This PDF file contains the front matter associated with SPIE Proceedings Volume 12162 including the Title Page, Copyright information, and Table of Contents.

Deep neural networks are especially vulnerable to adversarial example, which could mislead classifiers by adding imperceptible perturbations. While previous researches can effec- tively generate adversarial example in the white-box environment, it is a challenge to produce threatening adversarial example in the black-box environment till now, where attackers only have access to obtain the predicts of models to inputs. To conquer the problem, a feasible solution is harnessing the transferability of adversarial examples and the property makes adversarial examples can successfully attack multiple models simultaneously. Therefore, this paper explores the way to enhance transfer- ability of adversarial examples and then propose a Nadam- based iterative algorithm (NAI-FGM). NAI-FGM can achieve better convergence and effectively correct the deviation so as to boost the transferability of adversarial examples. To validate the effectiveness and transferability of adversarial examples generated by our proposed NAI-FGM, this study conducts the attacks on various single models and ensemble models on open Cifar-10 and Cifar-100. Experiment results exhibit the superiority of NAIFGM that achieves higher transferability than state-of- the-art methods on average against black-box models.

Based on polarization independent optical splitters and beam splitters, a new logic gate design method based on self collimation effect is proposed. Using the characteristics of light self-collimation transmission in the photonic crystal, combined with the optical splitter, the optical path difference is introduced, so that the corresponding phase difference between the incident light is generated. It makes the light beam produce interference effect at the beam splitter converge, so as to realize the logic function of NOR gate, NAND gate and XOR gate. Simultaneously, the plane wave expansion method and the finite difference time domain method are used to simulate and analyze the logic gate in the TM polarization mode to verify its logic function. The analysis of the results shows that the logic gate designed by self-collimation effect and linear interference is compact, smaller in size, easy to manufacture, short in time response, high in transmittance, and more suitable for the development of optical integration.

Satellite communication with its unique advantages, the use of high-throughput multi beam technology, to achieve global connectivity, has broad application prospects. Based on the development of new satellite payload technology, in order to improve the utilization of system resources, it is necessary to solve the problems of resource allocation and MIR/CIR rate guarantee between different VNO terminals under different beams, different gateway and different satellites. The method for bandwidth distribution is applied to the HNO on how to distribute the satellite bandwidth to two or a plurality of VNOs on a specific beam which respected to one Global Bandwidth Manager (GBM). The presented algorithm of bandwidth distribution for the HNO has the ability to calculate on how much bandwidth to be allocated to one or more VNOs across multiple beams, gateways and satellites which respected to a global bandwidth subscription. Based on this method, the bandwidth allocated to a specific VNO will not exceed the network capacity even if their resource requirements are higher than the available resources.

High-level synthesis technology is a technology for circuit design using high-level languages such as C, C++, or System C. This design method is suitable for algorithm engineers to implement calculation prototypes and is not suitable for hardware engineers to design circuits from the register transfer level. In addition, from the perspective of agile development and code reuse, C, C++, and System C are far from another hardware description language, Chisel. However, there are relatively few electronic design automation tools for Chisel, so we propose a circuit design flow based on CGRA hardware, in which Chisel designs the PE unit of CGRA. However, Chisel has few electronic design automation tools, so a circuit design flow based on CGRA is proposed. Dedicated data flow graph generation tools and typical coarse-grained reconfigurable arrays are implemented for this design flow. This article describes a design process based on CGRA and compares the design process with HLS through the DCT circuit design process. Through the experimental comparison of the two circuit implementation processes, it is found that, first of all, the correctness of the CGRA-based design process and the availability of data flow diagram generation tools. Secondly, from a design point of view, the use of Chisel will increase the overall development speed and increase the reusability of design files. Finally, from the perspective of toolchains, part of the tools in the new design process uses open-source toolchains, which reduces the cost of using toolchains and promotes the development of toolchains to a certain extent.

In order to detect the faint and small targets in the high-speed moving radar in the complex electromagnetic environment, a new method for tracking the faint and small targets in the complex electromagnetic environment is proposed. First of all, pre-processing and dynamic planning of the radar image sequence. Multi-stage hypothesis testing is performed on the candidate trajectories after frame accumulation. After the target trajectory is judged, the trajectory is tracked to realize the detection of faint small targets. Experimental results show that the proposed method can effectively improve the detection performance of small and dim targets, and it takes less time, with a minimum time consumption of 16.7s.

