Abstract：Modulation recognition of radio signals plays a vital role in radio monitoring and spectrum management. As the deep learning network in artificial neural network has the powerful ability of representation learning which can automatically extract various complex features from the original data, exploring the modulation identification of radio signals based on deep learning is one of the main development trends in the field of radio monitoring. This paper introduces some application results and existing problems of deep learning in radio signal modulation recognition. Combined with the actual needs of the work, this review puts forward some ideas for deep learning in the modulation recognition of radio signals, such as further improving the recognition range and the recognition accuracy, especially at low SNR; seeking some new deep learning hybrid architecture for radio signal modulation recognition.

Abstract：Face recognition technology is an important research field for deep learning. In order to overcome the shortcomings of traditional open-loop face cognition mode and deep neural network structure, and to imitate human cognition model of real-time evaluation of cognitive results to self-optimized regulate feature space and classification cognition criteria, drawing on the theory of closed-loop control theory, this paper explores an intelligent face cognition method with deep ensemble learning and feedback mechanism. Firstly, based on the DEEPID neural network, an unstructured feature space of face images with a determined mapping relationship from the global to the local is established. Secondly, based on feature separability evaluation and variable precision rough set theory, a face cognition decision information system model with unstructured dynamic feature representation is established from the perspective of information theory, to reduce the unstructured feature space. Thirdly, the ensemble random vector functional-link net is used to construct the classification criterion of the reduced unstructured feature space. Finally, the face cognition result entropy measure index is constructed to provide a quantitative basis for the self-optimization adjustment mechanism of face feature space and classification cognition criteria. The experimental results show that the proposed method can effectively improve the recognition rate of face images compared with the existing methods.

Abstract：In order to verify the assume that stock price movement is similar to the past，pricing movement is simply dividend into up and down by K-Nearest Neighbor algorithm for forecasting. Sliding window method is used for comparing which historical period is more similar to the current in data feature. Multiple KNN models construct ensemble models for the strategy generalization and return adjustment. The CSI500 price is used for verification. With the predication, single KNN model wins 76.72% return with fee return from 2017 to Sep. 2018，remote historical period is more similar to the current in data feature，and ensemble models are better in risk control. This model verifies the stock price is similar with K-Nearest Neighbor character, which could be used as an investment timing strategy.

Abstract：Due to the influence of the process, actual frequency of clock generator module always has some deviations from target, it is necessary to trim the clock module whose frequency is not as it was expected. Compared with traversing method, dichotomy has higher convergence rate. Auto trim circuit was designed based on dichotomy, and measurement error caused by synchronous of referenced pulse was analyzed. Auto trim circuit counted output of clock generator module during high level of referenced slow pulse which was generated by automatic test equipment, and trim value was changed using dichotomy, achieving the purpose of saving measure time.

Abstract：A power Darlington array layout structure is designed based on the 70 V high voltage bipolar process. The structure is suitable for the output stage layout of the single-chip high-voltage power device. It consists of symmetric interdigitated array unit structure, and has the characteristics of large current driving capability, uniform heat dissipation, good stability, and high area utilization. The interdigitated-emitter and interdigitated base ensure a high speed while taking into account the matching and symmetry of the array power transistor. The two-sided design of the interdigitated base increases the effective emitter perimeter and improves current conduction capability. The improved equi-plane wiring reduces the wiring steps and makes the high current layout possible. Simulation based on calibre proves the layout has the ability to output current of 9 A.

Abstract：The accuracy of symbol timing synchronization has a great impact on the demodulation performance of digital communication systems. The dPMR communication system requires the symbol synchronization of the receiver to have fast acquisition and good tracking performance. A timing estimation algorithm is proposed for this requirement. The algorithm combines the advantages of preamble timing algorithm and digital square filtering algorithm. Firstly, when the preamble of the burst information is captured, the preamble timing algorithm is used to implement high-precision fast timing estimation. Then, a digital square filtering algorithm is used to implement tracking correction for timing estimation at intervals of 384 symbols. At the same time, a simple FPGA implementation scheme is proposed. Compared with the classical matching filter timing algorithm based on synchronous waveform, it not only improves the demodulation performance of the receiver but also saves hardware resources.

Abstract：In order to improve the unified shading processor efficiency of the mobile graphics processor and reduce the number of accesses between it and the off-chip memory, this paper presents a four-port texture cache architecture. This architecture uses texture mapping based on Mipamp algorithm and the storage mode of different single-port cache is selected based on Level of Detail(LOD), which improves the hit rate of texture cache.In addition, in order to improve data throughput, four ports supporting 4 parallel reading are designed. FIFO buffer is designed for prefetching memory data and reducing memory latency. Using SV to build an experimental platform to test texture images，the results show that the average hit rate is about 92.5%, and the rate of data throughput is nearly 4 times that of the single port cache.

