. However, the traditional regular expression matching process is serial matching, resulting in low performance that cannot meet the needs of modern data processing. To address the issue of low performance in traditional regular expression matching, an accelerated regular expression matching algorithm based on multithreading and state prediction is proposed, named μFA. This algorithm performs character value comparison based on vector instructions to obtain the number of trusted characters that can be skipped directly. Simultaneously, it utilizes multithreading acceleration and state prediction techniques to achieve segmented matching processing of strings. By delimiting dangerous character regions, it determines the correctness of the final matching results for each segment. Experimental results show that the throughput is 10.12 to 91.36 times higher than the original DFA algorithm and 1.08 to 2.97 times higher than the ßFA algorithm.

Abstract：Ethernet Control Automation Technology (EtherCAT) is an important application technology of industrial control at present. PLC based on EtherCAT bus can integrate control two-bus system. In the implementation procdure, users need write lots of PLC codes, and asynchronous systems will also have the problem for asynchronous message. In order to solve this problem, a configuration modeling and graphical debugging technology of the two-bus device equipment based on logic configuration software is proposed. The technology includes message synchronization function between asynchronous systems, two-bus protocol analysis and state machine establishment, on-line real-time debugging function and visual interactive interface design function points.

Abstract：Eddy current testing is a method of identifying the properties of a material by detecting the tiny eddy current signals induced in a metal conductor. When using eddy current testing technology for metal classification, the weak signal may be completely overwhelmed by noise, making it very difficult to extract the effective signal. To address this practical problem, a metal classification system based on a lock-in amplifier was designed according to the relevant detection method. In the design, the weak test signal is amplified with low noise and correlated with two orthogonal reference signals, then collected after low-pass filtering circuit, and the collected signal is calculated to achieve detection of the weak signal. The experimental results show that this system has high accuracy and sensitivity, greatly improving the signal-to-noise ratio, achieving rapid classification of metal materials, with a classification accuracy of up to 93%.

Abstract：A miniaturized X-band dual-channel radio frequency(RF) transceiver module is designed by using system in package(SiP) technology, which integrates devices based on a variety of process. Receiving channel and transmitting channel are integrated into a cavity, and time-sharing control for transmitting and receiving is achieved. Double-cavity structure is used and different cavities are vertically interconnected through the ball grid arrays to reduce the module size significantly. The size of SIP module is 21 mm×16 mm×3.8 mm. The main measured technical specifications of the modules are as follows: The P-1dB of receiving channel ≤-10 dBm, receiving signal gain 30～35 dB, isolation degree of the receiving channel ≥55 dB, receiving noise figure ≤8 dB; gain of transmission channel 10～12 dB, maximum output power of transmission channel ≥12 dBm, second and third harmonic suppression ≥60 dBc, clutter rejection ≥55 dBc, the module can function properly within the temperature range of -55～+85 ℃. The measured results are basically agreed with the simulation results.

Abstract：This article designs a transmitter system for unmanned surveillance radar, aiming to achieve efficient detection and accurate monitoring of the surveillance area. The transmitter system uses FMCW (Frequency Continuous Wave) radar technology, and its core is a linear continuous wave signal. In view of the specific needs and characteristics of the unmanned monitoring field, the scheme determination and indicator analysis of the transmitting module were carried out, and the power amplifier and antenna were simulated and designed. On this basis, the structures of the power amplifier and antenna have been effectively improved. Finally, a physical test of the system was conducted, and the results reached the expected transmission power of 7 W, proving that the proposed design solution meets the actual design and application requirements and verifies the feasibility in this field.

Abstract：The unsteady airflow will cause the distortion and delay change of the signal propagation path, resulting in the delay and distortion of the received satellite signal. This will lead to increased ranging and direction finding errors in the positioning system, which will affect the precision and accuracy of the positioning results. Therefore, a new multi-satellite anti-interference hybrid positioning method under the influence of unstable airflow is studied. Preprocess the unstable turbulence data at high altitude, and verify the turbulence data authenticity quickly. The flow field description equation is constructed based on the momentum conservation theorem. The multi-satellite motion equation is established by using the carrier phase observations and satellite antenna phase center coordinates, and the anti-interference hybrid positioning model is constructed. Global Navigation Satellite System (GNSS) technology is introduced to solve the multi-satellite hybrid anti-interference model by using the signals of multiple satellites to calculate the receiver's position, speed and time. The experimental results show that, for static targets, the research method has a small deviation in the positioning of the target N, S and E. When using the mixed positioning of near-earth satellite, medium altitude satellite and geosynchronous satellite, the mean square value of positioning error is always lower than 0.002 cm. For the dynamic target, the method is not disturbed by clutter, and the target is stably located. Moreover, the proposed method has lower localization delay and higher localization efficiency.