Shot boundary detection is the main task in the preprocessing stage of content-based video operations. There are unforeseen illumination change and motion effects in a video due to the complexity of video content, which may lead to the detection of wrong shot boundaries. This paper proposes a novel shot boundary detection method to solve this problem. The method mainly includes two parts: abrupt and gradual transition detection. In the first stage, CIEDE2000 color-difference and adaptive threshold are used to find the possible abrupt transition frames. Then BRISK feature is utilized to extract real abrupt transition frames. In the next stage, the brightness change of video frames is utilized to detect the frame group that may be gradual. Then CIEDE2000 color-difference along with cumulative frame algorithm is used to detect actual gradual transition frames. The experiment is evaluated on the TRECVid2001 and ClipShots datasets. Experimental results show that the method proposed in this paper can improve the precision of shot segmentation.

Based on the modernization of GMDSS, the characteristics of maritime digital broadcasting (NAVDAT) system and its performance requirements for antenna are analyzed, and a scheme of Medium Frequency broadband antenna for NAVDAT is proposed. The simulation results show that the antenna designed in this scheme has high bandwidth, and the input impedance of the antenna presents narrow band and changes smoothly, it can well meet the broadband communication requirements of NAVDAT.

In directional modulation (DM) networks, conventional power allocation strategy is actually an equal-power allocation method. However, in practice, each antenna is equipped with its own power amplifier and is restricted. This paper considers a more practical situation that the power of each antenna sent to the receiver is different. Here, we study the optimal power allocation (OPA) strategy with multiple-input single-output (MISO) in DM communication system. Explicitly, we first formulate a secrecy rate (SR) maximization problem subject to the total power constraint, with designing the useful information power vector and artificial noise covariance matrix. For the non-convex problem, we propose the maximizing secrecy rate based on successive convex approximation (Max-SR-SCA) method to solve it. Subsequently, in order to reduce the computational complexity, we propose the maximizing product of signal-to-leakage-plus-noise-ratio (SLNR) and artificial-noise-to-leakage-plus-noise-ratio (ANLNR) (Max-SLNR-ANLNR) method to obtain a suboptimal solution. Simulation results show that the proposed two methods have better SR performance than the conventional PA factor scheme. In addition, the proposed Max-SLNR-ANLNR method forms a single energy main peak at Eve position, which indicates it realizes precise jamming.

With the popularity of smart wearables and smart homes, any product with WWAN or WIFI functionality is likely to be subject to Specific Absorption Rate (SAR) testing in order to ensure its security. According to regulatory requirements, SAR testing is mandatory once the usage distance and maximum power exceeds the specified exemption limits. Any SAR value can consequently be deduced from electric fields measurement within a given dielectric medium, usually assumed as homogeneous for practical convenience. In this paper, six antenna numerical models are designed and simulated using SAMCAD software to obtain two-dimensional electric field data and three-dimensional electric field data, and the twodimensional data are used as input training for the GAN to generate three-dimensional data. The experimental results show that the electric field reconstruction method proposed in this paper is effective.

A small-size MIMO re-configurable loop antenna for laptop applications with a compact area is proposed. It consists of two antenna elements and each antenna element has one SP4T RF switch used for re-configuring the antenna to realize 800-980 MHz, 1570-4210 MHz and 4520-6000 MHz band coverage (GSM850/900, DCS1800, PCS1900, LTE33/34/35/36/37/38/39/40/41/42/43 and 46 bands). The antennas are 70 mm × 10 mm located at the top corners of a 300 mm × 200 mm FR4(εr=4.4 and loss tangent=0.02) substrate of thickness 2 mm. By changing the states of the RF switch from state 01 to state 04, the antenna covers the low bands as well as the high bands. The antenna has good isolation and good radiation efficiency for all the bands.

With the explosive growth of mobile communication information brought about by the maturity of 5G commercial use, the energy consumption caused by grid communication is increasing. Promoting the emission reduction and intelligentization of electric energy and communication network is the key to the current grid construction. The impact of dispatched multi-core and multi-chip power terminals is prone to problems such as increased load in local areas and data penetration. Therefore, it is necessary to reduce current energy consumption and optimize power emissions. This paper analyzes the current problems of power grid energy consumption, and puts forward corresponding suggestions on reducing energy consumption, and provides certain ideas for constructing an effective balance between power supply and demand interaction and energy optimization.

Aiming at the problem of poor quality of filtered multi-carrier transmission in current 5G communication system, a filtered multi-carrier transmission side for 5G communication system is proposed. Firstly, the cross ambiguity function is used to collect and identify the filtered multi-carrier transmission deviation. Secondly the OFDM synchronization algorithm is used to eliminate noise interference and effectively correct the information. The processing interval of filtered multi-carrier transmission is optimized to ensure the quality of filtered multi-carrier transmission. Finally, the simulation results show that the filtered multi-carrier transmission method for 5G communication system can better ensure the filtered multi-carrier transmission efficiency and security of communication system, and effectively improve the quality of information transmission.

