The impact of input light polarization on the demodulation results in the 3×3 coupler-based current demodulation scheme is analyzed. By utilizing the features of a single circularly polarized optical fiber, this influence can be eliminated.
In interconnected networks, networks with smaller strong diameters tend to have better fault tolerance. This is because when the strong diameter of the network is small, the vertices are more closely connected and the path of information propagation is shorter, making it easier for the network to maintain communication and functionality in the event of an attack or failure. Whereas, when the strong diameter is large, the vertices are more distantly connected and the information propagation path is longer, which increases the vulnerability of the network and makes it more susceptible to attacks or failures. Therefore, strong diameter is one of the key metrics to measure and optimize the fault tolerance performance of the network. In this paper, a strong product network of cycles and paths is constructed by strong product and the strong diameter of this network is investigated. Specifically, we determine the minimum strong diameter of the strong product of even cycles and paths, and the upper and lower bounds of the minimum strong diameter of the strong product of odd cycles and paths. In addition, we give a lower bound on the maximum strong diameter of the strong product of cycles and paths, and present a conjecture on the maximum strong diameter of the strong product graph of cycles and paths.
Algorithms are commonly used to generate patrol routes. However, in cultural heritage protection patrols, traditional route planning methods often neglect the specific security requirements of cultural artifacts. Thus, there is a need for a specialized route planning algorithm and evaluation method tailored for cultural heritage protection to enhance security. This study proposes a method to optimize cultural heritage protection patrols, utilizing the Analytic Hierarchy Process (AHP) for comprehensive evaluation of pre-patrol route planning, actual patrol effectiveness, and patrol costs.During the pre-patrol route planning phase, to prevent the predictability of patrol routes and mitigate risks to cultural artifacts, this study employs an enhanced ant colony algorithm to generate multiple routes. It evaluates factors such as route differentiation and coverage of key patrol points, determining multiple patrol schemes based on these evaluations. To assess actual patrol effectiveness, the study analyzes patrol personnel GPS trajectories and extracts critical information. This optimization approach employs multidimensional quantitative analysis of the entire patrol system to enhance the efficiency and security of cultural heritage protection patrols.
This paper investigates the use of the diff - iteration method based on spectral interferometry for measuring the micro - dispersion of optical devices (optical fiber, lens, etc.). Using an optical frequency comb, based on the phase demodulation of dispersion interference spectrum, by employing the carefully designed diff - iteration method to solve the dispersion curve at any position and any order. Our approach is proficient in precisely measuring micro - dispersion across a broadband spectrum, without the need for cumbersome wavelength scanning processes or reliance on complex high - repetition - rate combs, while enabling adjustable resolution. The efficacy of the proposed method is validated through simulations and experiments. The approach ensures high precision, while maintaining a simple system structure, with realizing adjustable resolution, thereby propelling the practical implementation of precise measurement and control - dispersion.
In this paper, the problem of constructing an optimallinear code is considered. Four new codes with parameters [18,12,6], [21,16,5] over $\mathbb{F}_{9}$ , [20,15,5], [19,14,5] over $\mathbb{F}_{7}$ are determined. Their generator matrices and weight distributions are presented and these four linear codes are new and have better parameters than the known results.
Frequency hopping sequences (FHSs) play a crucial role in wireless communication, while Reed-Solomon (RS) codes hold a good algebraic structure. In this paper, two classes of FHSs are constructed from punctured RS codes, both containing new parameters. The two classes of FHSs both meet the lower bound of peak nontrivial Hamming correlation given by (3), and under certain constraints both classes can meet the upper bound of set size given by (2). Moreover, the methods in this paper can be regarded as a generalization of some previous methods.
