Search

Edge computing can be widely used in unmanned aerial vehicle (UAV) inspection, field operation control, power consumption information collection and other businesses in the power Internet of Things scene. Edge computing offloads functions such as data processing and applications to network edge nodes near the terminals to provide low-latency services and ensure service quality. However, with the explosive growth of business terminals, the capacity of single edge node is limited and it is difficult to meet all business requirements at the same time. Therefore, this article proposes a task allocation mechanism based on cooperative edge computing. Firstly, a task allocation model based on cooperation of two edge nodes is established to minimize the average task completion delay while meeting business requirements. Secondly, the Two-edge-node Cooperative-task Allocation based on Improved Particle Swarm Optimization (TCA-IPSO) algorithm is proposed, which applies the crossover and mutation strategy in genetic algorithm to improve the particle swarm optimization algorithm, and solves the problem that the task allocation scheme in cooperation is prone to fall into a local optimum. Finally the simulation results show that the proposed TCA-IPSO algorithm reduces the average task completion delay by 53.8% and 36.0% compared to the benchmark and QoS-based Task Distribution (QBTD) algorithm.

With the rapid development of Internet of Things (IoT) technology, Network Function Virtualization (NFV) is introduced in the edge network to provide flexible and personalized service. However, there still exist some problems to be solved, such as high cost, unbalanced load, and low availability. Therefore, a reliability-and-energy-balanced Service Function Chain (SFC) mapping and migration method is presented for IoT applications. First, aiming at improving network performance and reducing expenditure, an SFC mapping algorithm based on cost optimization, load balancing, and reliability is proposed to map SFC requests onto the network and provide backup. Second, aiming at optimizing resource configuration, an SFC migration method based on energy consumption and quality of service is proposed to integrate network resources. Simulation results show that the proposed method outperforms the compared algorithms by 15.5% and 24.55% in the acceptance ratio of SFC requests and the overall costs, respectively.

Breast tumor segmentation is a preliminary step for breast tumor diagnosis. It is expensive to obtain pixel-level labels for breast tumor segmentation. We aim to perform breast tumor segmentation in breast ultrasound (BUS) images with image-level labels. To this end, we propose a class activation mapping with level set (CAM-LS) method. Given a training dataset consisting of BUS images with or without breast tumors which naturally contain image-level labels, a classification network is trained for breast tumor recognition. Class activation maps (CAMs) are generated for positive predictions, which highlight discriminative regions of breast tumors. Anatomical constraints are used to reduce the search space for breast tumors. For breast tumor recognition, the proposed method achieves sensitivity of 92.0%, precision of 92.8%, specificity of 95.2%, accuracy of 93.9% and F1-score of 0.924. For breast tumor segmentation, the proposed method obtains Dice similarity coefficient (DSC) of 73.5 ± 18.0% and interaction-over-union (IoU) of 60.8 ± 19.7%. The proposed CAM-LS method performs automated breast tumor segmentation with image-level labels only and is demonstrated in the experiments to have good generalization on different ultrasound platforms.

Breast anatomical structures are distributed layer by layer in breast ultrasound (BUS) images. Breast anatomy decomposition (BRAND) has several clinical potentials in providing localization constraints for breast lesion detection and supporting multimodal registration. To take advantage of convolutional neural network (CNN) and reduce the requirements of manual labels, we propose a semi-supervised deep learning method for BRAND. A CNN is first trained on labeled training data. Pseudo labels for unlabeled training data are generated by smoothing the network's predictions with simple linear iterative clustering (SLIC) super-pixel algorithm. The proposed approach achieves Dice similarity coefficient (DSC) of 89.6 ± 9.3%, 88.2 ± 9.7%, 84.6 ± 17.7% and 89.5 ± 21.6% for subcutaneous fat, mammary gland, muscles and thorax layers, respectively, which outperforms the reported results. The proposed approach is demonstrated to be sufficient for BRAND in BUS images both with and without breast lesions.

The electronic medical record (EMR) of liver cancer covers a large amount of key information, including the pathological stage of the tumor, the location of the tumor, and the size of the tumor, which helps doctors quickly understand the patient’s condition and make diagnosis. However, conceptually complex and multi-terminological EMR makes doctors hard to retrieve useful information, which not only leads to low work efficiency, but also may miss the key information of pathological understanding. Named entity recognition (NER) technology can help doctors quickly screen out key entities and improve the efficiency of clinicians. In this study, we proposed two model structures (Damped Pointer Network and Dynamic Fusion) to improve the accuracy and recall rate of entity recognition. The model structures were well fitted the EMR with small size of samples and many technical terms in the form of uncommon words. Experiments showed that the F1-score (harmonic average of accuracy and recall rate) of the model structures proposed in this study was 98.56%, which was 2% higher than other frequently-used recognition technologies like BERT and BERT-CRF.