In order to quickly scout the radar signal and accurately estimate the pulse arrive time, this paper proposes an estimation algorithm of pulse arrive time based on the differential filter in time domain. The simulation result shows that the proposed method has high accuracy of time estimation and the process time is shorter than the algorithm in frequency domain. So it can be used in engineering application.

With the development of quad-rotor UAVs, UAVs have been widely used in civilian fields such as aerial photography, agricultural production, surveillance and rescue. Controlling drones is a complex and difficult task for human operators. Therefore, it is of great significance to reduce the burden of human operation and improve the efficiency of drones when performing tasks. With the development of some novel human-computer interaction technologies and the continuous improvement of drone autonomy, manipulating drones can become easier and more interesting. In this work, we designed a method that uses operator gaze to assist drone control and target following. We conducted verification experiments on the Tello UAV, and the experimental results show that eye tracking interaction makes the human operation process more natural, and the combination of eye movement behavior and semi-autonomous drone control further reduces the burden on operators and improves the reliability of drone missions.

NVMe over TCP is a key technology for building large-scale high-performance storage systems. It can realize NVMeoF (NVMe over Fabrics) storage network based on existing data center network infrastructure and standard TCP/IP software protocol stack. In this paper, we design and implement the Load-aware NVMeoF message processing mechanism LANoT (Load-aware NVMe over TCP). Firstly, the interrupt coalescing technology based on PDU aggregation is used to alleviate the interrupt storm problem and achieve high throughput. Secondly, matching the special message processing mechanism, which can effectively improve the key performance indicators according to the I/O characteristics of different dedicated queues. This paper implements the LANoT prototype system in Linux kernel. The performance test results show that comparing to NVMe over TCP implementation in standard Linux kernel, LANoT can significantly improve IOPS metrics and reduce CPU resource consumption by more than 50%.

Based on the ring resonator and microcavity, this paper designs a new method for realizing all-optical logic gates. Firstly, manufacture defects in complete triangular lattice silicon to produce waveguide coupling and linear interference effect. Secondly, the proposed structure was simulated by plane wave expansion method and finite difference time domain method. Finally, view the simulation results and analyze the performance index parameters of the proposed all-optical logic gate in combination with experimental data. The results show that the extinction ratio of the proposed all-optical logic AND gate can reach 8dB, the structure size is small, the response time can reach 2.42ps. At the same time, the value of the proposed all-optical logic OR gate output can reach 0.72, the structure size is small, and the response time can reach 2.41ps. The performance is favorable and meets the experimental requirements.

DNS, is one of the oldest and most basic important applications of the Internet. Because normal services must access services through DNS, and malicious programs also need to be connected through DNS. There are so many DNS servers on the Internet that we can choose freely. However, in the campus network environment, if users are allowed to choose DNS server freely, it will cause confusion in network management and cannot systematically monitor and schedule traffic. We can't force users to set up DNS servers, which will cost a lot of manpower and management costs. In this paper, we focus on DNS redirection technology. We use the improved DNAT（DNAPT） to implement DNS redirection. So, we discuss a new application to implement DNS redirection. And through DNS redirection can solve many campus network security problems.

In order to meet the integration, miniaturization and lightweight requirements of green communication for deep aerospace or lunar missions, it is essential to design an integrated device to replace both the measuring microwave radar and spaceto-space communication machine for rendezvous and docking. As the main measuring device, the microwave radar provides relative motion parameters between two dynamic spacecrafts, including relative rang, radial velocity, azimuth and elevation angles. The space-to-space communication machine is used for mutual communication between spacecrafts. In this paper, the integrated design of microwave measurement and space-to-space communication is carried out, to achieve high-precision r

With the gradual rollout of 5G network commercial deployment, as the main force of indoor deployment, 5G small cell is becoming more and more popular. In recent years, there are few high-performance channel estimators for small cell. The research work of this paper is to achieve a low-delay, low-overhead, relatively universal channel estimation module, which provides a reference for the design of channel estimator for 4G/5G small cell. By designing the instruction set of a dedicated instruction set processor, the function of channel estimation was realized on the vector processor. Through instruction acceleration, compared with the existing general vector processing instruction set, the instruction overhead of the LS estimation module is reduced by 50% and the Wiener interpolation module in the frequency domain is reduced by 37.5%. And the performance of Wiener filter interpolation in frequency domain is 0-2db better than that of linear interpolation in frequency domain, no performance penalty associated with the computing cost reductions.

