Abstract:As a cross-domain operational domain spanning both traditional and non-traditional security fields, electromagnetic space has become a strategic commanding height in the competition for national security. The capability of electromagnetic space security system is regarded as a core variable reshaping the pattern of warfare and determining the right to victory. The development and evolution of the concept of electromagnetic space security were systematically reviewed, and the " three-dimensional" leap characteristics displayed in its development process were summarized. A system architecture of electromagnetic space security and its essential connotations were proposed based on multiple perspectives. The current development status and technical challenges of electromagnetic space security were expounded. Some suggestions with multi-level and highly targeted were proposed for the development of the electromagnetic space security technology system. It provides theoretical and practical references for safeguarding national electromagnetic space sovereignty and electromagnetic space security.

Abstract:UAV (unmanned aerial vehicle) swarms hold broad application prospects in both military and civilian domains, and are developing rapidly. Communication and networking technologies tailored for swarms are crucial for enabling their formation systems and synergistic effectiveness. To this end, key technologies, research status, and development trends in UAV swarm communication and networking were systematically reviewed. Specifically, typical application scenarios and networking requirements of UAV swarms were analyzed. Key challenges and corresponding solutions were thoroughly discussed at the physical layer, data link layer, and network layer, respectively. Furthermore, leveraging the cyber-physical fusion characteristics of UAV, the deep coupling effects among communication, computation, and control, along with their joint optimization methods, were analyzed and summarized. Regarding intelligent empowerment, the intelligent architecture integrating " intent understanding, environment adaptation, and resource scheduling" as a trinity, alongside the latest research progress, were examined. For engineering applications, critical issues such as high-frequency band communication hardware optimization and antenna lightweighting were explored. The prospects for integrating UAV swarms into future integrated space-air-ground networks and the directions for next-step development were outlined. This review aims to provide references for top-level planning and design, as well as for future research directions regarding key technological breakthroughs in the field of UAV swarm communication and networking.

Abstract:To systematically summarize the research progress of attention mechanisms and radar image intelligent processing, this paper reviewed the development context of attention mechanisms and elaborated on typical models, represented by channel attention, spatial attention, self-attention, and hybrid attention. On this basis, the innovative applications of various attention mechanisms in radar image intelligent processing were discussed from the perspectives of tasks such as radar image preprocessing, target detection, image segmentation, and target recognition. To verify the application performance of attention mechanisms, a comparative analysis of different attention mechanisms was conducted on the basis of radar image target detection tasks. Future research directions were discussed from four aspects: attention mechanism interpretability, efficient attention mechanism design, attention mechanism optimization for multimodal fusion, and attention mechanism design in foundation models.

Abstract:High dynamic range imaging aims to recover the luminance and color information of real-world scenes, thereby overcoming the common problems of highlight saturation and shadow detail loss in conventional sensor imaging. It has been extended to applications such as autonomous driving and virtual reality/augmented reality. However, artifact removal in dynamic scenes remains a central challenge. To address this issue, the related datasets and evaluation metrics were systematically reviewed, the major research advances were comprehensively summarized, and the inherent causes of imaging model deficiencies and the current technical bottlenecks were further analyzed. It also compared and analyzed the performance differences among existing state-of-the-art methods from the perspectives of model generalization ability, computational complexity, and inference time. Building on recent development trends, three levels of important research directions were further identified, namely fundamental core challenges, key performance optimization, and frontier technology exploration, with the aim of providing a useful reference for both academic research and engineering practice.

Abstract:A systematic review of hyperspectral underwater target detection under complex water conditions was presented from three perspectives: imaging mechanism, characteristic modeling, and algorithm design. Starting from the underwater hyperspectral imaging mechanism, existing methods were categorized into five groups: spectral prediction, spectral restoration, band selection, pixel classification, and feature construction. Their differences and connections were compared in terms of mechanism-consistent modeling, distortion correction, representational robustness, and interpretability. Analysis show that current methods exhibited distinct characteristics in prior dependency, information utilization, and cross-scene adaptability, and the technical approaches are evolving from mechanism-oriented analysis toward mechanism-data synergy, as well as the integration of generative modeling and feature construction. On this basis, the major challenges in environmental adaptability, reliability modeling, and generalization are further summarized, and future directions are discussed, including differentiable physical modeling, uncertainty characterization, and cross-scene generalization mechanisms.