Faster-than-Nyquist (FTN) technology introduces additional inter-symbol interference (ISI), which significantly impacts the bit error rate (BER) performance of FTN multiple input multiple output (MIMO) optical wireless communication (OWC) system. To address this issue, an end-to-end (E2E) FTN-MIMO OWC system utilizing a long short-term memory (LSTM) autoencoder has been proposed. This approach effectively eliminates the influence of ISI and achieves complete signal recovery. Simulation results demonstrate that our proposal improves BER performance to different degrees under various conditions compared to conventional systems using maximum likelihood (ML) mothed. Specifically, with acceleration factors of 0.9 and 0.8, our proposal demonstrates an improvement in BER performance by 1.6dB and 2.2dB respectively, effectively mitigating the complex mixed ISI caused by FTN shaping and atmospheric turbulence channel.
The appearance of large language models (LLMs) and related products has generated widespread attention and lively discussions in both industry and academia. Given their extensive applications and remarkable achievements in numerous fields, LLMs have rapidly emerged as a frontier research topic that is actively explored by numerous scholars. This paper delves into the application and current research status of LLMs in the field of programming tutoring. Firstly, we outline the applicability of LLMs in programming tutoring, emphasizing the new opportunities brought by their powerful language understanding and generation capabilities to programming education. Subsequently, through empirical research, we validate the effectiveness of LLMs in correcting programming errors. The research results have confirmed the effectiveness of LLMs in bug localization and program repair, demonstrating the significant potential and practical value of large language models in Programming Tutoring Systems. By continuously optimizing model performance, we can further advance the development of programming education.
Diabetic Retinopathy (DR) is a severe complication of diabetes that significantly impacts vision and can potentially lead to blindness. The effective diagnosis and timely treatment of DR is crucial preventing vision loss. Convolutional Neural Networks (CNNs) have emerged as a prominent tool in the field of medical image analysis for this purpose. This paper proposes a novel ensemble residual deep learning model that integrates parallel convolutional filters with shallow convolutional neural network enhanced by a channel attention mechanism for the classification of DR. The proposed model leverages the strength of residual learning and attention mechanism to improve feature extraction and classification performance. To validate the effectiveness of the model, this study conducted extensive experiment on DR data, where the model demonstrated robust diagnostic capabilities. The results were remarkable, achieving accuracy of 97.69 % , F1-score of 97.77 % , sensitivity of 97.02 %, specificity of 98.18%, precision and recall both at 97.77%. Furthermore, the model attained an area under curve of 98%, highlighting its potential to serve as an effective supplementary tool for the analysis of DR images. These findings underscore the model's potential to assist healthcare professionals in the early and accurate diagnosis of DR, ultimately contributing to better patient outcome.
The missing data issue is common in practical traffic data collection, making accurate completion critical for various applications such as transport planning and route navigation. A lot of approaches have been proposed to address this issue, but many challenges remain, especially in designing efficient online completion algorithms that adapt to the streaming nature of traffic data. In this paper, to jointly aggregate multidimensional correlations, we utilize streaming tensors to represent traffic data and propose a dynamic tensor completion model based on Tucker decomposition to estimate missing values. Furthermore, we develop an efficient online algorithm to track the optimal core tensor and factor matrices, thanks to an alternating optimization framework. Experiments are conducted on two public traffic datasets to evaluate our method against some state-of-the-art baseline approaches. The results show that our method achieves completion performance closer to that of non-dynamic methods in a running time close to that of dynamic methods.
Emergency response decision-making is a core element in addressing unexpected safety incidents at various cultural heritage sites. Existing emergency plans often face challenges such as dispersed information, lack of integration, and high decision-making costs. To address these issues, this study proposes a reasoning method for emergency response measures to unexpected cultural heritage safety incidents based on the Generalization-Generation Pattern. This method first applies grounded theory, combined with large models, to efficiently and accurately extract key information from unstructured texts to construct a structured knowledge graph, promoting information sharing and supporting the reasoning of emergency response measures and optimizing resource allocation. Finally, by using the trained Generalization-Generation Pattern, we can predict response measures for cultural heritage safety incidents and analyze common patterns in similar situations, providing decision support for similar events. In practical applications at demonstration sites, this pattern has successfully provided reasonable response recommendations for specific safety incidents, demonstrating its potential in improving emergency response efficiency and reducing decision-making costs.
