The pattern of global regional conflicts is undergoing a disruptive transformation. Publicly available reports indicate that the overwhelming superiority of unmanned aerial vehicles (UAVs) over traditional armored and air defense systems in the Nagorno-Karabakh conflict, the first combat employment of the Kinzhal hypersonic missile, the saturation attacks by low-cost unmanned swarms interwoven with artificial intelligence (AI)-assisted decision-making in the Russia-Ukraine conflict, and the offense-defense cost crisis triggered by Iran’s large-scale combined drone-missile strike on Israel collectively reveal a core fact: the air and space combat domain is undergoing a paradigm shift driven by cutting-edge technologies and validated through recent conflicts. The traditional combat paradigm centered on high-performance manned platforms and reliant on information superiority for precision strikes is being fundamentally challenged across three dimensions："speed" "quantity" and "energy." These correspond respectively to the "dimension-reduction strike" of hypersonic weapons against existing air and missile defense early-warning and interception logic, the undermining of traditional defense economic foundations by intelligent unmanned swarms through cost asymmetry, and the new defensive era of "light-speed interception" and near-infinite ammunition heralded by directed-energy weapons. Drawing on recent local conflicts worldwide and current developments in airborne weapon technology among major military powers, this paper systematically analyzes the three core technology clusters driving this paradigm shift, examines their convergence mechanisms and operational impacts, summarizes future development trends, and offers insights and recommendations for addressing emerging air and space challenges in the new era.

Storage and transportation launch canister/container covers, as critical components of launch payload storage and launch systems, directly influence launch safety and reliability, operational response time, and battlefield survivability. The complex and ever-changing modern battlefield imposes increasingly stringent performance requirements on these covers. Composite frangible covers have become a research focus at home and abroad in this field by virtue of their advantages such as light weight, corrosion resistance, high design flexibility, rapid response, and low maintenance cost. Aiming at the integrated design methodology for composite frangible covers of storage and transportation launch canisters/containers, this paper reviews the research progress of composite frangible covers from three aspects: material and process selection, structural design, and numerical simulation. It also summarizes and analyzes the critical design factors for composite frangible covers of different forms, providing reference for optimal design of composite frangible covers.

With the large-scale proliferation of low, slow, and small (LSS) unmanned aerial vehicles (UAVs) on the battlefield, traditional air defense systems have revealed problems such as cost asymmetry and low operational efficiency. How to effectively address the new challenges posed by LSS UAVs has become an urgent issue for militaries worldwide. To address the severe challenges of cost-effectiveness imbalance and generational capability gap facing traditional air defense systems, this paper proposes a new defensive paradigm of unmanned counter-unmanned. The core of this paradigm lies in utilizing friendly unmanned systems to detect, identify, jam, or destroy adversary unmanned platforms. Starting from fundamental concepts, this paper analyzes its unique advantages, discusses key supporting technologies, and reviews the development status of typical counter-unmanned equipment represented by drone interception, aiming to provide reference for research and practice in related fields.

This paper examines the developmental trajectory and underlying logic of U.S. military operational concepts. Employing historical analysis, comparative study, and interdisciplinary research methods, it systematically reviews the evolutionary lineage from "AirLand Battle" to "Joint All-Domain Command and Control" since the 1980s. The analysis reveals that U.S. military operational concepts are undergoing a paradigm shift from "reactive response" to "designing warfare". The core driving forces stem from the interplay of strategic demand pull, technological innovation enablement, and inter-service competition-integration. Accordingly, this paper proposes measures including constructing a "strategy-concept-doctrine" linkage system, strengthening forward-looking planning for "technology-tactics integration," and innovating cross-domain collaborative training models, providing theoretical support for enhancing all-domain operational capabilities.

Against the backdrop of accelerated evolution in intelligent warfare, the U.S. military's joint operational concept system centered on Joint All-Domain Command and Control (JADC2) has attracted widespread academic attention. However, in the process of its dissemination and interpretation, many key concepts face risks of oversimplification, over-technologization, and even misinterpretation. From the perspective of system-of-systems operations, this paper analyzes and differentiates the U.S. military's joint operational concept system. By reviewing its development history and logical framework, this paper innovatively proposes a layered analytical framework centered on the evolution of combat systems: at the top level, it reveals the transformation of operational paradigms from force-centric to intelligence-centric; at the operational level, it categorizes operational concepts across four dimensions—cross-domain, distributed, confrontational, and agile; and at the support level, it focuses on enabling specific missions. This paper also outlines three future trends: intelligentization, multi-domain integration, and resilient self-adaptation, providing reference value for deepening related research.