In order to solve the complexity of the fault caused by the nonlinear circuit and the variation of component parameter tolerance, the fault diagnosis method of analog circuit based on the SVM is deeply studied and used in the analog circuit. First, in order to solve the problems of the low efficiency and the low precision of traditional SVM parameter optimization methods, take QPSO as the starting point and improve the diversity of particle swarm optimization as the goal, combine with the emerging swarm intelligence optimization algorithm QPSO(quantum-behaved PSO) to improve the SVM parameter optimization; Secondly, in view of the problem of higher dimension and lower diagnosis efficiency caused by the increase of attribute types, from the perspective of feature selection, the IBQPSO algorithm is used to combine the feature to be selected with the parameter optimization of SVM, in order to improve the classification accuracy and generalization ability of the SVM; Finally, an example of typical filter circuit is given to show that the processed method should be feasible and effective to be used in the fault diagnosis of analog circuit of photoelectric equipment.

The bubbles in the liquid film are important to improve the chemical reaction rate and the efficiency of heat and mass transfer. However, bubbles in annular flow have not been studied intensively like waves on film surface or droplets. Planar Laser-induced Fluorescence (PLIF) was employed to investigate the film bubbles in vertical upward annular flow. The subject of this paper is to study the correlation between the area, the gas holdup of bubbles in the liquid film and gas-liquid superficial velocity. The results show that the above parameters decrease with the increase of the gas superficial velocity, however, with the increase of liquid superficial velocity they are not always increasing. The variation of these parameters with the liquid superficial velocity is also diverse under different gas superficial velocity. The statistical characteristics of bubble area are studied and it can reveal how the mean bubble area decreases as the flow conditions change.

Falling film is widely applied in many industries, such as evaporation, condensation, chemical reaction and distillation. In order to investigate the influence of the liquid temperature on the wave characteristics in the free-falling film that is subjected to the sensible heating, instantaneous measurements of the film thickness using planar laser-induced fluorescence are performed. Experiments are carried out at liquid temperature that ranges from 24 °C to 52 °C. The spatiotemporal flow structure of falling film is reconstructed and analyzed at different temperatures. It is found that the frequency and the amplitudes of the large waves change with the temperature simultaneously, and the substrate regions become more smoother and wider with the temperature rising. The profiles of film thickness probability density function and the power spectrum density function are presented and discussed.

With the development of radar technology, frequency agile radar has been widely used in military because of its benefit in anti-jamming situation. In the battlefield, some friendly radars of which the signals are overlapped in the frequency range may interfere with each other when they work synchronously. The interference among friendly radars is mainly caused by the sideband signals of each other, and the spectrums of interference signals are known as they are emitted from friendly equipment. In this paper, the response of radar receiver is analyzed when sideband part of interference signal is injected into the receiver, and a calculation model of sideband signal response is proposed. The probability distribution of frequency alignment between the interference signal and the frequency agile radar receiver is analyzed. Furthermore, an interference probability calculation model of frequency agile radar that is interfered by the sideband signal jamming is proposed based on the jamming power calculation model. With modeling the frequency agile radar and friendly radars in the battlefield situation, the typical friendly radar jamming scenarios are simulated in computer, and the results produced by the calculation model are compared with the simulation results to verify the validity of the calculation model proposed in this paper. The interference probability calculation method of frequency agile radar based on the spectrum of interference signal proposed in this paper can be further extended to calculate the jamming probability of frequency agile radar in electronic countermeasure environment.

With the rapid development of deep learning and computer-vision, accurate identification of medical imaging has become one of the most important factors in medical diagnosis and decision-making. To this end, we propose a data-driven approach named Evolutionary Convolutional Deep Network (ECDN) to detect malaria parasites, which can use evolutionary algorithms to automatically generate deep neural networks, and optimized its network topology structure during the evolution process. Extensive experiments based on the large-scale thin-blood smear images data validate the effectiveness of ECDN for detecting malaria. To be specific, it has the advantage of being able to automatically generate an optimal network structure without the need for any prior knowledge of constructing a neural network, as compared to a traditional artificial convolution network. The experimental results show that the model robustness of ECDN is better. When the training set and test set are divided according to the ratio of 6 and 4, the best result is achieved, and the accuracy rate reaches 99.98%, which provides an important basis for this research.

