The rapid development of Unmanned Aerial Vehicle (UAV) technologies has brought non-negligible information security risks. Existing research literature lacks structured and systematic review and summary. From the perspective of key management, this paper reviews representative relevant studies in recent years, constructs a typical conceptual model for UAV key management, and conducts comparative and classified analysis from four dimensions: application scenarios, cryptographic mechanisms, security properties, and evaluation methods. On this basis, it summarizes the core unsolved problems in current theoretical research and scheme design and prospects the key research directions in the future.
To address the critical challenge of ensuring reliable disaster data backhaul under frequent natural disasters and fragile ground communication infrastructures, a backpressure-aware collaborative learning empowered data flow backhaul framework for aerial-ground integrated networks (AGIN) is proposed. First, a data flow backhaul architecture is established by leveraging cooperative networking between unmanned aerial vehicles (UAVs) and ground facilities. Second, a QoS guarantee mechanism based on queue priorities dynamically coordinates resources to ensure low-latency backhaul for high-priority services. Thirdly, a data flow backhaul optimization approach based on backpressure-aware collaborative learning incorporates queue backlog differences into the reward function to prevent accumulation and adds a global conflict penalty to facilitate multi-UAV coordination and mitigate next-hop selection conflicts. Simulation results indicate that the proposed approach enables low-latency, energy-efficient, and high-throughput data flow backhaul under dynamic networks.
The 13.56 MHz RFID system is a short-range wireless communication technology that uses magnetic fields. The Radio Law restricts the intensity of the magnetic field used for communication to a low level, limiting the communication area of 13.56 MHz RFID system. Furthermore, communication performance is easily degraded by the surrounding environment. For some applications, such as paying for all items in a shopping cart at once, it is desirable to expand the communication range. To solve this problem, the authors investigate a simple method for improving the system's communication performance by adding a parasitic coil to 13.56 MHz RFID system. In this paper, they evaluate the communication performance when one or two parasitic coils are placed between a table-type 13.56 MHz RFID reader/writer and a tag.
With the increasing frequency of extreme weather and natural disasters, traditional ground communication networks struggle to meet emergency demands due to inadequate coverage and slow recovery. Complex terrain and inefficient resource allocation also lead to spatial-domain coverage holes, impacting fault location accuracy and emergency repair efficiency. To address these issues, a low-latency service bearing architecture for space-air-ground integrated networks is constructed, which accounts for spatial-domain coverage holes. Secondly, a low-latency service bearing and access control method for space-air-ground integrated networks is proposed considering spatial-domain coverage holes. Furthermore, a low-latency unmanned aerial vehicle (UAV) relay dispatch and channel joint selection algorithm based on coverage hole-aware Q-matching is proposed. Simulation results demonstrate that the proposed method significantly mitigates spatial-domain coverage holes and reduces average queuing delay for critical service bearing.
In recent years, videos have become an integral part of daily life and there is an enormous volume of video content. As a result, finding specific videos or scenes within this vast amount of content has become a difficult task. In order to enable quick scene recognition by visualizing both video content and relationships between frames, the authors introduce a method for visualizing the subjects and their motion by arranging video frames in a spatial layout based on their correlations. This method calculates frame correlation matrix by applying Principal Component Analysis (PCA) and arranges the video frames in a spatial layout based on the principal component scores. From the visualization results by scatter plots and 3D space, they found that the scatter plot of minimal frames and 3D space arrangement of video frames were more effective than traditional video playback for scene recognition and workload reduction.
This article explores methods for the automatic generation and visualization of animation plots, aiming to improve the efficiency and quality of animation creation. The proposed system leverages neural network technology, incorporating a model for scene generation and a compatible visualization engine. Through the neural network, the system can autonomously generate cohesive and innovative animation scenes, while the visualization engine converts these elements into realistic 3D models, offering real-time, intuitive feedback to animators. Experimental validation was conducted to assess the quality of the generated animations and visualization effects, incorporating user feedback. The results show that the system excels in coherence, creativity, and visual appeal, with users particularly praising the realism and ease of interpretation of the animations. The automatic generation and visualization methods presented offer new possibilities for advancing the efficiency and creative potential of animation production.
As power systems evolve, diverse power services increase bandwidth demands, posing challenges like variable transmission loads and slow data transfer. Routing reconstruction dynamically adjusts paths, balances loads, and reduces delays, ensuring reliable power service data. However, current technologies lack global state awareness, integrated risk-delay optimization, and efficient algorithms. This paper introduces a unified model and risk evaluation framework for both data and power transmission networks. Considering the enduring operational demands of the transmission network, a joint minimization strategy is devised which focuses on minimizing both transmission delays and risks. Furthermore, a coupled Q-learning methodology for collaborative network operation is introduced, which sets routing priorities and resolves differences via cost to enhance routing results. Simulations validate that the proposed methodology drastically decreases transmission risks and delays.