Analysis of radars systems requires representing the radar operation and performance in the context of the overall system and the external environment. The radar performance must be evaluated, and the impact of radar operation on system performance must be quantified. In this paper, we simulate radar system by using the Radar Designer app in MATLAB, analysis the system to decide whether a set of performance requirements, such as pd and SNR, have been satisfied according to the specifications. This paper also discusses the loss of radar system in rain, and the performance of radar system when detecting Small UAS.

Query response time is an imperative performance parameter for applications driven by database systems. This paper proposes a novel cache strategy to speed up queries, especially in search-within-result. Redis is implemented as a cache server for Mongo database. We convert multiple query conditions into a set of filters and concatenate the filters to be capable of being keys in Redis. Thus, there is a cache hit if a subset of the filters is cached, not requiring that a query is exactly the same as previous queries. This leads to an increase in cache hit ratio and accelerates queries. Experimental results on a real dataset show the effectiveness of our proposal, achieving a 71.27% improvement of the average query response time.

The selection of targets for communication jamming directly affects the establishment of electromagnetic dominance at battlefields and it is the core content for planning combat operations in battlefield communications. Modern warfare is full of high-intensity conflicts and confrontations, and the battlefield information acquired from observation is incomplete and inaccurate sometimes. The existing methods to select targets to attack by communication jammers focus more on pre-war planning, are not closely integrated with the dynamic changes in the combat process, and do not take the cause-effect relationship into full consideration. This paper proposed a method of constructing dynamic Bayesian networks to select targets to attack by communication jammers, provided the algorithm flow of the method, and conducted a simulation analysis with an arithmetic example.

Electric energy is one of the important boosting forces for the progress of social and economic development and the improvement of people's quality of life. As the basic carrier of electric energy, the transmission and transformation project can be completed and put into operation on schedule, which is of great significance for ensuring the safe and effective supply of electric energy. Therefore, based on the analysis of construction schedule management problems and influencing factors of power transmission and transformation projects, this paper constructs a support vector machine-based construction schedule early warning method model, and verifies the validity of the model based on actual engineering cases. This model can provide a reference for improving the management level of the construction schedule of power transmission and transformation projects.

In order to improve the efficiency of full bridge LLC resonant converter interleaved parallel (FBLLC-SP) system, and make its theoretical loss closer to the actual operation loss, In this paper, the parasitic parameters of cascode GaN HEMT are introduced into the loss model , to set up an accurate loss model of FBLLC-SP system, which is suitable for single phase shift (SPS) open-loop control and three phase shift (TPS) double closed-loop control; In this paper, the optimal dead time calculation method is given, and the theoretical loss of FBLLC-SP system under single phase-shifting (SPS) open-loop control is analyzed, and the accurate calculation methods of transient loss of GaN HEMT with different modes are proposed; The experimental system is built. The experimental results show that the loss value of the loss model in this paper is about 0.6% higher than the actual operation loss of the system, that is to say, the refined loss model established in this paper can be used to analyze the actual operation loss of the system.

In order to study the cross-slanted self-bucking-restrained corrugated steel plate shear wall (CCSPW), the ABAQUS finite element software model was verified by establishing a test piece test data, and the effect of axial compression ratio on the seismic performance of bidirectional inclined self-proofed corrugated steel shear wall was simulated. The results show that the finite element calculation results are in good agreement with the experimental data. In a certain range of axial compression ratio, with the increase of axial compression ratio, the energy dissipation capacity of CCSPW increases, and the yield load and ultimate load of specimens decrease slightly, which is not conducive to the stability of bearing capacity and ductility of the structure.

In view of the current situation that the numerical simulation parameter analysis of frame structure mainly focuses on the macro characteristics such as stiffness and displacement, the author uses ABAQUS finite element (FE)software to establish a three story two span steel reinforced concrete plane frame and carry out pseudo dynamic analysis. The reasonable constitutive relationship of concrete and steel, the constraint relationship between concrete and longitudinal reinforcement, stirrup and section steel, the reasonable seismic wave input mechanism and the structural damping of plane frame are fully considered in the modeling. By comparing the displacement and acceleration time history results in the finite element with the test results, the difference between them is small, which shows the rationality of the modeling method, and further analyzes the meso characteristics of plastic energy dissipation of the plane frame. The results show that the proportion of plastic energy dissipation of frame beam is greater than that of frame column; According to the material, beam and column reinforcement are the main energy consuming components of plane frame, followed by concrete, and the energy consumption of section steel is relatively small; The modeling method can efficiently and accurately simulate the pseudo dynamic performance of plane frame, and can be used as a reference for subsequent practical projects.