Abstract:The existing backoff mechanism of the SPMA(statistic priority-based multiple access) protocol relies on static function models and has single dimension of optimization parameters, making it unable to adapt to dynamic transmission and multi-priority requirements in UAV ad hoc networks. To address this issue, the dynamic decision-making process of node selection for backoff time in the SPMA protocol was modeled as a Markov decision process, and an intelligent backoff strategy based on the DDQN(double deep Q-network) was innovatively proposed. This strategy comprehensively consideres factors such as service priority, thresholds, and channel load, and adopts the DDQN algorithm to select backoff time within a finite and discrete action space. Simulation results show that, compared to traditional binary exponential backoff strategies and logarithmic function-based backoff strategies, the proposed strategy can reduce the transmission delay for low-priority services by up to 33.3%, increase the initial backoff success rate by 18%, and effectively improve the transmission success rate and adapt to the variation of network scale well.

Abstract:With the growing demand for nanometer-scale precision in scientific research and industrial production, compliant nanopositioning technology capable of generating and controlling motion with nanometer-level precision has emerged as a key enabler for high-end equipment. The positioning range and resonant frequency determine the workspace and dynamic response of nanopositioning systems. However, there exists an inherent trade-off between them, making it difficult to optimize both simultaneously. This paper focused on the compliant nanopositioning technologies for achieving long range and high resonant frequency. From the perspectives of system configuration and key performance indicators, the intrinsic mechanism underlying the trade-off between positioning range and resonant frequency was introduced. Furthermore, the efforts and practices of researchers worldwide in addressing the trade-off, including the design of nanopositioners based on different actuation methods, as well as related control issues and control strategies, were reviewed. The current research status was analyzed, and future development trends were discussed, providing useful references for the advancement of high-end equipment and precision instruments in our country.

Abstract:With the growing demand for robotic operation in unstructured environments, fixed-configuration robots are increasingly limited in adaptability, fault tolerance, and task versatility. Modular self-reconfigurable robots, composed of standardized homogeneous or heterogeneous modules, can reorganize their topology to achieve morphological and functional reconfiguration. Since existing studies are commonly classified by geometric configuration or connection topology and therefore do not fully reflect the essential differences in reconfiguration mechanisms, this paper reviewed modular robots from the perspective of reconfiguration principles. They were categorized into four groups: non-self-reconfigurable, mobile self-reconfigurable, pose-transformation-based (translation/rotation) self-reconfigurable, and joint-motion-based self-reconfigurable systems. The mechanical principles, representative prototypes, control characteristics, and application scenarios of each category were systematically summarized, and their respective advantages and limitations in compactness, mobility, control complexity, energy efficiency, and environmental adaptability were compared. This review provides a reference for configuration design, strategy selection, and application optimization of future modular self-reconfigurable robots.

Abstract:To address the critical challenges where traditional materials in aerospace and high-end advanced equipment are approaching their physical limits and struggling to meet the stringent SWaP (size, weight and power) requirements, the frontier of 2D (two-dimensional) materials is the primary focus. This paper systematically elucidated their application potential, derived from their atomic-level thickness and quantum confinement effects. By comprehensively reviewing the latest advancements in five core areas—stealth and electromagnetic shielding (survivability), high-performance sensing and detection (perception), lightweight protection and anti-corrosion (defense), high-efficiency energy and power (logistics/support), and quantum technology and information security (computing)—it revealed the intrinsic correlations and mechanisms connecting microscopic properties to macroscopic performance. Furthermore, the key bottlenecks restricting the engineering implementation of 2D materials were analyzed, including wafer-scale high-quality fabrication, long-term stability in extreme environments, and the standardization of testing and evaluation. Based on this analysis, and incorporating emerging technologies such as AI (artificial intelligence)-assisted design and heterostructure stacking, an outlook was presented for achieving multi-functional integration and intelligent systems based on 2D materials towards the development of next-generation smart equipment. This review aims to provide theoretical support and forward-looking insights for securing a strategic technological edge in the future.

Abstract:Two-dimensional semiconductors are identified by the international roadmap for devices and systems as key candidate materials for future sub-nm nodes, owing to their atomic-scale thickness, smooth surface without dangling bonds and capability to suppress short-channel effects. Focusing on the current status of the full-chain development of two-dimensional semiconductors from basic materials science to system-level integration, the intrinsic physical advantages over traditional silicon-based materials and the progress in preparation processes were systematically analyzed. The latest progress and technical bottlenecks of core process modules including contact resistance engineering, gate dielectric integration and device architecture evolution of two-dimensional semiconductor transistors were reviewed in detail. Meanwhile, the development trajectory from early single-transistor verification to large-scale integrated circuits was traced comprehensively, and the collaborative challenges among materials, processes and design during the integration process were analyzed. The unique potential of two-dimensional semiconductors in emerging paradigms such as in-memory sensing and computing, neuromorphic computing and van der Waals heterogeneous integration is further discussed.