Patients with heart failure require long-term or frequent hospitalization, which places a heavy burden on medical resources. The six-minute walk test is a simple and cost-effective method for assessing aerobic capacity and endurance. It does not require specialized personnel or sophisticated equipment and involves recording walking distance, blood pressure, heart rate, and oxygen saturation level within a fixed time interval. In this study, we provided patients with heart failure with a smart watch and an application tool, enabling them to perform the six-minute walk test at home. The application allowed patients to upload their test data on cloud storage, which were examined using feature correlation analysis, regression modeling, and other techniques. The goal was to explore the most influential features that correlated with outpatient records and provide effective reminders to patients with heart failure to monitor their health status during their daily lives, which would reduce medical resource consumption.
Fine-tuning large language models (LLMs) for sports injury prevention and treatment in resource-constrained environments poses significant challenges due to memory demands and growing size of data. This paper proposes an efficient full-parameter fine-tuning approach based on Gradient Low-Rank Projection (GaLore) to reduce memory usage. Further, a data augmentation strategy for sports injury prevention and treatment is utilized to finetune a question-and-answer (Q&A) model with 0.5B parameter on consumer GPUs with 24GB memory. Experiment results show that the proposed method enhanced by GaLore is superior to SOTA methods such as low-rank adaptation (LoRA) in terms of convergence accuracy, training time, memory consumption, and indicators of BLEU-4 and ROUGE-2. Meanwhile, the empirical effect of injury prevention Q&A cases indicate that Qwen2-0.5B-Instruct trained by the proposed method have obvious advantages in professional knowledge understanding and overcoming hallucinations.
Bionic visual perception technology captures optical signals from gas insulated metal enclosed switchgear (GIS) partial discharge by mimicking the biological visual system, achieving real-time detection and recognition of partial discharge phenomena under complex electromagnetic environments. However, existing technologies often consider only a single spectrum and do not account for differentiated thresholds for various discharge phenomena, affecting imaging accuracy. This paper proposes a multi-spectral bionic visual perception optimization method for GIS. First, a multi-spectral bionic visual perception framework is constructed. Second, an optimization problem is formulated to maximize the average imaging accuracy of all GIS discharge phenomena. Next, a two-stage event-driven deep Q-network (DQN) optimization method is proposed, learning the optimal light intensity change threshold through two-stage closed-loop feedback, including offline and online learning. Finally, the superior performance of the proposed method is validated through simulations.
With the continuous development of power systems, power grid work order data provides a large amount of data support for power facilities. However, the long sequence of text in power grid work orders brings challenges to computer processing and analysis. In order to improve the accuracy of power grid work orders, this paper proposes a long sequence text vectorization method for power grid work orders based on leader grey wolf optimization-bidirectional long short-term memory and bidirectional encoder representations from transformers (Bi2-BERT). Firstly, a long sequence text feature extraction method is proposed, which performs the long sequence text feature extraction by BiLSTM, to improve the global optimization performance. Secondly, a text vectorization method for power grid work orders based on increment-stock bilevel-driven BERT is proposed to improve the model learning efficiency and solve the catastrophic forgetting problem. The results show that the proposed algorithm has high accuracy and convergence speed in long sequence text vectorization of power grid work orders.
With the circulation of massive electric measurement data, data anomaly caused by security attacks imposes security risks on reliable operation of smart grid. Long short-term memory (LSTM) based data circulation monitoring and security risk anomaly evaluation has been intensively studied. However, some issues remain unsolved, including learning overfitting and large prediction error. In this paper, we investigate fuzzy learning to infer the abnormal level of security risk. In particular, an adaptive grey wolf optimization-LSTM-fuzzy petri network (AGWO-LSTM-FPN) based electrical measurement data circulation monitoring and security risk anomaly evaluation algorithm is proposed. Specifically, AGWO is utilized to optimize LSTM parameter updating and improve traffic prediction accuracy. Furthermore, FPN is combined with multi-dimensional monitoring indicators to enhance anomaly level evaluation. Simulation results illustrate the excellent performance of AGWO-LSTM-FPN.
This paper combines the dance 3D space simulation sequence diagram with video motion recognition technology, filters, denoises, grays and background removal the collected dance video images, analyzes the motion characteristics of people in the sequence diagram, uses support vector machine to learn and train 3D space models, classifies and recognizes people's dance movements, and extracts 3D-SIFT and optical flow characteristics of various areas of human body. Form a three-dimensional space simulation sequence diagram, reduce and normalize the extracted features, get the feature vectors of various characters, and input them into the classifier to realize the recognition of dance movements. The results show that the combination of 3D-SIFT and optical flow can realize the dynamic change of human static information, the illumination invariance of SIFT features can make up for the illumination sensitivity of optical flow features, and the optical flow features can solve the instability problem of determining the key points of SIFT features.