Aiming at infrared dim small target detection, this paper proposed a spatial and temporal fusion detection algorithm based on mathematical morphology. Firstly, the obtained infrared image is filtered in spatial domain by morphological Top-hat filtering, so that most of the background and clutter in the original infrared image are suppressed. In time domain, some fixed background is removed and the target is enhanced by three frame difference filtering, and the corresponding rules are used for fusion. The fused image is segmented by adaptive threshold to obtain the potential target points. Finally, according to the characteristics that the real target has continuity and regularity in time domain and the noise points are randomly distributed, the pipeline filtering algorithm is used to further filter the false alarm points and detect the motion trajectory of the target. Simulation results show that comparing with the Top-hat filtering and the time domain three frame filtering algorithm, the proposed fusion algorithm can effectively detect the target and accurately detect the false alarm target correctly.

The steering problem of multi axle vehicles in sandy soil has always been an important issue in terra-mechanics. Numerical simulation method has become an important research method to study the vehicle-terrain contact problems. In this paper, the discrete element method is used to establish the sandy soil pavement model, and the multi-body dynamics method is used to establish the vehicle body model. The coupling of two methods are used to carry out the driving simulation of multi axle vehicles on sandy soil pavement, and the force of tires and the speed and trajectory of vehicles are analyzed. The results show that the DEM-MBD coupling method has great application prospects in vehicle-terrain simulation.

With the development of Internet big data, Internet evaluation and Baidu index database can obtain comprehensive tourist behavior characteristics and travel experience of Longji yao and zhuang ethnic terraces tourists, using Python language to analyze Longji yao and zhuang ethnic terraces tourists' travel time and focus of attention. Propose of optimize marketing strategies for the source market; expand visibility and develop tourism for ethnic minorities; optimize services and enrich the experience activities and other countermeasures.

Aiming at the problem of penetration flight parameter design of sea skimming aircraft, the analysis model of sea skimming anti-ship aircraft penetration capability is established based on the detection capability of shipborne radar, infrared detection system and aircraft survivability. The influence of flight Mach number on sea skimming aircraft penetration capability is studied. Under the atmospheric standard condition, the relationship between the Mach number of sea skimming flight and the operating range of enemy infrared detection system is analyzed. Considering the radar sea clutter jamming effect and the hitting probability of aircraft flying across the sea, the cruise altitude and maximum covert flight Mach number under different sea conditions are studied. The results show that the higher the sea state, the closer the sea clutter jamming area is to the target, and the smaller the corresponding maximum stealth penetration Mach number.

With the wide application of Block chain, the phenomenon of malicious attack of Block chain also increases, which makes Block chain face the risk of data damage, tampering, leakage and so on. And with the increase of data and application, the efficiency of Block chain also needs to be improved. Aiming at the relatively low computational efficiency of the hash function in the block chain, which affects the overall computational efficiency of the block chain and even leads to the security hidden trouble in the block chain, a safe and efficient hash algorithm is proposed. The experimental results show that on the premise of ensuring the security of the algorithm, the efficiency of the hash algorithm is improved, so as to improve the performance of the block chain system.

Streamline is one of the most inspiring technologies to achieve fundamental goal of scientific insight from resulting numerical simulations. Drawing on an intuitive interpretation in terms of flow field, it is an ideal tool to illustrate and describe a wide range of phenomena encountered in the study of application-domain vector fields. In this paper, we propose a parallel streamline algorithm based on Dynamic Node Tree (DN-Tree), to improve the efficiency of streamline visualization algorithm with efficient utilization of multi core computing resources. Firstly, we realize fast location of grid cells via constructs a dynamic node tree in parallel. The grid cells are divided into coarse-grained and fine-grained partitions in terms of organizing and managing the dataset, and the index structure is established to conveniently narrow the alternative areas in the integration process. Secondly, we reduce memory requirements of streamline integration by the methods of data blocks attribute extraction and employ parallel execution of seed integral tasks to leverage multicore computing resources. Finally, the performance is analysed by comparing our algorithm with existing streamline algorithms using datasets of different scales. The results show that the proposed algorithm improve the efficiency of streamline visualization algorithm and can achieve a higher parallel speedup.

In the context of global warming, how to improve the innovation efficiency in China’s renewable energy industry is of great significance. In this paper, we use the panel data of listed companies of China’s renewable energy industry from 2012 to 2018 as a sample and apply the distance output stochastic frontier model to measuring the investment (R&D input) efficiency by considering the factors affecting the innovation efficiency of China’s renewable energy industry. Conclusions show that the innovation capacity of China’s renewable energy industry is negatively correlated with economic development level, and positively correlated with the market structure, ownership structure, and R and D expenditure structure. From 2012 to 2018, the innovation capacity of most listed companies in the renewable energy industry is on the rise.

