Admission control governs Quality of Service (QoS) in cognitive IoT networks (CIoTNs) in which secondary nodes opportunistically share spectrum with preemptive primary users. The optimal policy is of threshold type, but the threshold table is indexed by the 2N fully occupied channel configurations, so both model-free learning and on-chain enforcement scale poorly in N. We show that this table has low intrinsic complexity: across 55 randomised continuous-time Markov decision process (CTMDP) instances, a monotone step function of the aggregate secondary-user drain rate σ(n) carrying only 1.6–2.3 distinct levels reproduces 81–87% of the 2N table entries exactly, the remainder erring by a single queue slot. We give a quasi-static argument for why σ(n) is the right scalar summary. Using this regularity as an inductive bias, an index-pooled Q-learning read-out reduces discounted policy-value loss by up to 4.4× (1.55% to 0.36% at N=4, six seeds) at the largest sample budget, but adds variance and was worse than the per-configuration baseline in one of eight cells; we therefore also specify a deployment gate and report it as untested. We further report a negative result: projecting learned value differences onto the concave cone guaranteed by the structural theorem is inert at N=4 and yields at most a 0.22 percentage-point gain at N=3, because recovery is limited by configuration coverage rather than by shape violation. Finally, we implement Service Level Agreement (SLA) enforcement as a Solidity contract and measure it on a local Ethereum Virtual Machine. Index compression cuts policy-installation gas by 241.8× at N=16 (48.79M to 0.20M gas) and keeps installation in one transaction, but raises the per-decision cost by 14.4–18.4k gas; it is therefore a feasibility mechanism for large N and for frequently re-committed policies, not a uniform improvement.

Intelligent Intrusion Detection Systems (IDS) routinely report accuracy above 99% on standard benchmarks, yet real‑world deployment remains limited, an operational divergence we formalize as the Performance–Adoption Gap (PAG). This Preferred Reporting Items for Systematic Reviews and Meta‑Analyses (PRISMA) compliant systematic survey integrates 153 empirical IDS studies (2009–2025), an extended UTAUT2 adoption survey of 300 security professionals, and the distributed‑systems literature to examine this gap from a sociotechnical perspective. We contribute: (1) a transparent six‑stage PRISMA pipeline with reproducibility artifacts; (2) a Six‑Axis Sociotechnical Taxonomy that extends four technical IDS dimensions with two empirically grounded human‑factors axes; (3) a quantitative meta‑analysis showing a mean accuracy of 97.1% but a median minority‑class recall of 0.648 and an average cross‑dataset accuracy drop of 18 percentage points; (4) a formal definition of the PAG as a divergence measure; and (5) a ten‑stage Sociotechnical IDS Architecture that maps detection, explanation, and analyst‑interaction stages to distributed edge–cloud and federated deployment scenarios. The analysis highlights scalability constraints, including inference latency, horizontal partitioning, and model‑synchronization overhead, that directly affect cluster‑based IDS deployments. Finally, we outline a prioritized research roadmap, identifying Retrieval‑Augmented Generation augmented (RAG-augmented) explainability as a high‑leverage direction for bridging technical performance with operational adoption in distributed and federated IDS environments.

Swarm robotics has emerged as a transformative paradigm for accomplishing complex tasks through the coordinated operation of large numbers of simple robots. As swarm sizes scale from tens to hundreds of agents, effective control strategies become increasingly critical to prevent congestion, maintain system throughput, and ensure safe coordination. This review surveys the state of the art in swarm robotics control, with particular emphasis on coordination mechanisms and congestion management. Four major control paradigms are examined: centralized trajectory planning, reactive collision avoidance, spatial partitioning, and learning-based adaptive methods. Beyond these established approaches, the paper explores emerging hybrid frameworks, including Centralized Training with Decentralized Execution (CTDE) architectures, Graph Neural Network (GNN)-based coordination, and hybrid rule-learning integration. For each paradigm, a formal mathematical analysis is provided, covering computational complexity, convergence properties, and stability guarantees. To support rigorous and reproducible evaluation, this paper proposes a standardized assessment framework comprising mandatory performance metrics, scalability benchmarks, and systematic ablation protocols. Finally, open challenges are identified and future research directions outlined, with the aim of advancing the development of robust and deployable swarm robotic systems.