In recent years, with the construction of large-scale urban gas pipelines, accidents such as fires, explosions caused by gas pipelines have been increasing. The catastrophic effect of these accidents on people and their properties highlights the urgent need to develop appropriate safety strategies for those who are in risks. However, there are no effective measures to identify gas pipelines with hidden danger due to occupation or crossover. The aim of this study is to quickly identify a large number of gas pipelines with hidden dangers at one time and locate the positions of the targets that cause such hidden dangers. This paper proposed a rapid hidden danger detection method for determining geographic information of hidden danger targets and their associated gas pipelines. It consists of three main parts: an overview of the rapid detection method, the general identification process and the identification of hidden danger targets with their associated gas pipelines. This paper proposed matrix-based "hidden danger identification area", "potential hidden danger target area" and "primer number identification method" to find all hidden danger points at one time. This method contributes to the safety of cities and becomes an important part of smart cities.

In order to simulate the spaceborne early-warning radar echo, a clutter simulation model is proposed. In spaceborne radar clutter simulation, the problem of two-dimensional widening in the range-Doppler domain of traditional simplified model is analyzed, as a result of beyond the precondition of the conventional airborne clutter simulation model. For the spaceborne condition, a new simplification is proposed. Taking into account the accuracy and efficiency of the simulation, a simplified spaceborne warning radar clutter model is proposed. The simulation results show that the new model improves the accuracy of the clutter simulation compared with the traditional one, and the simulation efficiency is taken into account as well.

The simulation of spatial-temporal correlated sea clutter returns is of great significance in the research of radar detection and performance analysis. Traditional simulation methods generally introduce a recursive formula into the Spherical Invariant Random Process (SIRP) method to build the Spatial-temporal correlation properties of the sea clutter. However, the methods are computationally intensive and inefficient due to the recursion. In this paper, an efficient construction method of spatial and temporal correlation is proposed for the first time. The new method has reduced the computational burden and became parallelizable. The simulation results have been analyzed in this paper, which show the feasibility and accuracy of this method.

PHM (Prognostics and Health Management) is a key technology in autonomic logistics equipment. It is also the newest development in the current equipment system, BIT (Built-In Test), because of its condition monitoring abilities. PHM technology marks a transformation from condition monitoring to health management. The PHM design philosophy has also been adopted in the development of new generation military and civilian equipment. In the complex systems of military and civilian equipment, such as electromechanical systems and flight control systems, the PHM design has been adopted as a new function design. The diagnostic capabilities of PHM are its core foundation, and its design level has a direct influence on the overall efficiency of PHM. This influence is mainly reflected in its fault detection capabilities, and its comprehensive processing and analysis capabilities. In order to validate the PHM fault diagnostic design level, we propose a new diagnostic capability validation method for complex systems. Our proposed method is based on testability validation technology, and provides a guideline for technological approaches for PHM diagnostic validation that evaluates the integrity, rationality, and effectiveness of PHM diagnostic capabilities. Moreover, our proposed method can expose design defects and weak links, allowing for an iterative optimal design.

Policy-based concept and service-oriented design are two promising techniques in network management. In this paper we show that integrating policy-based management with service-oriented design can significantly improve the autonomic and flexible capabilities for management. In particular, we propose an architecture that covers policy modeling, specification and enforcement in a service-oriented manner. We develop the key components of this architecture: notification collection service, eXPath parsing engine, policy event matching component, policy condition evaluation component, policy action parser, policy action executor, policy repository, and web services pool. The results from our preliminary performance evaluation demonstrate that the proposed architecture generates an acceptable policy response delay and transmission bandwidth cost.

Failure instances in distributed computing systems (DCSs) have exhibited temporal and spatial correlations, where a single failure instance can trigger a set of failure instances simultaneously or successively within a short time interval. We investigate an effective approach to predict correlated failures of computing elements (CEs) in DCSs. Correlated-failure patterns are modeled using the concept of probabilistic shared risk groups (PSRG). Firstly, we design a new structure for PSRG, named SPSRG, to describe features of correlated failures. Then we exploit an association rule mining technique in a parallel way to generate and update our SPSRG using information of CE-failure states. Finally, we propose a correlated failure prediction approach to evaluate the probabilities of upcoming failures from the SPSRG. The experimental results show that the proposed approach outperforms other approaches in failure prediction performance in terms of precision, recall and F-measure. Moreover, it allows employing customizable thresholds by which the trade-off between precision and recall can be adjusted for various requirements.