Abstract:Satellite clusters, as a distributed collaborative spacecraft system, possess significant application value in areas such as Earth observation, on-orbit assembly, and deep space exploration. Given the constraints of the dynamic space environment and limited on-orbit computing resources, the primary challenge is to devise an effective technique for cluster trajectory planning to ensure the successful execution of cluster tasks. Based on typical satellite cluster systems both domestically and internationally, related application scenarios and developmental tendencies were summarized. The development status of satellite cluster trajectory planning methods was comprehensively elaborated. From the perspectives of Euclidean space and manifold space, the advantages and disadvantages of existing methods were discussed. Starting from recent popular machine learning techniques, the development status of satellite cluster trajectory planning methods that combine deep learning and reinforcement learning with traditional approaches was introduced. Finally, the challenges were encapsulated, and future research avenues were anticipated.

Abstract:To meet the experimental requirements for developing wide-speed range aircraft, research on a supersonic continuous variable Mach number wind tunnel with a Mach number range of 3 to 4.5 was conducted. A scheme of a continuous variable Mach number nozzle was proposed, followed by the design of the supersonic continuous variable Mach number wind tunnel and the calibration of its flow field. Based on P-M (Prandtl-Meyer) expansion theory, a nozzle with single-degree-of-freedom adjustment was designed to achieve continuous variable Mach number. Numerical calculations and calibration experiments were adopted to verify the variable Mach number scheme. Numerical calculation results show that the flow field of the variable Mach number nozzle is uniform at different Mach number. The nozzle centerline Mach number calibration gave the Mach number root mean square, maximum deviation and error. The exit Mach number-expansion angle relation fits P-M theory. The supersonic continuous wave-damping variable Mach number wind tunnel achieves the design goal of continuously changing the Mach number by controlling the nozzle rotation angle while ensuring a high-quality experimental flow field, providing a simplified scheme for experimental research equipment of supersonic variable Mach number flow.

Abstract:In data engineering, the enhancement of processing effectiveness for real-world massive raw data, characterized by being multi-source, heterogeneous, and high-noise, via the application of artificial intelligence methods is currently regarded as a research hotspot. Based on the general research framework of data engineering, the latest research progress in intelligent data engineering methods was systematically reviewed in accordance with the design of three key stages: data cleaning, data linking, and data discovery. Additionally, the principles and effectiveness of the methods related to each key stage were analyzed in detail. Furthermore, an outlook on future research in data engineering is provided in combination with the development trends of intelligence.

Abstract:In response to the significant challenges faced by artificial intelligence techniques, particularly deep learning, in terms of interpretability, heavy data dependence, and limited robustness, the theoretical methodologies and recent advances of PL (prototypical learning) were systematically reviewed. By clarifying the fundamental concepts and mathematical formulations of prototypical learning, a prototype generation framework was established that encompasses statistical machine learning, deep feature-driven representations, and semantic representation perspectives. Prototype-based mechanisms for single-modal and multimodal data augmentation and fusion were analyzed, and the underlying rationale for overcoming data quality bottlenecks was elucidated. Particular emphasis was placed on the effectiveness of prototypical learning in interpretable deep neural networks, fuzzy rule-based reasoning, causal abduction, and time-series analysis. Furthermore, the evolutionary dynamics of prototypical learning across interdisciplinary domains, including generative learning, capability enhancement of large-scale models, and graph learning, were explored. By synthesizing the distinctive advantages of prototypical learning in representation efficiency and logical transparency, its critical role as a key enabling technology for constructing trustworthy and efficient artificial intelligence systems was highlighted. Finally, future development trends of prototypical learning were discussed, including directions related to generative artificial intelligence, collaboration with large models, and sustainable learning.

Abstract:CGF(computer generated force) is the core component of military simulation systems. Traditional modeling methods suffer from bottlenecks including rigid knowledge representation, scarcity of high-quality samples, insufficient modeling of decision complexity, and lack of behavioral evolution capability. Large language models provide a new paradigm to address these issues. This paper systematically clarified the enabling paths of large models from three dimensions: data and knowledge enhancement, decision intelligence generation, and capability iterative evolution. Focusing on five key modules—perception, decision-making, action, role, and memory—this paper elaborated on the large-language-model-based CGF decision-making behavior modeling framework, sorted out the technical implementation routes and representative research achievements of each module, and summarized key technical characteristics and application status. Furthermore, future research directions were proposed from five aspects: decision real-time performance, decision quality, decision fidelity, decision evaluation system, and decision risk control. The findings can provide a systematic reference for intelligent CGF research and the intelligent upgrading of military simulation.