The proliferation of Internet of Things (IoT) devices has introduced significant security challenges. Resource-constrained devices face sophisticated threats but lack the computational capacity for advanced security analysis. This study investigates optimal security task allocation in Cognitive IoT (CIoT) networks. It specifically examines when IoT devices should process security tasks locally or offload them to Mobile Edge Computing (MEC) servers. The problem is formulated as a Continuous-Time Markov Decision Process (CTMDP). The study demonstrates that the optimal offloading policy has a threshold structure. Security tasks are offloaded to MEC servers when the offloading queue length is below a critical threshold, k & lowast;. Otherwise, tasks are processed locally. This structural property is robust to changes in MEC server configurations and threat arrival patterns. It ensures an optimal and easily implementable security policy under the exponential model. Theoretical analysis establishes upper bounds on the performance of AI-based security controllers using the same models. The results also show that standard model-free Q-learning algorithms can recover optimal thresholds without any prior knowledge of the system parameters. Simulations across multiple reinforcement learning architectures, including Q-learning, State-Action-Reward-State-Action (SARSA), and Deep Q-networks (DQN), confirm that all methods converge to the predicted threshold. This empirically validates the analytical findings. The threshold structure remains effective under practical imperfections such as imperfect sensing and parameter estimation errors. Systems maintain 85% to 93% of their optimal performance. This work extends threshold Markov Decision Process (MDP) analysis from classical queuing theory to the context of CIoT security offloading. It provides optimal and practical policies and model-free algorithms for use by resource-constrained devices.

This paper presents a novel Deep Q-Learning (DQL) framework for multi-robot navigation that addresses the critical problem of target congestion in swarm robotics systems. The framework employs a Centralized Training with Decentralized Execution (CTDE) paradigm, where a single DeepQ-Network (DQN) agent learns a global coordination policy during offline training, and individual robots execute learned policies using only local Radio-Frequency Identification (RFID) based observations during deployment. Scalability refers to decentralized execution, while training is conducted under fixed swarm configurations. Unlike traditional vision-based obstacle-avoidance methods, we propose an RFID-based information-processing module combined with Robot Constraint Rules (RCR) that provide explicit heuristic guidance for flow regulation, while the DQN learns to optimize its application. The Probabilistic Finite State Machines (PFSM) and RCR modules define structured safety constraints, while coordination strategies are learned within this constrained action space. The system is modeled as a Discrete-time Markov Decision Process (DTMDP), which is subsequently integrated with PFSM to capture both reactive behaviors and learned coordination strategies. The PFSM framework enables the formalization of human-traffic-inspired behaviors (such as following, lane-changing, etc.) that arise within the constrained decision-making framework through reward-driven learning rather than hard-coded rules. The framework demonstrates empirically stable convergence under structured constraints. Simulation results demonstrate that our method significantly reduces the average robot workload by up to 66% and improves system efficiency by 51% compared to baseline reactive collision-avoidance methods without congestion control, while maintaining target occupancy density below 50% and ensuring deadlock-free operation even with 100 robots.

