Soft errors and aging are two major factors impacting the long-term reliability of integrated circuits, especially for aerospace applications. Double-node upset (DNU) and bias-temperature instability (BTI) are typical soft errors and aging, respectively. First, this article presents a C-elements (CEs) based DNU-recoverable latch, namely the basic latch, without aging mitigation to ensure low overhead. Second, in order to alleviate BTI-induced aging, the BTI-mitigated CEs are proposed so as to construct the aging-mitigated CEs based DNU-recoverable latch, namely ACDRL, for long-term reliability enhancement. By optimizing the key CEs internal structure, the stress time of transistors in feedback loops can be effectively reduced to mitigate the impact of BTI on the latch. Pertinent simulation results demonstrate that the proposed latches have DNU recoverability; the soft error rate increase due to BTI is reduced by roughly 35% for ACDRL compared with the basic latch; the delay of ACDRL is almost not affected, and the area and power increase of ACDRL are limited compared with the BTI-unmitigated latches.

Heterogeneous unmanned swarm systems (HUSSs) have gained extensive deployment owing to their superior collaborative capabilities in novel battlefields. However, their heterogeneity and complexity also pose significant challenges for ensuring mission success, making the accurate evaluation of mission reliability of great importance. Addressing limitations of traditional approaches in handling system heterogeneity, multi-layered structures, and epistemic uncertainties, this study proposes a conditional evidential networks (CEN)-based approach to evaluating mission reliability in HUSSs, to reduce assessment inaccuracy caused by epistemic uncertainty. This study constructs a multi-layered collaborative capability evaluation framework for HUSSs, which comprises the basic capability layer, the collaborative interaction layer, and the mission effectiveness layer. By means of indicator mapping and weight propagation, this framework enables quantitative evaluation of capabilities. Subsequently, elaborating the approach to evaluating state credibility between underlying capability indicators and higher-level capabilities. Building upon this foundation, the integration of CEN with quantitative weighting enables effective evaluation of mission reliability for HUSSs. Finally, the effectiveness of the approach is validated using a heterogeneous unmanned aerial vehicle (UAV) swarm as the research object. The mission reliability of the HUSSs is calculated to be 0.5079.

The reliability of Unmanned Swarm Systems(USSs)represents a critical research domain essential for ensuring the safety and stability of system operations.Given the dynamic nature of mis-sions and environments,it is essential to equip the reliability models of USSs with appropriate evolu-tionary capabilities to enhance the credibility of the evaluation process.However,much of the work in this field is primarily designed for Unmanned Equipment(UE)or components,limiting its applicabil-ity to USSs.This study proposes a multi-agent-based short-cycle reliability evolution model for USSs.An evolution framework for the agent-based reliability model is developed to effectively integrate the evolution elements,evolution timing,and evolution strategies of the USS reliability model.The inter-nal modules and data flow of agents are designed to describe the associated elements of USSs and sup-port model evolution.Moreover,the study emphasizes the Adaptive Adjustment of Failure Propagation Paths(AAFPP)and the Online Addition and Removal of System Agents(OARSA),supported by associated trigger mechanisms and evolution strategies.A case study involving a 6-UAV swarm for surface reconnaissance validates the approach,demonstrating an average of 23 model evolutions per simulation,with AAFPP accounting for 54.56％of adjustments.

Traditional backscatter communication systems typically operate in a single frequency band and suffer severe performance degradation in the presence of strong interference. To address this limitation, this paper proposes a multichannel backscatter communication (MCBC) system that enables data transmission across multiple frequency shifts using a single tag. Unlike prior approach that rely on multiple tags or antennas, our designed system achieves frequency diversity by serially shifting frequencies at the tag side, thereby enhancing robustness without increasing hardware complexity. To further improve communication reliability, we introduce a reliable channel selection algorithm that dynamically evaluates and selects the optimal frequency channel based on real-time bit error rate (BER) prediction. Experimental results demonstrate that the proposed MCBC system achieves a 3.5 dB gain on signal-to-noise ratio (SNR) at the BER level of 10-5 compared to the conventional backscatter systems even under complex interference. The results also show that our scheme maintains stable and energy-efficient communication across diverse environments, including real-world WiFi deployments.