The United States has conducted research on boost-phase missile defense concepts since the 1960s, with sustained and uninterrupted efforts continuing to the present day, reflecting its consistent emphasis on boost-phase missile defense operations. By systematically reviewing the development trajectory of U.S. boost-phase missile defense systems and examining the progression of major programs from the Anti-Ballistic Missile (ABM) Treaty era to the post-ABM Treaty era, this paper analyzes the primary difficulties confronting U.S. boost-phase missile defense system development. On this basis, lessons and experiences from such development are summarized, yielding several insights for future technological development in this domain.

To address the inability of traditional tree-like air-defense and anti-missile command architectures to cope with new air and space threats such as hypersonic weapons and stealth platforms, as well as problems including difficult combat force coordination and inefficient command relationships, this paper systematically conducts research on structural design and simulation analysis of air-defense and anti-missile operational command architectures. By integrating system-of-systems engineering and complex network theory, a weighted complex network model is constructed, encompassing four categories of core nodes—command and control (C2), information acquisition, information processing, and interception and strike—and three categories of directed interactive relationships: hierarchical command and control, all-domain information sharing, and cross-unit collaborative feedback. A horizontal collaborative flat command architecture suited to networked air-defense and anti-missile operations is proposed. Considering the temporal characteristics of air-defense and anti-missile operations, dynamic evaluation indices are established, covering distance timeliness, synergy, connectivity, and overall effectiveness. Comparative simulation analyses with traditional and weak-center architectures show that the horizontal collaborative architecture effectively shortens average path length while improving clustering coefficient and average node degree compared with traditional architectures. This indicates that its overall network effectiveness, robustness, and survivability are significantly superior to those of traditional and weak-center architectures. The findings provide theoretical guidance and quantitative methodological support for constructing efficient and resilient air-defense and anti-missile command architectures.

To address the dynamic air defense firepower allocation problem of intercepting multiple hostile targets with various types of surface-to-air missile (SAM) weapons in key-point air defense operations, this paper proposes a method based on non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) and an improved multi-attribute decision-making method. Firstly, a dynamic weapon-target assignment optimization model considering multi-stage and multi-mission requirements is established. Based on this, NSGA-Ⅱ is employed to solve the model and generate the Pareto optimal solution set. Then, an improved multi-attribute decision-making method combining grey relational analysis (GRA) and combinative distance-based assessment (CODAS) — GRA-CODAS — is utilized to select the optimal allocation scheme from the Pareto solution set. Finally, a weapon-target assignment simulation experiment is conducted using a set of multi-target air raid cases. Results demonstrate that compared with traditional decision-making methods, the proposed method not only obtains the optimal scheme, but also balances operational requirements including maximizing overall damage effectiveness against enemy targets and minimizing own-side operational costs. The generated weapon-target assignment scheme ensures the accomplishment of the mission to inflict severe damage on all incoming hostile targets while significantly reducing operational costs, proving to be more reasonable and effective.

From the perspectives of command and operational employment, this paper introduces the combat command architecture of the U.S. aircraft carrier strike group (CSG). It examines its command and control (C2) and force employment in strike operations, air and missile defense (AMD), and anti-submarine warfare (ASW). It comprehensively describes how the CSG, under a flat command structure, employs coordinated operational strategies to accomplish various combat missions. Finally, it summarizes the distinctive characteristics and advantages of the CSG command structure, aiming to provide reference for research on command relationships of modern naval formations.

To address the challenge of identifying the importance of heterogeneous nodes in complex combat systems, this paper proposes a critical node identification method based on operational loop effectiveness. The method first constructs an operational loop model, identifies typical operational loop structures via subgraph isomorphism algorithms, and computes loop closure probabilities through Monte Carlo simulation. On this basis, system effectiveness and node importance metrics are derived. Case analysis demonstrates that command and control nodes occupy a central position in the system, and their removal leads to a significant decline in system effectiveness. Comparative analysis indicates that removing nodes identified by the proposed method results in a faster decline in system effectiveness compared with conventional methods, thereby enabling more precise critical node identification. Meanwhile, dynamic node removal experiments further reveal that node importance dynamically evolves with the system state. The findings provide a scientific basis for combat system survivability design and critical node protection.