In recent years, tuned liquid dampers (TLDs) have emerged as a focal point of research due to their remarkable potential for structural vibration mitigation. Yet, progress in this field remains constrained by an incomplete understanding of the fundamental mechanisms governing sloshing-induced loads in liquid-filled containers. Aqueducts present a distinctive case, as the capacity of their contained water to function effectively as a TLD remains uncertain. To address this gap, the present study investigates the generation mechanisms of sloshing loads under non-resonant cases through a two-dimensional (2D) computational fluid dynamics (CFD) model developed in ANSYS Fluent. The incompressible Reynolds-Averaged Navier-Stokes (RANS) equations are solved, while the Volume of Fluid (VOF) method captures the evolution of the air-water interface. Turbulent flow behavior is modeled using the RNG k-s approach. The ensuing results reveal the dynamic characteristics of the horizontal force (Fh) and the fluctuating component of the vertical force (Fvf). Fh is predominantly governed by the inertia of the deep-water region and its phase varies coherently with the aqueduct's acceleration. With increasing excitation amplitude (A) and frequency (f ), the contribution of deep-water inertia to Fh intensifies markedly, accounting for 82.6-92.1% of the total horizontal load at an excitation amplitude of 0.15 m and frequencies of 1.0-1.6 Hz. The extreme values of F Fvf arise primarily from asymmetric static pressures induced by free-surface fluctuations, which are further amplified when wall gaps appear at large amplitudes (A >= 10 cm) and high frequencies (f >= 1.4 Hz). Unlike resonant cases dominated by free-surface resonance, non-resonant sloshing loads are principally driven by deep-water inertia and motion-induced surface asymmetry.

Research on characteristics of tsunami forces acting on conical islands and acting on cylindrical structures placed on top of a conical island, as well as the generation mechanisms of the tsunami forces, is limited. The total tsunami force acting on conical islands and its generation mechanisms are studied in cases with different initial water depths and relative wave heights. The results indicate that the peak values of the positive and negative tsunami forces, which increase with both initial water depths and relative wave heights, have almost the same magnitude. The initial submerged state of the conical island does not change the variation trend of tsunami force peaks with relative wave height. Then the tsunami forces on the structures located at different positions on the conical island are compared and analyzed, and the results show that when the conical island is not submerged at the initial state, the front cylinder has the largest peak value; when the conical island is submerged at the initial stage, the rear cylinder has the largest peak value.

Pounding tuned mass damper (PTMD), an improved passive damping device based on the traditional tuned mass damper (TMD), dissipates vibration energy by pounding when the main structure vibrates fiercely and dissipates vibration energy in a traditional TMD mode when the main structure vibrates slightly. PTMD has been demonstrated to be effective and robust to reduce vibration of the main structure in air and in still water in previous studies. However few studies focus on PTMD damping performance to the vortex-induced vibration of the main structure in water flow. This study proposed a new form PTMD with double L-shaped cantilever beams and streamlined mass blocks, and also designed the test model to validate the damping performance of the PTMD. Results show the test model works well both in air and in water flow, and the new form PTMD can reduce vibration of the main structure when the main structure vibrates in air. The new form PTMD is demonstrated to be effective and robust to reduce vibration when the main structure vibrated by vortex shedding in water flow. Furthermore, comparison reveals that the new form of PTMD has better damping performance when working in water flow than in air.

Solitary wave is often used to simulate tsunami propagating in deep water and breaking solitary wave is often used to simulate tsunami bore propagating in shallow water or on land. The breaking solitary wave force on box-girder, which has been widely used in bridge engineering in coastal areas of China, receives few attentions. This study aims to investigate characteristics and generation mechanism of breaking solitary wave force on box-girder numerically. A numerical wave flume with a 1:20 slope was built firstly, then the solitary wave generation ability, wave deformation and wave breaking on the slope, as well as wave force calculation precision, are validated. The water depth 0.6 m, the slope gradient 1:20 and the distance between slope top and box-girder 2.0 m remain unchanged, while the wave height and clearance changes in different cases. The time histories of horizontal force and vertical force on box-girder can be divided into three and four stages respectively according to their characteristics. The surface of box-girder is decomposed into a series of panels to facilitate exploring tsunami bore force generation mechanism. Results show horizontal force is dominated by static pressure on upstream vertical panels and vertical force is mainly contributed by static pressure on upstream horizontal panels and on panels in the chambers. Tsunami bore overtopping the box-girder deck impacts the top panel vigorously and results in the peak value of negative vertical force.