Abstract：This paper introduces a high-speed, low-offset sense amplifier with low supply voltage in UMC 28 nm CMOS process. The paper presents a novel structure of the sense amplifier which bases on the traditional differential amplifier, class AB latch and other circuits. It′s designed and verified in Cadence. The simulation results show that the proposed design exhibits 0.2 mV/0.8 mV offset voltage, 63 ps/44 ps delay, 0.37 mW/0.44 mW power dissipation respectively with 1.05 V supply voltage when the clock signal at the 5/10 GHz. Therefore, the proposed sense amplifier is satisfied for the analog-to-digital converter of high-speed interface JESD204B.

Abstract：Aiming at the limited fault diagnosis and performance monitoring capabilities problem of satellite power control unit (PCU), the testability design is done for the PCU and the on-orbit fault diagnosis and performance monitoring system based on BIT technology is built. The BIT module consists of three analog boards and one core board. The analog board can perform high-speed sampling of dozens of analog signals, while the core board can run fault diagnosis and performance monitoring algorithms, and other works such as data storage, data packaging, and communication. The result of fault diagnosis and performance monitoring is sent to the communication interface after packaging. The verification experiment shows that the system can accurately detect and locate the fault of the PCU fastly without affecting the normal operation of the PCU. Meanwhile the performance of the PCU, such as bus voltage ripple, can be monitored.

Abstract：Nowadays, the performance requirements of time unified system in various fields are getting higher and higher. In the past, most time unified system used NTP for time synchronization, but the protocol can only provide millisecond-level synchronization accuracy, and the master clock is generally fixed, which cannot meet the requirements of high precision and high robustness in application fields such as measurement and control. Based on this, this paper conducts an in-depth study on the application of high-reliability time unified system based on PTP timing. The system adopts the PTP protocol for timing, which can make the system time synchronization precision reach nanosecond level, and based on the BMC algorithm，the system is designed and implemented the double mode of manual setting and automatic selection for the dual-system hot standby function, which satisfies the high-performance requirements of the time unified system in the application fields such as measurement and control. Compared with the previous time unified system, the system greatly improves the timing accuracy and robustness of time unified system.

Abstract：Aiming at the complex problem of telemetry time series data interpretation, a new method was proposed which can consistently analyze and judge the telemetry data in launching tasks of vehicles. By establishing Person, Spearman, and Kendall correlation coefficients, the relevant analysis approach will obtain the assessment results based on the evidence theory. The experimental results show that the results of data consistency are correct, which can significantly improve the efficiency of data checking and comparison, change the current situation that data interpretation relies on manual and data judgment methods, and play an important role in data rapid assessment and analysis, especially in fault analysis.

Abstract：With the development of energy Internet and the implementation of "cloud, big, material and mobile" strategy, the traditional centralized access authentication mode of power Internet of Things terminal is difficult to meet the ubiquitous application needs. Based on the super-ledger technology of alliance blockchain, this paper studies the distributed access authentication technology of power Internet of Things terminal, combines the key sharing and distribution protocol in secure multi-party computing with the super-ledger consensus algorithm, and proposes an effective access authentication scheme for power Internet of Things terminal. The simulation results show that the proposed scheme has advantages in authentication time and resource consumption.

Abstract：RS422 is widely used in industrial and defense communication interfaces due to its low cost and high reliability. However, due to its short transmission rate and short transmission distance, its application range is limited to low-speed, short-distance transmission. Aiming at such problems, this paper analyzes the reasons for restricting the transmission rate and distance of RS422, and proposes a method to increase the communication speed of the interface in software. This method is that we perform 6B/10B encoding on RS422 data, and increase the level conversion to make the processor better edge judgment. The DC component in the traditional 8B/10B encoding is removed from the encoding to achieve unique encoding and has one bit error correction function. Experiments show that by increasing the coding method, the transmission rate of RS422 is closer to the theoretical value, which has certain reference significance for improving the throughput of RS422 in engineering.

Abstract：The security of traditional wireless communication relies on upper layer encryption and cannot protect the security of the physical layer. In order to overcome the shortcomings of traditional wireless communication system encryption, artificial noise generation and random precoder are used to study the secure transmission method of large-scale multiple input multiple output(MIMO) downlink system under relay cooperation. The estimated channel state information is derived and the rate expression is derived,and then the confidentiality rate on the user node is analyzed. The performance gap is analyzed and compared for the minimum mean square error precoder(MMSE) and zero forcing(ZF) precoder at the base station. Theoretical analysis and simulation results show that this method can design a secure physical layer transmission strategy for large-scale MIMO relay systems, and the MMSE precoder exhibits better security performance than the ZF precoder.