Transportation systems are one of the most common types of complex systems in modern society, and their highly reliable operation is of great significance to factories, cities, and other entities. Selective maintenance strategies are commonly used to flexibly ensure system reliability. However, the continuity of transportation system tasks and the time-varying nature of external demands lead to significant gaps in existing approaches. Furthermore, the conflict between cost and reliability exacerbates the difficulty of planning selective maintenance for the system. Therefore, this study presents a multi-objective optimization method for selective maintenance of transportation systems with time-varying demands, which generates a comprehensive maintenance plan encompassing maintenance timing, nodes, and actions. We first analyze the characteristics of transportation systems under continuous and time-varying demands and proposes a scenario-adaptive reliability measurement method. Then, leveraging the advantages of the rolling horizon control framework in handling continuous problems and considering the time and cost of different maintenance actions, a multi-objective optimization model for selective maintenance is constructed to enhance the system's future performance and reduce costs. Subsequently, a tailored multi-objective optimization method is developed by introducing a multi-population collaborative differential evolution algorithm with a mutation mechanism into the traditional pigeon-inspired optimization algorithm, which avoids premature convergence and improves solution quality. Finally, the feasibility and effectiveness of the proposed method are verified in a coal transportation system.

The existing SLAM (Simultaneous Localization and Mapping) systems often suffer from increased positioning errors, inaccurate map construction, and challenges in real-time multimodal sensor data fusion in dynamic environments. This paper proposes enhancements to the SLAM system using nonlinear optimal filtering and deep learning to improve adaptability in such conditions. The study employs an Unscented Kalman Filter (UKF) for nonlinear state estimation, while deep feature extraction of environmental images is conducted via Convolutional Neural Networks (CNN). Semantic edge detection integrates Fully Convolutional Networks (FCN) and Canny edge detection techniques. The Extended Kalman Filter (EKF) is utilized for multimodal data fusion to optimize positioning accuracy across vision, lidar, and inertial measurement unit (IMU) sensors. Real-time motion estimation is achieved through an event-based camera combined with an optical flow algorithm, enhancing speed and accuracy in dynamic scenes. Experimental results demonstrate that the proposed SLAM system achieves an absolute trajectory error (ATE) as low as 0.067 m across datasets, with over 90

With the rapid advancement of semiconductor technologies, latches become increasingly sensitive to soft errors, especially triple node upsets (TNUs), in harsh radiation environments. In this article, we first propose a high-performance and area-efficient latch, namely, HALTRAV, featuring complete TNU-recovery. The storage portion of HALTRAV consists of 28 interlocked source-drain cross-coupled inverters (SCIs) for complete TNU-recovery with area efficiency and low delay. To mitigate the issue that node-upset-recovery verifications for existing latches highly relies on electronic design automation tools, we further propose an algorithm-based verification method that can automatically verify the node-upset-recovery of latches, which greatly simplifies the reliability-verification flow. Simulation results demonstrate the TNU-recovery of HALTRAV and also show that HALTRAV achieves 40.38%, 8.17%, and 31.89% reduction in delay, area, and delay-power-area product (DPAP) on average, respectively; however; it is at the cost of power as compared to typical latches that are TNU-recoverable. Comparison results also demonstrate the moderate sensitivity of HALTRAV to the impacts of the process, voltage, and temperature (PVT) variations.

The increasing complexity of warfare environments and equipment has significantly raised the workload of the equipment support system, making it more vulnerable. Therefore, it is essential to study the vulnerability of equipment support systems. Based on the analysis of the components of the support system, this paper proposes a multidimensional performance model for the system, which includes the dimensions of equipment availability, personnel reliability, and consumable satisfaction. Furthermore, a vulnerability assessment method is presented. In addition, based on multi-agent simulation theory, a simulation modeling framework for the vulnerability of equipment support systems is proposed. This framework includes the analysis of support mission processes, the agent-based simulation modeling architecture, and the system vulnerability simulation evaluation process. Finally, a simulation calculation is performed using an unmanned aerial vehicle (UAV) formation as a case study, which validates the accuracy and scientificity of the proposed method.

As semiconductor technology advances, radiative-particle-induced soft errors and power consumption are becoming major concerns for digital circuits in aerospace applications. Radiation hardening by design and magnetic tunnel junctions (MTJs) are widely employed to address these concerns. In this paper, a novel latch, called TNURML, that can completely recover from triple-node upsets (TNUs), is proposed. The embedded MTJs provide non-volatility and are compatible with traditional CMOS processes. The TNURML employs a TNU-recovery module as well as a pair of MTJs for backup and recovery operations. Extensive simulations demonstrate the excellent TNU-recovery capability and non-volatility of the TNURML latch at the cost of slightly increased area overhead. The proposed TNURML latch reduces 43.63% of delay and 48.23% of power on average when compared to the state-of-the-art latches.

