Abstract：This article mainly introduces the current status of 5G antennas, and shows several mainstream smart terminal antenna design methods in the past decade. Based on MIMO antenna, frequency resettable antenna and slot antenna, their working modes, principles and corresponding advantages and disadvantages are explained, and two important technologies of antenna decoupling and bandwidth expansion are introduced in detail. Finally, the future development trends and possible improvements of 5G antennas are prospected.

Abstract：This paper reviews the reliability of TSV(through silicon via) in three aspects: thermal stress, process and piezoresistive effect. The thermal stress reliability of TSV is reflected in the great difference of thermal expansion coefficient between different materials, too large thermal stress may lead to interface delamination and cracks; the reliability of TSV process is reflected in the continuity of side wall and the quality of filled copper; the carrier mobility in the active region will be affected by the thermal stress of TSV. A KOZ(keep out zone) area around TSV is defined. The KOZ is set to the region where the carrier mobility does not exceed 5%. When the carrier mobility exceeds 5%, the timing of the circuit may be destroyed and the integrated circuit will fail.

Abstract：Swarm intelligence has significant advantages in solving nondeterministic polynomial(NP) problems or problems with too much search space. In this paper, pigeon inspired optimization(PIO) is applied to the feature selection of intrusion detection systems. The Sigmoid-based PIO(SPIO) and Cosine-based PIO(CPIO) algorithms were proposed to select the features of the intrusion detection data set KDDCUP99 and conduct experiments with the method of machine learning to build the model and evaluate the results.

Abstract：In order to extract user requirements effectively in the field of car styling intelligent design, a method for extracting car styling requirements based on multi-feature TFIDF(word frequency-inverse file frequency) text analysis is proposed. Firstly, a large number of unregistered professional vocabularies is obtained through mutual information and boundary degrees of freedom to optimize the vocabulary after simple word segmentation. Next, in order to solve the problem of the limitations of the classic TFIDF algorithm, vocabulary and emotional feature factors are introduced to get user demand feature candidates set. Finally, effective user needs are obtained according to the threshold. The experimental results show that the multi-feature TFIDF text analysis algorithm has certain advantages in feature extraction, and can effectively extract user needs of the car styling in the text.

Abstract：Most of the existing researches on accelerators of binary convolutional neural networks based on FPGA are aimed at small-scale image input, while the applications mainly take large-scale convolutional neural networks such as YOLO and VGG as backbone networks. The hardware of convolutional neural network is optimized and designed from the two aspects including the network topology and pipeline stage, so as to solve the bottleneck of logic resources and improve the performance of the binary VGG network accelerator. CIFAR-10 dataset resized to 224×224 was used to verify the optimized design of VGG convolutional neural network accelerator based on FPGA. Experimental results showed that the system achieved 81% recognition accuracy and 219.9 FPS recognition speed，which verified the effectiveness of the optimization method.

Abstract：In order to meet the special requirements of portable and wireless monitoring device for pilot in daily simulation training, a wireless EEG monitoring system based on brain wave module TGAM is designed and implemented, which adopts the Nordic 2.4G ultra-low power wireless platform to realize wireless transmission between the acquisition equipment and the monitoring computer.The system has the characteristics of low power consumption, miniaturization and easy to use. It can monitor the EEG of the pilot during the simulated training in real time and feed back the indexes of attention and meditation.

Abstract：The current situation of oil pipeline engineering site management is investigated, and the current site management problems are analyzed and identified. At the same time, the site management requirements are obtained. Based on the requirements, intelligent site architecture is constructed based on the cutting-edge technologies such as Internet of Things, cloud computing, big data, artificial intelligence and BIM, which realizes automatic data collection and processing, construction process visualization, real-time monitoring of construction process and cooperative work of all participants. Cloud computing and big data technology are used to integrate, store, analyze and mine data, which can assist managers in decision-making. Using BIM model to integrate all kinds of information, and working with multiple personnel, it can achieve integrated management of project objectives, change the existing site management mode, and promote the existing site to gradually move towards standardization, science and intelligence.

Abstract：In order to solve the problem of pet lost and difficult to retrieve, a design of pet anti lost wearable device based on Internet of Things technology is proposed.In the short distance of the system, through the low-power Bluetooth 4.2 module, the WeChat applet of mobile phone can control and play the music of wearable device and light up the colorful led to prompt the pet′s position. At the same time, with the help of the speech recognition module, the wake-up system can be used, and the user can also complete the above operations by saying the secondary keywords.Outside the Bluetooth signal range, the user can check the location and speed information collected by the GPS + GPRS location module on the pet wearable device in the WeChat applet. The experimental results show that the system has perfect functions, the Bluetooth control distance of WeChat small program end reaches 80 m, the GPS positioning accuracy error of wearable device is controlled within 2 m, and the voice recognition function is stable, which can help the owner to find and control the pet in time，and avoid the potential safety hazard and property loss caused by pet loss.