Abstract：A GPS tracking algorithm is studied and applied to locate a spinning vehicle on which a single-patch antenna is mounted. This algorithm is able to track global positioning system(GPS) signals with intermittent availability at a lower rotational speed, but with enough accuracy to determine the rolling rate. Firstly, the received signals of single rolling antenna is analyzed and modeled. Then, based on the fact that the received signal has amplitude modulation information relative to the rolling characteristic, demodulation is performed using the symmetry of the rotational amplitude to detect the rolling rate and angle of the projectile. Actual test results show that，for using amplitude information of I/Q signals，rolling rate and rolling angle of spinning vehicle can be calculated in real-time by using proposed method，and calculation accuracy can meet requirements of most practical applications.

Abstract：Be aimed at application requirements of VLSI(very large scale integration) in space environment, current researching status in China and abroad of radiation harden of FPGA are introduced in the paper. With a simple description of space radiation environment and SEE(single event effect) and analyzing the structure and fault characteristics of SRAM-based FPGA, a hardware platform design of single event effect mitigation for configuration, monitoring, readback verification and Scrubbing of Xilinx Kintex-7 series FPGA based on HRU(high reliability unit) is proposed. The process of protecting Kintex-7 series FPGA and the composition of SEU injection test system are elaborated. The platform is used in a project and has passed functional testing and related environmental tests, and it has provided a design reference for space application of VLSI.

Abstract：Edge computing is a computational model for all terminal devices from cloud to edge path, which emerges in the era of interconnection of all things. The existing server pattern under cloud computing model cannot satisfy the application requirements of edge computing model. It mainly embodies in the limitation of edge device resources, the real-time nature of edge device services and the intelligence of edge device decision-making. Aiming at the characteristics of application requirements of edge computing model, this paper proposes a service pattern suitable for edge computing model—computing as a server pattern(CaaS)，and abstracts the task model, computing model and service model provided by this pattern.

Abstract：The training of large-scale corpora is an important basic work for the automatic detection of Chinese texts using the trigram N-gram algorithm. Faced with up to one million pieces of data to be processed by the new media platform per day, there is a computational bottleneck in the construction of a single-node trigram N-gram language model lexicon. Based on the deep research of the trigram N-gram algorithm, the idea of trigram N-gram parallelization algorithm based on MapReduce programming model is proposed. In the MapReduce programming model, the arithmetic tasks are evenly distributed to m nodes. The main task of the trigram N-gram algorithm in the Map function part is to calculate the number of times the local words are matched with the first two words, while the main part of the Reduce function，its task is to merge the number of occurrences of the statistical word matching in the Map part to generate global statistical results. The experimental results show that the MapReduce-based trigram N-gram parallelization algorithm running on Hadoop clusters has good performance and scalability. For a 12 billion word-per-day training corpus data set, the algorithm is obtained in a cluster environment. The rate of training results is more linear.

Abstract：Based on the TensorFlow framework, this paper builds convolutional neural networks to recognize the defects in the electroluminescent images of cells. It selects the exposed data set that contains the different types of defects in the cell. Based on the traditional VGGNet network, the full convolution neural network is used for training, and this paper analyzes the training effects of different loss functions and dropout probabilities on data set. Experiments have shown that the algorithm accurately recognizes whether the cell is defective. The study also shows that the compression network structure greatly increases the training rate of the algorithm, which makes the simplified model more portable and provides an effective solution for a wide range of real-time defect recognition.

Abstract：The single channel theoretical bandwidth of the UM-BUS can be up to 200 Mb/s. When 16 channels are used for concurrent transmission, the theoretical bandwidth can be up to 400 MB/s. So the test system needs to establish a communication channel of no less than this bandwidth between the data acquisition terminal and PC. In PCIe1.1, the theoretical bandwidth of the four channel transmission is up to 1 GB/s, which meets the transmission requirements of the UM-BUS test system. Thus, this paper realizes the application scheme of PCIe1.1 x4 link channel of UM-BUS bus test system, and gives the BMD transmission scheme of PCIe bus based on FPGA. The test result shows that the actual transmission speed of the scheme can reach 550 MB/s, which satisfies the bandwidth requirements of the UM-BUS bus test system.