Abstract:Modernization of social governance is an important component of modernizing the national governance system and governance capacity. With the integration and development of location-aware technologies such as mobile communications and satellite positioning have generated massive mobile big data with spatiotemporal identifiers, providing an important opportunity for high-precision, real-time, and scientific transformation of social governance. This paper summarized the 5V +5C characteristics and main sources of mobile big data, reviewed research progress in mobile big data-driven human mobility pattern mining and mobile network model construction, and systematically examined its typical applications and frontier advances in key social governance scenarios, including poverty identification, economic assessment, epidemic prevention and control, and emergency response. On this basis, this paper further analyzed the limitations of mobile big data in terms of representativeness, data bias, privacy protection, and usage compliance, and also discussed future research prospects in multi-source data fusion and spatiotemporal semantic alignment, privacy-preserving computation and real-time response in disaster scenarios, artificial society generation, and large model-driven digital simulation.

Abstract:With the rapid expansion of image data, large-scale image retrieval faces increasingly stringent efficiency requirements. Deep hashing is a key research direction in this field by mapping high-dimensional features into compact binary codes, thereby simultaneously enabling deep semantic learning and efficient image retrieval. Existing methods can be classified into three categories according to the extent of supervision utilized: unsupervised, weakly supervised, and fully supervised. Specifically, unsupervised methods mine latent semantic information from unlabeled data by modeling intrinsic data structures; weakly supervised methods extract effective supervisory signals from noisy or incomplete user-provided tags; and fully supervised methods rely on complete class labels to accurately model semantic relationships. The core ideas and representative achievements across these three categories were systematically reviewed, and comprehensive comparisons of retrieval performance for representative methods were conducted on multiple mainstream datasets. Moreover, despite significant progress, deep hashing still confronts substantial challenges in adapting to dynamically arriving data and achieving effective collaborative modeling in cross-modal scenarios. Future research should prioritize incrementally scalable hashing via continual learning, cross-modal hashing leveraging pre-trained models and so on, thereby promoting deep hashing toward greater efficiency, scalability, and real-world applicability.

Abstract:CIL (class-incremental learning) aims to enable models to maintain discriminative ability on previously learned classes while incrementally acquiring new ones, a process in which catastrophic forgetting often occurs. This paper provided a comprehensive survey and analysis of CIL and its development trends. It clarified the definition of CIL and distinguished it from other incremental learning settings. Mainstream approaches were categorized and summarized from five perspectives: memory replay, parameter and optimization constraints, model prediction calibration, model architecture design, and transfer of pre-trained models. In addition, the commonly used evaluation metrics and datasets of CIL were reviewed, and its applications in typical vision tasks such as image generation, object detection, and semantic segmentation, as well as in emerging areas including video understanding and 3D vision were summarized. Finally, the future research directions of CIL were prospected.

Abstract:Robotic grasping perception is a fundamental prerequisite for autonomous manipulation and embodied intelligence. The technical paradigm is undergoing a profound shift from analytical methods based on explicit geometric modeling to intelligent perception frameworks driven by data-driven learning and enhanced semantic reasoning. Research on robotic grasping perception was systematically reviewed along the lines of paradigm evolution. The evolutionary process was described through three progressive stages: analytical geometry-driven methods, visual data-driven methods, and semantic understanding and reasoning enhancement. Representative algorithms and key technical pathways for each stage were examined and analyzed. Through a comparative analysis of input modalities, data requirements, generalization ability, and task adaptability across different paradigms, the advantages and limitations of various methods in unstructured environments were summarized. Furthermore, the evolution of grasping datasets from planar benchmarks to large-scale comprehensive data was systematically traced, and the quantitative evaluation system composed of task reliability and proposal accuracy was analyzed. Prevailing challenges, including sim-to-real transfer, inference efficiency, cross-modal information fusion, and the extension to complex tasks, were identified. Future development trends that integrate embodied foundation models with dexterous manipulation were discussed to provide references for building general-purpose robotic grasping systems with high generalization performance and robust task comprehension.