Abstract：With the rapidly increasing interests in ocean observation and ocean Internet of Things, low-cost underwater acoustic communication equipment draws more and more attention from the research community. The low complexity and high reliability multi frequency shift keying(MFSK) modulation provides an effective technical solution. However, the typical time-space-frequency characteristic of the underwater acoustic channel will cause severe amplitude fading to the MFSK underwater acoustic communication. Considering that the chip MAX9814 has a large-range automatic gain adjustment capability, it is adopted for anti-fading processing of underwater MFSK communication. The experimental results show that MAX9814 can effectively suppress the amplitude fading caused by the underwater acoustic channel, thus improve the performance of the underwater acoustic MFSK communication.

Abstract：For the problem of computing offloading decision in mobile edge computing, this paper proposes an offloading decision algorithm based on enhanced learning in multiuser MEC system. According to the task model, calculation model and channel state, ODRL algorithm makes the task unloading decision and uses reinforcement learning method to solve the optimal computing unloading strategy. Simulation results show that the ODRL algorithm proposed in this paper has lower total system cost compared with the baseline strategy.

Abstract：With the development of the Internet of Things technology, the field of smart home has risen rapidly, and has bright prospects for development. The design and implementation of the intelligent Magic Mirror system in the smart home industry is designed to enrich and optimize the quality of life at home and enhance the taste of the home environment. The intelligent Magic Mirror uses the Raspberry Pi based on the Linux environment as the hardware platform, and adopts the three-layer architecture design of the Internet of Things. It uses the hardware devices at the perception layer to obtain information and collect, and the HTTP/2 at the network layer calls cloud API interfaces and other communication technologies to implement data transmission and exchange, and the collected information and data are analyzed and fed back by the cloud computing technology at the application layer. The intelligent Magic Mirror system is designed and implemented to operate stably, and can well realize basic functions such as the main interface information display, voice interaction function, real-time monitoring function, and so on.

Abstract：MVC(Model-View-Controller) mode is the traditional way of Android application development (Activity/Fragment is used to represent the Controller layer and XML files are used to represent the View layer). As the number of project pages increases, the logic complexity increases, and the Activity file becomes bloated and the code coupling is obvious. Improvement is not conducive to the upgrade and maintenance of the project later. By studying and comparing the traditional MVC development model and the mainstream MVP(Model-View-Presenter) development model, it is found that the MVP development model can better solve the above problems. At the same time, the efficiency of the official HttpURLConnection class for HTTP network requests can not meet business needs, and the responsive network request framework of Retrofit2+OkHttp3+RxJava2 has higher response efficiency. Based on the "Changguang Satellite Cloud Extreme Vision" project as the background, this paper studies and verifies the feasibility of combining the MVP mode with the responsive network request framework of Retrofit2+OkHttp3+RxJava2.

Abstract：Aiming at the problem of left-behind children, empty-nest elderly, and people with reduced mobility, an autonomous family companion robot was designed. Based on the architecture of robot operating system(ROS), firstly, an autonomous mobile family companion robot was built with Jetson TX1 as the core, then the software was designed, which makes it not only able to communicate with people in natural language, but also use eye expressions and humanoid upper limbs for physical communication, and realize human-machine interaction through User Interface(UI). The rationality of the robot design is verified by the physical test, and the robot can meet the company work of daily family and continue to be developed to do more work.

Abstract：Text recognition in nature scene is currently applied in various intelligence equipment. The first step of text recognition is to precisely locate the text. In the Pixel Link text location methods, there are mainly two problems: too much background information is incorporated in the text region, and the test accuracy is insufficient. Aiming at these issues, an improved text location method was proposed to precisely locate the text in the natural scene. At first, an attention mechanism was incorporated into the original network. By focusing on the weight relationship between feature channels in the generated feature map, one can improve the weight coefficient of effective feature channels, and suppress the weight of inefficient or invalid feature channels. In the second, by changing the form of data set annotation, the coordinate points can be expressed as a series of sequence forms, so that the text lines can be framed adaptively in the LSTM model. At last, the located object is rotated according to the angle between a pair of vertexes in the polygon frame, and is subsequently fed to the text recognition interface to obtain the final character. Finally, the text detection test is carried out on the open data set and self-built data set. The experimental results show that the improved detection method is superior to the original method on different dataset, and the accuracy is similar to the current leading method.

Abstract：A new design of video stream encryption system is introduced，combined with the interleave technology，which is widely used in wireless communication. This paper provided a new key distribution and management system and a modified Output-Feedback encryption operate mode. The provided method keeps the structure of the transport steam packet, has a low protocol cost and a high level of security, compatible with the AES and SM4 encryption algorithm, thus can be easily implemented in hardware and widely used in wireless image transmission and digital video broadcasting system.