To address the emerging need for enhancing equipment capabilities through software iteration under future warfare modes, this paper analyzes the state of the art and characteristics of the prevailing software factory research and development model based on development, security, and operations integration (DevSecOps), and introduces the status of advanced software factory construction at home and abroad. Focusing on the unique characteristics of domestic equipment software development, this paper comprehensively elaborates the fundamental approach and key measures for constructing software factories in the equipment domain from five aspects: technical architecture, reuse construction, team organization, process norms, and tool platforms. Finally, combined with practice, it introduces the general process and methods for local construction of the software factory model.

To meet the urgent demand for "detect-to-destroy" capabilities in long-range precision strike under cross-domain joint operations, the core supporting role of space-based information chains is becoming increasingly prominent. To address the issues of rigid resource scheduling and insufficient response timeliness in existing space-based information systems, this paper proposes a real-time construction and resource scheduling method for the space-based information chain. Firstly, the operational requirements for closing the time-sensitive kill chain with space-based information support are analyzed. On this basis, a layered processing architecture for the space-based information chain is constructed, and a core dynamic resource cooperative scheduling algorithm is designed. The algorithm models resource scheduling as a multi-objective optimization problem with time constraints, which is solved using an improved genetic algorithm to achieve intelligent orchestration of space-based resources including reconnaissance, communication, and computing. Simulation results demonstrate that the proposed processing architecture and scheduling algorithm can reduce the average end-to-end time of the kill chain by approximately 50%, and increase the mission success rate from 22% to over 90% in severely contested environments, significantly improving the timeliness, resilience, and resource utilization of the space-based information chain.

To realize threat evasion for hypersonic vehicles with high-fidelity dynamic characteristics under multi-process constraints, this paper proposes an intelligent decision-making method based on safe reinforcement learning (SRL). Firstly, the hypersonic dynamic model and multi-constraint model are constructed. Secondly, an intelligent maneuvering strategy decomposed into longitudinal and lateral dimensions is designed based on vehicle and mission characteristics. Subsequently, the constrained maneuvering decision problem is formulated as a constrained Markov decision process (CMDP). The constraints are incorporated into the classical deep deterministic policy gradient (DDPG) training framework via the Lagrange multiplier method, and a multiplier update rule is designed based on constraint violation frequency, thereby achieving policy network training under multi-process constraints. Finally, the algorithm is tested across multiple scenarios and evaluated for generalization capability. Results demonstrate that the proposed method significantly improves training efficiency, achieving an effective balance between rapid policy learning and constraint satisfaction. The maneuvering strategy enables the vehicle to evade threats while satisfying multiple process constraints, with the success rate remaining above 95% in generalized scenarios.

To improve the trajectory prediction accuracy of high-speed flight vehicles under complex maneuvering conditions, this paper proposes a trajectory prediction method integrating temporal convolutional networks (TCN), bidirectional long short-term memory (BiLSTM), long short-term memory (LSTM), and attention mechanism, addressing the limitations of single neural networks in temporal dependency modeling and prediction stability. The proposed method first employs TCN to extract multi-scale temporal features; subsequently, BiLSTM and LSTM are utilized to collaboratively capture bidirectional and unidirectional dependencies within the sequence. Meanwhile, a multi-head self-attention mechanism is introduced to achieve adaptive weighting and fusion of multi-dimensional features. On this basis, a Bayesian optimization strategy is adopted to perform global optimization of key hyperparameters, effectively overcoming the limitations of traditional manual tuning and further enhancing model performance. Simulation results demonstrate that the proposed method significantly outperforms baseline models in terms of three-dimensional root mean square error (RMSE) across different prediction horizons. Specifically, within a 50 s prediction horizon, the maximum improvement in prediction accuracy exceeds 50%. Meanwhile, the proposed method exhibits favorable computational efficiency and adaptability to different prediction starting points and historical input lengths, providing a feasible technical path for engineering applications.

To address the high complexity of swarm target allocation, this paper constructs an operational intent preference function to solve the problem of inaccurate threat level estimation for different target types, and introduces Bayesian optimization to estimate weight values of threat degree elements in the model, thereby reducing the model's dependence on expert experience. Meanwhile, considering that the traditional particle swarm algorithm is prone to fall into local optima, a dimensionality-reduction particle swarm heuristic algorithm suitable for multi-constrained nonlinear problems is designed to optimize algorithm performance. Simulation results demonstrate that compared with conventional target allocation algorithms, the proposed swarm target allocation algorithm based on intent information significantly improves the effectiveness of operational intent allocation and algorithm stability in complex environments, enhancing the algorithm's environmental adaptability.