In radar down-to-sea detection, in order to determine whether the measurement results of two targets at different times in the same space area come from the same target, so as to facilitate the identification and tracking of the target, in this paper, a method for correlation determination of target observation results based on Bayesian statistics is proposed. Firstly, the derivation of this method is introduced. Then, the explanation of this method is given. Finally, the method is verified by simulation data, and the success rate reaches 99% under ideal simulation data, 91% after adding noise, which proves the correctness and practicability of the method.

According to the characteristics of multimodality, complexity and computer-aided segmentation of MRI brain tumor images, an improved UNet3+ brain tumor segmentation network based on full-scale jump connection is proposed. The main work is as follows: the residual structure is integrated into the feature extraction part of the network, and then the data correlation upsampling structure called Dupsampling is integrated into the upsampling part of the network to improve the quality of the feature map and prevent the gradient from disappearing. Finally, the hybrid loss function combined with cross entropy and dice is used to further improve the segmentation accuracy. The method was validated on BraTs2019 partial data, and the dice coefficients of the whole tumor, core tumor and enhanced tumor are 0.9186, 0.8419 and 0.8358 respectively, and the Hausdorff distance is 2.097, 3.672 and 2.008, which are better than most brain tumor segmentation models.

In the era of big data, the topic, how to use modern technology to build a suitable and perfect database for a specific demand, has become heated-discussed. The expectation given to databases is that database can organize the data, correlate the different data representations and mine data value. These requirements are in line with the original design intent of Neo4j graph database. This paper constructs the knowledge graph of listed companies based on Neo4j. The result shows that the knowledge graph, as a powerful form of graphical database, can sort out and mine the relationship between data. At the same time, it also provides a reliable system for the public to learn and understand all kinds of information of listed companies.

Robust localization in dynamic scenes is one of the bottlenecks in the development of visual SLAM. In recent years, many researchers have built new slam systems to avoid the interference of dynamic objects. However, it is not perfect to provide the original object mask with the existing semantic segmentation methods, which can’t completely cover the moving objects, especially the object boundary. In order to solve this problem, we proposed a dynamic SLAM method combining semantic information and multi view geometry in this paper. In our algorithm, semantic network is used to recognize and segment each object in the image frame. And then, the moving object processing link in the dynamic scene is realized based on semantic tags and multi-view geometry, so as to reduce the influence of moving objects on the pose estimation of SLAM algorithm and improve the scene adaptability of the algorithm. We evaluate the performance of our system with TUM dataset. Results are compared with the typical visual SLAM system to show that the absolute trajectory accuracy in the algorithm proposed has been greatly improved.

Knowledge Graph Completion (KGC) aims at predicting missing information for knowledge graphs. Most methods concentrate on learning entities’ representations with structural information indicating the relations between entities, while the utilization of entity attribute information is not sufficient for KGC. How to use the complex and diverse entity attribute information for KGC is still a challenging problem. In this paper, we propose a novel joint model Entity Structural and Attribute Embedding (ESAE) for KGC, which takes advantage of the entity structural and attribute information. Specifically, we first design a novel encoder Attribute Encoder (AE), which encodes both entity attribute types and values to generate the entities’ attribute-based representations. Based on AE, we use the structure-based and attribute-based representations in ESAE. We evaluate our method on the KGC task. Experimental results on real-world datasets show that our method outperforms other baselines on KGC.

This paper propose an image security attribute annotation technology based on CAN and WordNet to realize the flexible, controllable and automatic annotation of dynamic emerging image security attributes, and providing the basic support of security attributes for dynamic and accurate access control based on ABAC (Attribute Based Access Control). Firstly, the transfer-learning model CAN (Contractual Adaptation Network) with CBAM (Convolutional Block Attention Module) attention mechanism used to extract the entity labels of image data. Secondly, based on the extracted entity label set, propose a security attribute tree construction method based on WordNet. This method can construct a security attribute tree with continuous semantic generalization and coarsening granularity layer by layer, to ensure that the scale of security attribute space is controlled and meet the access control requirements of different granularity. The experiment results show that the annotation accuracy of this technology can reach an average of more than 91% under different control granularity.