This paper is concerned with the problem of designing output feedback controller and guaranteed cost controller of discrete-time singular piecewise-affine systems with time delay and norm-bounded time-varying parameters uncertainties. Firstly, the corresponding operating region of the time-delay singular piecewise-affine systems in state space is described as ellipsoid which can be solved by a set of vector inequalities. Then, based on a singular piecewise Lyapunov-Krasovskii function combined with S-procedure and some matrix inequality convexifying techniques, the output feedback controller of the time-delay singular systems is investigated, and meanwhile, the guaranteed cost controller of the time-delay singular piecewise-affine systems is solved in terms of linear matrix inequalities (LMIs). It is shown that the controller gains can be obtained by solving a family of LMIs parameterized by one or two scalar variables. Numerical examples are given to illustrate the effectiveness of the proposed design methods. Moreover, it is suitable for practical application.

In this paper, the problem of designing robust non-fragile guaranteed cost state feedback controller for a class of continuous-time singular piecewise-affine systems is considered. Based on a singular piecewise Lyapunov function combined with S-procedure and some matrix inequality convexifying techniques, the robust non-fragile stabilization condition is established, and meanwhile, the condition of constructing non-fragile guaranteed cost state feedback controller is proposed. Under the multiplicative perturbation, the design method of robust non-fragile stabilization controller is given in terms of the feasible solutions of linear matrix inequalities. It is shown that the controller gains can be obtained by solving a family of LMIs parameterized by one or two scalar variables. Finally, numerical examples are given to illustrate the effectiveness of the proposed design methods.

In this paper, we briefly analyzed the advantages of passive distance measurement based on Wireless Fidelity (WiFi). Based on the analysis of the principle of traditional interferometer direction finding, we derived the equations which are suitable for indoor distance measurement. Through theoretical analysis and simulation, we prove the correctness of the equations. Finally, we analyze and conclude the conditions of using the method, the scope of distance measurement, and the main influencing parameters of system accuracy.

The definition of network management interface is generally divided into three phases, requirements, analysis and design. In analysis phase, UML is used as the modeling language to present technology independent models, and these models can be mapped into multiple technology specific models. With the development of Web Services applied in network management domain, it is required to define Web Services-based information models in design phase. This paper applied MDA approach to realize models transformation. As the main work, this paper proposed detailed mapping rules describing how to map the existing source UML models into the target Web Services-based models. An automatic transformation approach using XSLT was proposed to implement the mapping rules, and experiments were made for verification. The proposed mapping rules improved the defects in related work. Furthermore, the work of this paper could help resolve some issues in current models transformation work which is still being accomplished manually by standard developers.

The surrounding area of the Lingding Bay has one of the highest economic development rates of China. Rapid industrialization and urbanization has resulted in extensive changes in land use, including the construction of harbors and embankments. For the analysis of the changes of the coastal zone multi-temporal Landsat images have been digitized by using the integrated RS and GIS technique, in combination with topographical and nautical data. From the change analysis, it can be concluded that the largest variations in the position of the coastline over time occurred in the Nansha Harbor Zone and Qianhai Bay, mainly caused by extensive harbor construction. Based on the research of the coastal change in recent decades, it is suggested that the human activities have become one of the major factors affecting the morphological processes of the Lingding Bay.

Digest E-government is becoming more and more popular among governments, accelerating e-government is now a major trend of today's world economic and social development. By constructing E-government, it helps to improve comprehensive application of information technology in different government sectors' business, and to realize the exploitation, utilization and accumulation of governmental information resources.

This paper proposes a new way to complete the interface model mapping from CORBA to ontology in the field of network management, which is Interface Model Mapping Adapter. The adapter consists of three parts, which are CORBA interface analyzer, mapping controller and ontology information database. The adapter can automatically map the CORBA interface model to ontology interface model. Therefore, we can use the output of the adapter to establish the Ontology-based Network Management Information Model, which is much helpful for the research of unified network management.

The paper is to solve how to map OWL information model to database which is very important for sharing information. First, relatively complete and exact mapping rules with the consideration of OWL syntax and semantics is designed. Then, the mapping system from OWL to relational database is implemented based on Jena using Java. Finally, the mapping system is tested by some OWL examples. The mapping rules in this paper are lossless and canonical, and every element in OWL can be stored in the database. The mapping system is advanced, simply operable and robust.