Abstract：Aiming at the existing problems of low real-time performance, unsatisfactory stitching effect and high cost in video stitching system at present, a real-time video stitching system was implemented on Xilinx Zynq platform using software and hardware collaborative design method after improving traditional ORB algorithm. The system used PS on Zynq to build an embedded Linux system, develop user interface programs, and implement task scheduling. The improving video stitching algorithm was packaged into an IP core that can be deployed into the PL on Zynq through Vivado HLS tools to achieve hardware accelerated processing. Experimental results showed that stitched video image had high clarity and no stitching gap, while this Zynq-based software and hardware collaborative processing architecture had greatly improved the speed of video stitching, which can meet the requirements of real-time processing.

Abstract：In order to ensure the real-time performance of the system and solve the problem that the perfect reflection method fails in the exposure environment, this paper proposes an FPGA-based dark channel automatic white balance correction algorithm. According to the light transmittance distribution of the dark channel of the color image, the reference pixels are selected to realize the automatic white balance correction of the color cast pixels. Experimental results show that the dark channel automatic white balance algorithm using FPGA correction can meet the real-time requirements of the entire system and eliminate the environmental impact of overexposure. At the same time, it can be effectively applied to dynamic scenes.

Abstract：The wireless charging technology can make the charger get rid of the restrictions of the line, realize the complete separation of electrical appliances and power supply, and have better safety and flexibility than traditional charging. However, at present, the wireless chargers in the market are expensive and thus are not conducive to promotion. Therefore, a wireless charging system with low cost and good charging performance is urgently needed to enter people′s lives. This paper provides a solution of the wireless charging SoC system based on off-chip Flash programming and loading, and elaborates the wireless charging system architecture. The problems of resetting watchdog and low power consumption in implementation are summarized, and the corresponding strategies are provided. Experiment results show that the proposed wireless charging system SoC system can meet the design requirements of miniaturization and low power consumption, and is able to provide good wireless energy transmission performance.

Abstract：The planar transformer has the advantages of small size, low leakage inductance, low conversion efficiency, and good heat dissipation performance. The design scheme of the planar transformer is very suitable for the design of ultra-thin switching power supplies. Small-sized ultra-thin switching power supply with high power density of 19.8 V(2.31 A) is developed using a planar transformer. Flyback converter is used as the main circuit and SG6742 is used as the control chip,parameter selection of transformer and PCB design of planar transformer are given. The circuit test results show that the flyback power supply has a wide input voltage of 88 V～264 V, the output voltage ripple is small, and the efficiency is greater than 87% under a wide voltage input range.

Abstract：In recent years, convolutional neural network plays an increasingly important role in many fields. However, power consumption and speed are the main factors limiting its application. In order to overcome its limitations, a convolutional neural network parallel accelerator based on FPGA platform is designed. Ultra96-v2 is used as the experimental development platform, and the design and implementation of convolutional neural network computing IP core adopts advanced design synthesis tools. The design and implementation of convolutional neural network accelerator system based on FPGA is completed by using vivado development tools. By comparing the recognition rate of GPU and CPU, the convolutional neural network based on FPGA optimized design takes much less time to process a picture than CPU, and reduces the power consumption of GPU by more than 30 times. It shows the performance and power consumption advantages of FPGA accelerator design, and verifies the effectiveness of this method.

Abstract：The planar integration design of the antenna system for soldier terminals is presented in this paper based on the requirements of the miniaturization and high integration. A four-antenna system including recieving and transmitting Ad-Hoc network antennas, satellite navigation and communication antennas is proposed to solve the 3D packaging and low-integration problems. The conventional 3D Ad-Hoc network antennas and circular polarized helix antennas are replaced by planar printed monopoles and horizontal dipoles with linear polarization, respectively. Simulated and Measured results show that the planar four-antenna system realizes onmidirectional radiation in horizontal plane in Ad-Hoc network frequency band and upper space coverage in satellite navigation and communication band. This antenna system can be applied in soldier terminals to improve its integration.

Abstract：For the test and transmission of rotating parts parameters, a design of a PPM modulation and demodulation system based on FPGA infrared data transmission is proposed, and a CRC check is added in the data transmission process, which increases the accuracy of the data transmission process. This design uses Altera(Intel)′s Cyclone10 series 10CL016YU256C8 chip, which has the characteristics of high speed, high bandwidth, high capacity, etc., suitable for high-speed data communication. Through the Quartus II development platform, it uses Verilog HDL hardware description language to write the corresponding modulation, demodulation and CRC verification programs. The paper introduces the algorithm of 4PPM modulation and demodulation, as well as the algorithm of CRC-CCITT check code, and carries on the timing simulation to the system through Modelsim. The simulation result verifies the correctness of the design.