With the deployment of internet of things-based monitoring terminals, massive electric measurement data are transmitted and shared among different grid company sectors. Consortium blockchain based consensus process has been adopted to enhance data traceability. However, previous studies face challenges including single-dimensional indication and lack of dynamic trustworthy degree update mechanism. Therefore, this paper proposes a node trust evaluation and selection algorithm based on improved fuzzy Petri network (IFPN). Firstly, the multi-dimensional node trust evaluation and selection framework for practical Byzantine fault tolerance (PBFT)-empowered electric measurement data sharing is constructed. On this basis, IFPN is proposed to dynamically evaluate measurement data node trust, and nodes participating in the PBFT consensus process are selected based on the Top-N algorithm. Further, the trustworthy degree of IFPN is dynamically updated based on the fuzzy deviation to improve the accuracy. Finally, the superiority of the proposed algorithm has been verified through simulation.
With the maturity of AR technology, AR is more and more used in various industries. from the perspective of the application of AR technology at this stage, the emergence of AR technology improves the visual effect in image animation, simulates the real scene in the viewing process, and makes the computer advanced human-machine interface with the basic characteristics of equipment interactivity and conceptualization, spread to the scope of film and television. As a creative emerging industry, influence animation has huge development space and prospects, and it also carries the functions and tasks of inheriting traditional culture. Combined with the specific situation of Lingnan culture, this paper introduces the combination of AR technology and Lingnan culture in visual, sound and interactive design, and extracts its creative elements. Compared with the simple data recording method, the difference between the two is 41.22%. Through specific ideas and programs, it provides a new digital method and idea for the protection and research of Lingnan culture.
Coronavirus disease 2019 (COVID-19) pandemic has produced a global health crisis that has had a deep impact on the way we perceive our world and everyday lives. Not only the spread rate of contagion and patterns of transmission endangered our sense of security, but the safety measures put in place to contain the spread of the virus also require social distancing. In this context of physical threat, social and physical distancing, the role of the different mass media channels and social media in lives of individual, social and societal levels cannot be underestimated.
Coronavirus disease 2019 (COVID-19) pandemic has produced a global health crisis that has had a deep impact on the way we perceive our world and everyday lives. Not only the spread rate of contagion and patterns of transmission endangered our sense of security, but the safety measures put in place to contain the spread of the virus also require social distancing. In this context of physical threat, social and physical distancing, the role of the different mass media channels and social media in lives of individual, social and societal levels cannot be underestimated.
Virtual power plant (VPP) plays an important role in improving the balance and regulation abilities of new power system. The safe and reliable operation is support by the VPP end-to-end communication network with differentiated multi-service bearing capability. For the requirement of unified and standard VPP end-to-end networking scheme, the VPP service communication metrics, as well as the communication network architecture of VPP aggregation and control are analyzed. Then, a multi-dimension hierarchical VPP end-to-end network evaluation index system is put forward. In addition, an end-to-end VPP network evaluation method considering differentiated time-sensitive and granular requirements of multiple services is proposed. Finally, the suitability analysis results of various end-to-end networking schemes and multiple services with differentiated time-sensitive and granular requirements are given, which plays a guiding role in establishing a unified standard VPP end-to-end networking scheme.
In this work, the authors propose a new biometric authentication system on mobile devices, enhancing security at these terminals and preserving user privacy. The proposed system uses a method of extracting strong features from minutiae with refinement of the method with regard to the further elimination of false minutiae by the calculation of geometric information (orientations and distances between minutiae) to obtain true terminations and stronger bifurcations facilitating the recognition of individuals. A series of tests carried out using a recognition and authentication application allowed us to achieve a false rejection rate of 13.81% and a false acceptance rate of almost zero (0.021%). The authors also propose a security model using hash functions and a random number to make the recognition system revocable, more difficult to compromise and thus reducing the risk of usurpation.
Cellular vehicle-to-everything (V2X) communication is essential to support future diverse vehicular applications. However, due to the dynamic characteristics of vehicles, resource management faces huge challenges in V2X communication. In this paper, the optimization problem of the comprehensive efficiency for V2X communication network is established. Considering the non-convexity of the optimization problem, this paper ulitizes the markov decision process (MDP) to solve the optimization problem. The MDP is formulated with the design of state, action, and reward function for vehicle-to-vehicle links. Then, a multiagent deep Q network (MADQN) method is proposed to improve the comprehensive efficiency of V2X communication network. Simulation results show that the MADQN method outperforms other methods on performance with the higher comprehensive efficiency of V2X communication network.