Abstract：With the large-scale emission of automobile exhaust, in order to prevent pollution of the atmosphere, it is more and more important to detect the exhaust gas. In order to automatically detect the blackness of automobile exhaust, this paper designs and implements a system for detecting Ringelmann number using image processing method for automobile exhaust. The Ringelmann number of automobile exhaust is obtained by photographing the automobile exhaust. At the same time, the detection result can be sent to the mobile phone APP through the 4G network, and the mobile terminal can also control the detection system to take photos, get analysis results and print the report. This research provides a feasible solution for vehicle exhaust blackness detection, which has certain practical significance.

Abstract：In order to meet the requirements of spacecraft miniaturization design, a single chip anti-fuse FPGA is used as main control chip for driving relay antenna. In order to meet the high precision requirements of antenna pointing, dual-channel rotary transformer with AD2S80 decoding is used to obtain position information data. In order to improve the speed stability of the mechanism, the stepper motor 256 subdivision sinusoidal PWM drive mode is adopted. Communication, position, current acquisition, fault error mechanism and stepper motor drive function are set in one, drive machine weight is reduced from 5~6 kg to 2.5 kg, and pointing accuracy increases to 5.5 degrees per thousand. The experimental results show that the drive control of digital relay antenna based on single chip FPGA can effectively improve the integration of machine and the smoothness of motion process, and also has great significance and application value to the improvement of satellite comprehensive performance.

Abstract：Aiming at the technical defects of the existing helium mass-spectrometry leak detection system, a design and implementation of the helium mass spectrometry leak detection system based on the Internet of Things technology is proposed. The system integrates embedded system, WiFi，ZigBee and consists of main control module, acquisition module and execution module. The main control module can integrate the data information of the helium mass spectrometer leak detector and the acquisition module. The execution module automatically completes the leak detection process, and predicts the stable time of the leak detection signal by gray algorithm, which can improve the leak detection efficiency. The test results show that the implementation scheme is effective.

Abstract：In order to solve the back corona phenomenon caused by high specific resistance dust in electrostatic precipitation process, the circuit topology of DC base voltage superimposed pulse voltage is adopted. Firstly, for the DC base voltage part, a three-phase AC voltage regulator circuit and a subsequent filter circuit are established, and the average values generated by the circuit at different firing angles are analyzed. Secondly, in the pulse generation part, the LCC resonance is discussed according to the mathematical model. Finally, according to the design requirements, the whole circuit is simulated with Simulink module in MATLAB, and compared with the calculated data, it has a high consistency, which verifies the correctness of the design idea.

Abstract：The tire pressure sensor is a wireless sensor that has been used in automobiles for many years, but the power supply has always been a bottleneck restricting the service life of the sensor. The article proposes a method for collecting heat energy in the environment, designs an energy harvesting circuit and achieves self-power supply of the tire pressure sensor. The thermoelectric generation(TEG) is used to convert the thermal energy accumulated inside the tire into electric energy. An ultra-low voltage electric energy collecting circuit is designed based on the LTC3108 power management chip. Experiments have shown that the designed circuit can collect energy as low as 100 mV, 10 mA, 1 mW, and successfully drive the tire pressure sensor after a period of accumulation.

Abstract：Aiming at the problems of difficult adjustment of resonant frequency and difficult debugging in traditional atomizing circuit, a new atomizing circuit with multi-frequency and multi-circuit controlled by program is designed.The circuit uses STM32 as the main controller, and generates a frequency-adjustable external driving signal by frequency synthesizer based on AD9833 programming.The switching function of the MOS tube is driven to control the input of the external signal, so as to excite the oscillator oscillation.Sampling feedback circuit is used to find the optimum resonant frequency, which improves the stability of atomization circuit, realizes the real-time frequency adjustment, simplifies the circuit structure and reduces the difficulty of debugging.

Abstract：OLED display device has the characteristics of spontaneous light, short response time, low power consumption, wide color gamut, wide operating temperature range, good seismic resistance and thinness, and has been developed rapidly in the field of small and medium size display.In this paper, the circuit design of the airborne color OLED display module is studied, with a 7.7-inch color OLED display as the display medium and FPGA as the main control chip, which realizes the functions of video display, brightness adjustment and night vision compatibility.

Abstract：According to the basic characteristics of inductive components and the mechanism of suppressing electromagnetic interference caused by the magnetic cores, the influence on the impedance of magnetic cores of the material type，core size and the number of turns is introduced in this paper. Meanwhile, an experiment which was based on the outdoor union of an inverter air-conditioner was designed to illustrate the relationship of impedance characteristics of a magnetic toroid and its ability to suppress electromagnetic interference. That is to say, the higher its impedance in the application is, the more it will suppress; furthermore, it needs to be evaluated according to the actual situation in the process of use because of the impedance characteristic difference between the test data and the data supplied by the manufacturer of a magnetic core, which can provide guidance for the rectification of Electromagnetic Compatibility test.