Uncrewed swarm systems (USSs) are injecting renewed vigor into societal development and economic growth, and their failure modeling is crucial for ensuring the safety and stability of system operations. However, traditional modeling methods are insufficient for describing the cross-layer diffusion characteristics of USS failure. In this context, proposed here is a multi-agent-based failure modeling approach to establish the groundwork for analyzing USS failure. A failure modeling framework grounded in agent-based modeling methodology is developed to efficiently capture and organize the multi-layer failure information of a USS. Expanding on this framework, the interaction mechanism of agents is designed to delineate the operational processes and failure propagation paths of the USS. A dynamic clock failure model, structural functions, and functional constraint matrices are integrated effectively to describe the cross-layer failure behaviors of the components, subsystems, and nodes of the USS. Furthermore, a unified modeling approach is proposed to describe the failure propagation within and between nodes. Finally, a case study of a USS comprising 16 uncrewed aerial vehicles and eight satellites is conducted to validate the effectiveness of the proposed approach. The simulation results reveal that the electric distribution board, voltage stabilizer, and flight control board significantly influence the mission completion.

Unmanned Swarm Systems (USSs) are injecting renewed vigor into societal development and economic growth. However, addressing the modeling of their external shocks remains insufficient. In this context, we propose an agent-based approach to model external shocks, ensuring the reliable and stable operation of USSs. We introduce a novel external shock modeling framework based on the multi-agent modeling and simulation (MABMS). Within this framework, furthermore, we present a unified modeling approach capable of integrating various shock models to enhance both accuracy and efficiency. Finally, a case study is used to verify the effectiveness of the proposed approach.

Swarm systems of unmanned aerial vehicles (UAVs) have emerged as a popular subject of research on the Internet of Things owing to their higher flexibility, efficiency, and reliability than single UAVs. However, little research has been devoted to investigating the vulnerability of UAV swarms, and the traditional network model cannot fully and synchronously characterize their communication-based and mission-based relationships. This study proposes a two-layer multi-edge complex network model to characterize the UAV swarm by considering its status of communication and collaboration for the given mission. The model contains a communication layer and a function layer. In addition, we consider the area of coverage of the UAV swarm, provide the definitions and methods of calculation of three factors influencing its performance, and use them to develop a method to assess its performance for the three typical missions of attack, reconnaissance, and jamming. Furthermore, we propose a framework for the vulnerability analysis of the UAV swarm that can analyze its process of failure and measure its vulnerability. Finally, we use a swarm consisting of 10 UAVs as a case to verify the effectiveness and accuracy of the proposed model.

To adapt to the characteristics of IoT tasks arriving online (i.e., the arrival pattern and time of tasks cannot be pre-dicted), delay sensitivity, and limited processing unit resources, to ensure the completion of tasks with low latency and to ensure the efficient and stable operation of the IoT platform, this paper studies the problem of online task distribution and scheduling in the unit manager of a data-driven Internet of Things (IoT) architecture. The problem is modeled and an efficient online task distribution and scheduling algorithm is proposed, named OnSche. We have validated the performance of the OnSche algorithm using real datasets and conducted a thorough verification. The experimental results show that under different parameter changes, OnSche is consistently superior to the baseline methods.

In the basic Vision transformer (ViT) model, the way to process the image is to cut the image into a certain number of patches as labeled (location coded) tokens, which are inputted to the multi-attention part of the model to extract the information, and then passed through the full connectivity layer to a classifier to classify the image. If all the Tokens are utilized for interaction, the complex ViT structure brings a huge amount of computation, which is a major drawback of the model. The ViT model itself is characterized by global interaction, and Attention is extremely sensitive to global information, so the over-pursuit of interaction with all the Tokens will bring about a sharp increase in complexity and a multiplication of the pressure on the hardware facilities, such as the reduction in the number of Tokens, which will bring about a reduction in the amount of information. If the number of Tokens is reduced, it brings about loss of information and unsatisfactory accuracy. In this study, we take the lightweight ViT model as a starting point, and preprocess the images: the images are chunked and then entered into the coding layer for information extraction respectively, and finally the respective cls labels are processed to get the classification results, which can solve the problem of over-attention to global attention of ViT as well as avoiding the loss of information due to the reduction of Tokens, and ultimately realizing the model's lightweighting. In conclusion, our model (PS-ViT) starts from two aspects, firstly, data enhancement is used to improve the generalization ability of the model; secondly, in the processing of input images, pre-blocking can both change the range of model information interaction and reduce the computation of the model in order to achieve efficient reasoning. Finally, a 40% reduction in computation can be achieved while maintaining accuracy.

