Search

The metasurface based on dielectric nanoantennas has lower ohmic losses than its counterpart with metallic nanoantennas. Here by utilizing coupled leaky mode theory(CLMT), the resonance mode of the unit cell composed of rectangular silicon nanorods with variable geometric parameters is proposed and then theoretically analyzed. Based on the array of unit cells, a silicon metasurface with the properties of broadband around 97.17 THz and high transmission efficiency over 70% are achieved at the frequency range from 400THz to 660THz. Furthermore, a new strategy is proposed and then adopted in this paper to alleviate the crosstalk between the adjacent cells. Owing to the characteristic of broadband resonance and linear phase modulation, the metasurface with the silicon nanoantennas could hold a great potential for flat optical devices in the visible frequency range.

The 5G wireless network which is regarded as the future critical infrastructure, is faced with many formidable challenges. To deal with them, resource sharing is the trend, including sharing the wireless resources and the infrastructure. Thus, the wireless network virtualization has been raised to integrate and abstract the resources and make network management more flexible. In this paper, a wireless network virtualization architecture is designed, which consists of the cognitive plane, the control plane and the data plane. These three planes collaborate to optimize the resource allocation according to the network condition. Then, we introduce the cell clustering method in the proposed architecture, which can largely reduce the complexity of resource slicing and make it possible in wireless network. Finally, we have implemented a trial environment to test the architecture. The test data indicate the network performance is obviously improved. Spectrum efficiency is doubled and the packet loss rate reduces to 1#x002F;20.

Android is paid more and more attention by many mobile phone manufacturers and software vendors. Due to defects of the Android and the huge potential economic benefits, there are more and more malicious codes. The majority of malicious applications will exhibit malicious behavior only if they interact with users. However, there is not a mature solution to traverse APKs automatically currently. By studying and analyzing the framework of Android system, we design and implement a system called APK-DFS which can traverse APKs automatically. This system can extract and recognize views in UI pages, and interact with these views via depth-first-search algorithm layer by layer; it establishes a UI storage stack and a UI trash can; it can also generate strings with specified format for views which require text input. We evaluate the system by testing it with APKs downloaded from Android markets. The results show that APK-DFS can simulate real users to trigger views in APKs effectively. For APK-DFS, in 30 min the average of effective trigger rate is 91%, and the average number of views that can be triggered in 50 events is 32.58. Compared with Monkey and PUMA, APK-DFS is the best one.

At present, Android malwares become more and more subtle and intelligent, after their invasion, they often detect whether the running environment is a real environment, to decide whether to perform their malicious behavior. Therefore, malware tend to execute different behavior when running in different environments. Benign applications will perform the same functions in different environments, their behaviors have a strong consistency. Based on this basic idea, we design an Android malware detection method based on behavior comparison analysis. First, design and development a number of specific different running environments, and then execute application in these environments. With the same event input, record and compare the behaviors of this application, calculate the difference, determine whether it is malicious. Under the guidance of this thought, we design and development the Android malware detection system EmuProtect. We evaluate EmuProtect system from the aspects of accuracy and validity, the results show that this system can effectively detect Android malicious applications.

An important aspect of research on integrating wind farms is the analysis of short circuit current contribution to the power grid. In this paper, the fault related features of doubly fed induction generators (DFIGs) are modeled using low voltage ride through (LVRT) test data sets. The dynamic behavior of DFIGs after fault occurrence is represented by a typical curve that is obtained using a curve clustering technique — the backward scenario reduction method. Then, two fault features (the maximum value of the short circuit current termed as peak current and the time to reach it), which are important for protection relay settings, are collected and analyzed using the probability density functions (PDFs). Two cases are considered in the analysis and some discussions are presented in the end.

Tensegrity traced back to the early 20's, the characteristics of the structure are potentially extensive applications in geometry and mechanics. In 2015 NASA researched out a tensegrity robot called SUPPERBall. This robot has been greatly improved compared to the previous tensigrity robot. However, they just focused on the design of the space structure of tensegrity robots to adapt more space constraints. The SUPPERBall robot uses the strings structure as the driving part, so the driving ability is limited. In this paper, a new tensegrity use the bar structure as the driving element and use the springs instead of string structure to make the whole robot have a good elastic performance and a good buffer in the condition of the impact force. This paper mainly does the following work. Firstly, according to relationship among nodes, connection matrix of the bars and the strings are built and then the mathematics model of tensegrity robot is obtained. Secondly, dynamics simulation of tensegrity robot is established in Adams software to analysis the movement locus of center of gravity and change in length of bars of tensegrity robot. Finally, it is proved that the application possibility of tensegrity theory used in the field of robot.