Alleviating the imbalance between grassland ecological protection and economy at this stage is conducive to strengthening grassland ecological construction in China. The multi-agent-based grassland ecological policy driving model adopts the idea of multi-agent simulation modeling, formulates various agent behavior rules based on real data, and realizes the interactive behavior between agents by changing the influence factors. The model takes the Xilin Gol League in Inner Mongolia as an example to simulate the interaction between herdsmen, herder's family and the government under different scenarios, visualizes the grassland land utilization rate, pressure index, per capita income, livestock breeding rate and so on in the next 30 years, and maximizes the ecological and economic benefits by adjusting policies and strategies. The simulation results show that the current herdsmen are gradually shifting to urban employment, and the government regulation can continuously improve the comprehensive benefits, which not only improves the herdsmen's living wellbeing, but also achieves the effect of protecting the ecological environment. The model can predict the future development trend of grassland and the living standards of herdsmen under different policies, which is helpful for the government to grasp the direction of decision-making and maintain the ecological balance of grassland.

This article is based on RSA algorithm analyzes the basic content, combining with the RSA algorithm in the file system design, auxiliary work component design, safety management component design, the concrete application in the design of business functions to study an example RSA algorithm implementation process in the smart card operating system, its aim is to improve the cognitive level of the role of RSA algorithm, meet the operating requirements of the smartcard operating system.

Coronavirus Disease 2019 (COVID-19) pandemic has had a huge impact on airport traffic, but few works have researched about this field. In order to have a further understanding of the influence, this paper studied the dataset Covid Impact on Airport Traffic. By exploratory data analysis and visualization of the dataset, this paper has a clear insight of how the airport traffic was influenced by COVID-19. This paper also analyzed the correlation of the data with confirmed covid-19 cases using an auxiliary dataset Time series summary of global confirmed covid-19 cases and forecasted the how would the data in Covid Impact on Airport Traffic change with the number of confirmed covid-19 cases. And finally, this paper estimated the financial loss of airports. The result of this paper showed the influence of COVID-19 from multiple aspects and it has instructive significance in some degree.

An efficient vector and scalar fused load store unit (VS-LSU) is designed under the open source RISC-V framework. It can support vector access instructions of different addressing modes. Compared with scalar units, it reduces the number of instructions while increasing the memory access speed. Through the fusion of vectors and scalars, the growth of port and buffer overhead is effectively controlled. The experimental results show that when the same number of memory access operations are completed, the 256-bit VS-LSU reduces the number of instructions executed by 75%, and the memory access speed increases by 72.1% than the scalar unit; the fusion optimization reduces the overhead in port signal and buffer by more than 90% compared to the conventional vector unit, and the overall performance has been significantly improved.

Based on the expanding rigid demand for medical care, internet medical care is developing rapidly at an unprecedented speed. According to statistics of the World Health Organization in 2016, depression has become the second burden disease in China and is expected to rise to the first in the world by 2030. With the continuous popularization and development of the Internet and mobile networks, more and more social platforms appear in the public perspective. Researchers can get whether the subjects have depression tendency after emotional analysis of emotional texts from social media, which is of great significance for online medical treatment and depression screening. The method of text emotion analysis is often used in traditional fields by previous researchers, such as consumer emotion index, review analysis of film and television products, and it is rarely used in the field of mental health. In addition, the previous prediction of mental diseases focused on the use of traditional methods, such as conversation diagnosis, CT examination, painting diagnosis, etc., lacking the participation of the Internet. The main research content of this paper is the emotional analysis of social media on mental health related texts. After preprocessing and classifying the text data crawled from social media, this paper obtains the mental health emotion analysis data set, and proposes LSTM neural network based on the pre-training language model and a variety of traditional classification algorithms for this data set. The algorithm achieves good results on the data set proposed in this paper and realizes the prediction of depression tendency based on social media.

Medical image segmentation plays an increasingly important role in the whole field of image processing. Among them, the method of tumor segmentation has been paid more attention because of its special clinical significance. To solve the problems of traditional pulse-coupled neural network (PCNN) in the field of medical image processing, an internalactivity-changed FCMSPCNN (IAC-FCMSPCNN) is proposed to segment pulmonary nodules. This method further optimizes and improves the synaptic weight matrix, link strength and dynamic threshold, and reduces the number of model iterations. Experimental verification on five images in PET-CT lung cancer image library shows that the proposed method has good segmentation effect and is more suitable for clinical medical image segmentation.

In recent years, traditional deep learning has been widely used in the time series prediction of air quality, but this kind of model has many shortcomings in the input selection of meteorological related data. Based on the effectiveness of meteorological input data, this paper selects the daily meteorological data of 35 monitoring stations in Guiyang from 2019 to 2020. Analyzes the correlation between environmental data and six pollutant concentrations through Spearman Rank Correlation Coefficient; the ventilation factors and stable energy proposed from the perspective of energetics are added to the input elements to predict the concentration of pollutants. The experimental results show that the Long Short-term Memory network (SA-LSTM) model based on self attention mechanism with ventilation factors and stable energy is better than other existing models in air quality prediction.