Underwater communication is crucial for subsea engineering. Currently, although a variety of available technologies are available for underwater communication, almost all of them are highly dependent on the conditions of the subsea environment. Hence, underwater communication is still a challenging issue. This article develops a new stress wave communication method that can be implemented along steel pipes for underwater data transmission. In this article, we analyze stress wave propagation along pipelines, and the corresponding channel response shows the features of dispersion and frequency selectivity. Thus, a suitable carrier frequency is then selected and binary phase-shift keying is applied to encode information into the carrier stress wave. The proposed single-input-single-output communication system is implemented on a steel pipe and experiments are conducted both when in air and seawater. Results demonstrate the validity and stability of the proposed method. The communication data rate of the method can reach 1000 b/s with a low bit error rate. The proposed method shows great promise for enabling underwater communication with lower loss and longer range than conventional underwater communication methods.

With the explosive development of the computer vision technology, more and more vision-based inspection methods enabled by unmanned aerial vehicle technologies have been researched on the crack inspection of the sundry concrete structures. However, because of the limitation of the low-cost unmanned aerial vehicle hardware, whose cost is around US$500, most of the vision-based methods are difficult to be implemented on the low-cost unmanned aerial vehicle for real-time crack inspection. To address this challenge, in this article, a new computationally efficient vision-based crack inspection method is designed and successfully implemented on a low-cost unmanned aerial vehicle. Furthermore, to reduce the acquired data samples, a new algorithm entitled crack central point method is designed to extract the effective information from the pre-processed images. The proposed vision-based crack detection method includes the following three major components: (1) the image pre-processing algorithm, (2) crack central point method, and (3) the support vector machine model–based classifier. To demonstrate the effectiveness of the new inspection method, a concrete structure inspection experiment is implemented. The experimental results indicate that this new method is able to accurately and rapidly inspect the cracks of concrete structure in real time. This new vision-based crack inspection method shows great promise for the practical application.

Remote Laboratory as a Service (RLaaS)-Frame is a web application framework and an Service-Oriented Architecture built upon the cloud-computing technology. Based on the RLaaS model, a new cloud-based framework, namely RLaaS-Frame, is proposed to flexibly and rapidly deploy the remote laboratory system. As a standardized framework, the RLaaS-Frame can be extended to other disciplines, such as physics, chemistry and biology. This RLaaS-Frame will accelerate the adoption of remote laboratory technology and benefit online education, academic research, and industrial applicat ions. To demonstrate the feasibility of the RLaaS-Frame, a Wiki-based remote laboratory platform and the mobile-optimized application framework have been integrated into the RLaaS-Frame. In order to combine the advantages of Virtualiztion Management and Docker, an optimized solution to support the RLaaS-Frame is proposed. The RLaaS model originates from promoting the traditional remote laboratory technology to the new cloud-based remote laboratory technology.

Although the high-resolution materials can improve the resolution of the conventional time-reversal imaging (TRI) algorithms, they also limit the applications of TRI. In this paper, a new TRI algorithm with high-resolution is presented. Since the proposed algorithm utilizes multiple time reversal operation steps to improve resolution, it can realize high-resolution without invoking any high-resolution materials. The results show the resolution of the proposed algorithm is superior to that of the conventional TRI.

As regular inspection for the bolt connection in inaccessible areas is difficult and costly, computer vision technology provides a suitable noncontact approach for real-time bolt looseness detection as an alternative to inspection approaches. However, computer vision still suffers from various impracticalities. In this paper, a new vision-based bolt looseness detection method is designed and implemented with the bolt images acquired by a camera at arbitrary positions around the bolts. The new method includes the perspective transformation of original images acquired, identification of bolt positioning with the convolutional neural network digit recognition, detection of bolt rotation angles using Hough transform line detection, and density-based spatial clustering of applications with noise. To demonstrate the effectiveness of the new method, an experiment with bolted connections is setup. The experimental results demonstrate that the new method can accurately detect the looseness of the bolts in the bolted connection.