As the semiconductor technology continues to advance, integrated circuits (ICs) are becoming increasingly sensitive to soft errors, e.g., double-node upsets (DNUs) and triplenode upsets (TNUs), induced by harsh radiation. In this paper, a low-cost latch design, namely ICLTR, using input-split inverters (ISIs) and C-elements to provide complete TNU recovery, is proposed. ICLTR consists of seven ISIs, seven 2-input C-elements and a clock-gated inverter, and all these elements are interlocked. Simulation results show the complete TNU recovery for ICLTR. The simulation results also show that ICLTR can save 59.5% of the transmission delay, 36.1% of the power consumption and 81.6% of the delay-area-power product (DAPP) on average when compared with the same type of TNU recovery latch designs.

Modern powerful CMOS chips are usually highly integrated and implemented with aggressively shrunk technology nodes. In radiation environment, under charge-sharing mechanism, one particle striking can simultaneously impact multiple nodes causing double-node-upsets (DNUs) and triple-node-upsets (TNUs). In this paper, we propose an Interlocked Dual-circle Latch Design, namely IDLD, with low cost and TNU recovery for aerospace applications. IDLD consists of four transmission gates and twelve 2-input C-elements (CEs) implemented in 22nm CMOS process. Simulation results demonstrate the complete TNU recovery as well as cost-effectiveness for the proposed IDLD latch.

Large pre-trained visual models, such as Vision Transformer (ViT), have shown excellent results in a variety of visual processing tasks. However, the large parameter sizes of these models make them difficult to deploy in applications that require limited resources for fast real-time inference. Existing model compression methods compress pre-trained ViTs into models of the same structure, but the compressed models fail to maximise performance on different tasks and consume significant computational and storage resources to adapt to downstream tasks. Pre-trained ViT models learn rich knowledge from large datasets and require different knowledge for different tasks, so removing task-specific redundant knowledge from pre-trained models is the key to achieve model compression while maximising model performance on different tasks. Based on this idea, we propose a novel ViT compression method, AdaViT. Firstly, a lightweight module is introduced, which adds less than 2% of extra parameters to realise the adaptive tuning of the model to different tasks to show the best prediction, and its effectiveness is proved by experiments. Second, inspired by the idea of knowledge distillation, we propose a new module replacement compression method, which effectively compresses the ViT by gradually replacing the Transformer module in the original ViT, and achieves task-oriented adaptive ViT compression through the combination of adaptive modules and model replacement methods. We evaluate AdaViT on several visual classification tasks and compare it with other ViT compression methods, demonstrating the effectiveness of task-adaptive ViT compression.

Achieving photothermal therapy (PTT) at ultralow laser power density is crucial for minimizing photo-damage and allowing for higher maximum permissible skin exposure. However, this requires photothermal agents to possess not just superior photothermal conversion efficiency (PCE), but also exceptional near-infrared (NIR) absorptivity. J -aggregates, exhibit a significant redshift and narrower absorption peak with a higher extinction coefficient. Nevertheless, achieving predictable J -aggregates through molecular design remains a challenge. In this study, we successfully induced desirable J- aggregation (λ abs max : 968 nm, ϵ: 2.96×10 5  M −1  cm −1 , λ em max : 972 nm, Φ FL : 6.2 %) by tuning electrostatic interactions between π-conjugated molecular planes through manipulating molecular surface electrostatic potential of aromatic ring-fused aza-BODIPY dyes. Notably, by controlling the preparation method for encapsulating dyes into F-127 polymer, we were able to selectively generate H -/ J -aggregates, respectively. Furthermore, the J -aggregates exhibited two controllable morphologies: nanospheres and nanowires. Importantly, the shortwave-infrared J -aggregated nanoparticles with impressive PCE of 72.9 % effectively destroyed cancer cells and mice-tumors at an ultralow power density of 0.27 W cm −2 (915 nm). This phototherapeutic nano-platform, which generates predictable J- aggregation behavior, and can controllably form J- / H- aggregates and selectable J- aggregate morphology, is a valuable paradigm for developing photothermal agents for tumor-treatment at ultralow laser power density.