Aimed at the characteristics of the diversity of traffic road targets and the complexity and variability of detection scenarios, it is difficult for general target detection algorithms to find an ideal balance between accuracy and speed, resulting in slower detection speed and higher false detection rate. An improved YOLOv4 traffic road target detection algorithm has been proposed to solve these problems. Firstly, a Context Exploitation method is introduced to reduce information loss at the highest level feature map in FPN. Secondly, the residual feature augmentation method is adopted to enhance the feature extraction of the convolutional layer of the YOLOv4 neck, which greatly improves the detection speed and also obtains an increase in accuracy. Finally, the augmented PANet is used to improve the feature fusion method and enhance the representation ability of the feature map. Compared with other classic methods on the VOC and TT100K datasets for road target detection, it is found that the improved YOLOv4 algorithm can effectively reduce the false detection rate of small targets and significantly improve the accuracy and speed of detection. Experimental results show that the improved YOLOv4 algorithm has an average accuracy of 2.42% higher than original YOLOv4 algorithm on detection, and the detection speed reaches 61.5 frames/s.

This paper summarizes and analyzes the technical and operational pain points faced by content auditing in the field of radio, television and network audio-visual, and puts forward the content auditing blockchain in combination with blockchain and artificial intelligence technology, so as to solve the current dilemma of content auditing from a technical perspective, improve the efficiency of content auditing and relieve the pressure of content broadcasting institutions.

In the long-term observation of geomagnetic field, the levelness of coil vector magnetometer will affect the measurement of magnetic direction. Therefore, we propose an electrical detection method of instrument non-levelness, which realizes the quantitative calculation of non-levelness by measuring the maximum and minimum values of the resultant magnetic field during detection. Then, a horizontal error model of coil vector magnetometer is constructed, and the mapping from instrument non-levelness to declination measurement error can be realized by vector operation. The results show that when the non-levelness of the instrument is less than 10", the measurement of magnetic declination variation deviates from the real value by 3.4%; When the non-levelness of the instrument is less than 20", the measurement of magnetic declination variation deviates from the real value by 13.5%.

With the rapid development of Virtualization technology and networks, Virtualization has become an important part of enterprise IT construction. Not only the original business system and the virtualized environment co-exist in the enterprise data center, but also the traditional software licenses based on hardware characteristics need to be migrated to the virtualized environment. This poses a challenge to traditional software authorization methods. This article analyzes the software authorization scheme in the distributed environment, hoping to provide help and reference for improving the traditional software authorization method.

Based on the historical data of Jiaxing Port, this paper uses ARIMA-RBF to predict the monthly cargo throughput of Jiaxing Port, builds an ARIMA-RBF model, and finds the optimal network parameters through training. On this basis, the ARIMA-RBF model is used to predict the cargo throughput of Jiaxing Port, and the actual data analysis is used to compare the ARIMA-RBF neural network, grey system theory and ARIMA prediction results. Finally, the prediction results of different models are compared to prove that the ARIMA-RBF neural network prediction is the optimal prediction model.

In image analysis, often on a comparison of geometric parameters of the image analysis, including width, area, etc., usually adopt the method of minimum circumscribed matrix splitting measurement, measurement of extracting the smallest rectangle of blade first commonly, because of the plant leaf surface is rough and irregular, also present irregular edges and corners, on the edge of this requests us in the course of processing, The parameters of the minimum peripheral matrix and the algorithm adopted have certain characteristics. Nowadays, target rotation method and vertex chain code method are usually used. But previous algorithms required high-speed computing. On the basis of the original, an optimization algorithm is proposed to analyze the plant leaves.

For super high-rise buildings, reasonable structural layout and design optimization help reducing the engineering costs while ensuring structural safety. However, structural designers generally create structural models through manual modelling, which is inefficient and difficult to compare and optimize structural design schemes. In order to solve this problem, after combing through the structural logic of super high-rise building structure, this research adopts the parametric design mode, and proposes an integrated design framework for parametric modelling, analysis and optimization of super high-rise building structure, and developed the design tool on the Grasshopper software. Using this tool, you can quickly create a three-dimensional structure model, and call Etabs software for structural calculations. The tool can also be used in combination with evolutionary optimization algorithms to construct a parametric optimization process for structural optimization. The effectiveness and practicability of the developed design tool is verified through a super high-rise building with frame-core tube structure. The illustrative example has been shown that the tool can help engineers quickly create and modify structural models, and the concrete material consumption can be reduced by 40% after structural optimization. The tool and design process can be extended to similar construction projects to improve the design efficiency of structural engineers and reduce the construction cost of the project.