Frequency hopping has good anti-interference ability. However, the high-speed movement between Low Earth Orbit (LEO) satellite and ground will lead to a large Doppler frequency of the received signal carrier. With the switching of the frequency hopping frequency point, the Doppler frequency will change abruptly, which will affect the subsequent signal demodulation. To solve this problem, a carrier tracking loop algorithm based on normalized Doppler rate estimation is proposed. The loop separates the frequency hopping agility factor from the phase and persistently locks the continuous Doppler rate. When the signal is detected to have high power interference at some frequency points, the algorithm performs delay locking to shield the disturbance to the tracking loop. The simulation results and analysis show that compared with the traditional phase-locked loop, the proposed algorithm can be well applied to frequency hopping carrier tracking in large Doppler dynamic and harsh interference environment.
In order to address the issue of statistical heterogeneity in wireless federated learning, we proposed a novel client selection method. We introduced a new metric called client quality, which takes into account the local data volume of clients and the testing accuracy of the previous round's model. Based on this, we proposed a client selection method called Client Quality-based Client Selection (CQCS) to improve the performance of wireless federated learning. Results show that the proposed CQCS client selection method effectively enhances the global accuracy of the model. Specifically, compared to the baseline algorithms FedAvg, FedProx, and Per-FedAvg, the model average accuracy improved by 6.7%, 4.8%, and 3.1%, respectively, on the MNIST dataset. On the Cifar10 dataset, the model average accuracy improved by 3.4 %,3.7 % , and 3.9%, respectively. Similarly, on the Fashion-MNIST dataset, the model average accuracy improved by 3.1 %, 3.8%, and 4.2%, respectively.
Physical feature extraction of underwater acoustic targets is crucial for target detection and identification, which is able to transform the physical properties of targets into quantifiable feature parameters. Structure extraction is performed by statistically analyzing some characteristics in the target echo signal, such as the number of highlights, intensity distribution, time interval. Traditional highlight structure extraction methods rely on manually designed signal processing techniques. However, these methods often face challenges in maintaining extraction accuracy under complex environmental noise. Deep learning methods effectively solve these problems by constructing multilayer neural networks to automatically learn feature representations from the data. To address the above issues, this paper proposes a feature extraction method for active echo highlight structure based on a Transformer network. The underwater acoustic target-echo highlight model has been designed. Furthermore, a deep learning model was constructed based on Transformer to achieve patch embedding and feature extraction of the input samples. Moreover, a joint loss function, including the regression loss of echo highlight structure, the regression loss of target scale, and the time extension correction loss, was designed to improve the convergence effect of the model and the accuracy of feature extraction. The model training results show that the prediction error of the number of highlights can be controlled within 5, and the prediction error of the target scale is controlled within 12 meters. In addition, the model is tested on actual underwater acoustic target echo data, and achieves a good prediction performance of the highlight structure on real target echoes.
Multiple Origin AS (MOAS) outsourcing mitigation is a defense countermeasure against prefix hijacking, during which the mitigation AS attracts and redirects the hijacked traffic to the legitimate origin AS by announcing the hijacked prefix. In order to enhance the effectiveness of mitigation, this paper proposes a node selection method for the mitigation AS. Specially, it takes multiple dimensions of AS topology into considerations: the hierarchical type, the Internet eXchange Point (IXP) access and the neighbor relationships of the AS. Experimental results show that the proposed method can achieve better mitigation effect against prefix hijacking compared other methods.