Abstract:In the inference process of mixture of experts models, matrix operators constitute the primary performance bottleneck, with those in the attention module and expert computation being particularly time-consuming. Although existing approaches have extensively optimized matrix operators on GPUs, the substantial differences between GPU and CPU architectures in memory hierarchy and compute units make these optimizations difficult to transfer directly to CPU platforms. To address this limitation, FlashMatrix was introduced as a matrix-operator optimization scheme tailored for CPU equipped with advanced matrix extensions. FlashMatrix incorporates an efficient data layout transformation strategy that avoids additional memory-access overhead caused by layout conversions, and employs a carefully designed micro-kernel for matrix multiplication that achieves an optimal compute-to-memory ratio through effective register reuse. Experimental results show that, compared with the state-of-the-art CPU matrix-computation library oneDNN, FlashMatrix delivers an average 2.5 × speedup. For end-to-end inference performance, FlashMatrix achieves a speedup of approximately 1.2×.

Abstract:APN (almost perfect nonlinear) functions, renowned for their optimal differential properties, have become a research focus in the field of cryptographic functions. This paper systematically reviewed the research progress of APN functions: first, it summarized the general methods for generating APN function examples; second, it refined the construction techniques of existing infinite families of APN functions and clarifies their specific constructions; third, it introduced the equivalence classification results of APN function examples and infinite families; fourth, it combed through the research conclusions on the cryptographic properties of APN functions, such as permutation property, algebraic degree, and nonlinearity; fifth, it reviewed some applications of APN functions in coding theory and combinatorial design; finally, the research prospects of APN functions were prospected. Currently, the construction of APN functions is still dominated by quadratic ones, and no infinite families of polynomials with higher algebraic degree have been found. Major challenges, such as the "big APN problem", remain unsolved. Future research may focus on constructing APN polynomials with non-classical Walsh spectra, discovering APN polynomials with higher degree, among others, and exploring their applications in coding theory and combinatorial design.

Abstract:A novel neutron source with high brightness, short pulse, and compact-size--LDNS (laser-driven neutron source) offers considerable promise for a wide range of applications, including non-destructive material testing, neutron imaging, neutron resonance spectroscopy, and nuclear astrophysics, positioning it as a vital supplement to conventional neutron sources. Beginning with an overview of the production methods for LDNS, the fundamental principles were outlined behind various production mechanisms. By reviewing the historical development and recent progress of LDNS, the substantial potential of it for both fundamental research and practical applications was highlighted. Diagnostic techniques and its principles for characterizing key parameters of LDNS were systematically introduced, with special emphasis on recent advances at home and abroad in measuring neutron yield, energy, pulse duration, and source size. Basing on above, this review proceeds to identify technical challenges in the development of LDNS and concludes with a perspective on future research directions.

Abstract:RLG(ring laser gyroscope) was pivotal to the technological revolution transitioning from platform inertial navigation to strapdown inertial navigation in the field of inertial guidance and navigation. To this day, RLG remained the dominant gyroscopic technology in the global high-end inertial sensor market. Firstly, the theoretical origins and feasibility concepts of RLG were traced. Then, the development and application history of RLG technology in China and abroad was systematically reviewed across four stages: initial research, technological breakthroughs, practical advancements, and mass production applications. RLG was invented in the 1960s, with the United States pioneering core key technologies. Building on a dual-track industrial model for commercialization and military-civilian use, its equipment adaptability and high reliability drove a revolutionary advancement in inertial technology from mechanical gyroscopes to optical gyroscopes, achieving widespread application in both civil aviation and military domains. China's RLG research began in the 1970s, gradually overcoming challenges in foundational materials, key technologies, and core processes. It achieved a historic leap from imitation to independent innovation, establishing a complete RLG industrial chain and ultimately achieving full self-reliance. Analysis indicates that RLG is a mature technology with proven performance and reliability, still very robust at the middle and high-end,and is becoming more and more compact as manufacturing advances. Future development priorities for RLG technology will include improving precision through quantum enhancement principal innovation, reducing costs via AI-enabled automated manufacturing, and minimizing size, weight, and power consumption through micro-nano integrated photonic technology.

Abstract:To promote technological innovation in ultrafast spectroscopy and broaden its interdisciplinary applications, this paper provides a systematic review of the current development status and challenges in ultrafast spectroscopic techniques. The fundamental definition of ultrafast spectroscopy was illustrated. The principles of key techniques including ultrafast pulse sources, pump-probe spectroscopy, two-dimensional coherent spectroscopy, and near-filed optics were introduced. By combining research examples such as charge transfer in organic photovoltaics, electron transfer in molecular chemical reactions, exciton fine energy level structure in quantum-confined materials, and lipid molecular photoswitches, the application scopes and unique advantages of various spectroscopic techniques were discussed in detail. The emerging opportunities and future challenges in the development of ultrafast spectroscopy were also summarized.