This study evaluates the usability of the Mobile Automated Fingerprint Identification System (MAFIS) mobile application within Philippine law enforcement using the PACMAD model. The objectives include assessing effectiveness, efficiency, learnability, memorability, error frequency, and cognitive load. A mixed-methods approach was employed, involving questionnaire surveys and usability tests with 21 participants from the Crime Laboratory - Fingerprint Identification Division. Results indicate highly positive perceptions of the MAFIS application across all usability attributes, with weighted mean scores consistently falling within the “Highly Acceptable” range. Qualitative findings highlight user satisfaction with interface design and features, alongside desired improvements for enhanced functionality. These insights inform the optimization of the MAFIS application to support law enforcement efforts, contributing to more effective crime prevention and public safety in the Philippines.
The Ga 2 O 3 /GaN/Ga 2 O 3 n-p-n double-junction ultraviolet (UV) photodetector was successfully fabricated by radio frequency magnetron sputtering. The device exhibits stable responses under 265 nm UV irradiation at -3 V bias, with a responsivity of ~130 mA/W, an external quantum efficiency of ~61%, a detectivity of ~8.21×10 8 Jones, which are 3.4, 3.4, and 8 times that of the GaN/Ga 2 O 3 p-n type device, respectively. The energy band diagram shows that due to the large valence band offset of Ga 2 O 3 and GaN, large number of holes accumulate at the Ga 2 O 3 /GaN interface closer to the cathode, which reduce the valence band offset at the junction to promote the transmission of photogenerated carriers and improve the device performance.
As an important means of obtaining information about the Earth's surface, remote sensing mapping is widely used in fields such as military, environmental monitoring, resource surveying, and disaster monitoring. Consequently, remote sensing image matching has always been a prominent area of research in image matching. Optical images closely align with human visual perception of the world, but they are susceptible to environmental influences, which can result in information loss or operational failure. LoFTR, as a dense matching model, is limited by its complexity and is not suitable for most equipment conditions. Therefore, this paper proposes a sparse remote sensing matching method based on the LoFTR descriptor, which utilizes RIFT feature points to extract the corresponding LoFTR feature map to obtain the corresponding feature descriptors, and performs registration by calculating cosine similarity. In the experiments, multiple sets of Optical and SAR remote sensing images were used as test data. The experimental results demonstrate that the proposed method effectively retains the advantages of LoFTR, reduces the model's complexity and parameter scale, and achieves good results in remote sensing image registration for both Optical and SAR types, with sub-pixel registration accuracy.
Power Internet of Things (PIoT) is a crucial technology for smart grids, and it also plays a momentous role in current electric power data transmission and sharing. Currently, an increasing demand for cross-domain interactions requires various devices belonging to different domains to cooperate, which makes traditional identity authentication mechanisms unsuitable to ensure information security and privacy during communication. In addition, centralized power IoT systems will also face concerns of single point of failure, data tampering, etc. To tackle these issues, we design TrustPIoT, a blockchain-assisted authentication scheme for cross-domain PIoT, which achieves access control between cross-domain devices while ensuring identity anonymity. We exploit an on-chain and off-chain storage mechanism to ensure the consistency of the data, by introducing a cloud server to coordinate with the blockchain network. To verify its validity, we also conduct experiments on Hyperledger Fabric to evaluate the performance of TrustPIoT.
With the increasing number of transmission and distribution networks, end-to-end optical layer slicing is expected to provide a more efficient transmission method for power communication networks. However, the coexistence of multiple communication systems such as synchronous digital hierarchy (SDH) and optical transport network (OTN) in current power communication networks poses challenges for end-to-end optical layer slicing. In this paper, against the backdrop of power communication networks where SDH and OTN communication systems coexist, we analyze the differences between SDH and OTN communication systems and consider two types of requests: bandwidth-constrained and latency-constrained. We construct an integer linear programming (ILP) model for slicing mapping and design a slicing request mapping strategy based on consistent transmission bandwidth and another slicing request mapping strategy based on minimal transmission latency. Surrounding the proposed content, we design simulation evaluation metrics for the two strategies and compare their performance with traditional strategies. We evaluate and discuss the effectiveness and efficiency of the two proposed strategies, providing a reference solution for the integration research of power communication networks under multiple communication systems.
