Abstract：Object tracking is an important part in computer vision and is widely used in military, medical, security and autonomous driving. Although great progress has been made, it is still a challenging research field due to the factors such as occlusion, rapid speed, motion blur, illumination and deformation. In recent years, the correlation filter tracking method, one of discriminant type, has attracted much attention due to its higher processing speed. We first introduces the basic knowledge of the object tracking and the correlation filter tracking, and the correlation filter tracking methods in simple stage, we also discussed the circular structure and the kernel trick, the combination of multiple feature channels and deep feature, scale research, boundary effect and the use of other information. Finally, the research direction and development trend of the method based on the correlation filter is given.

Abstract：Nowadays, with the increasing complexity of digital circuits design, the post silicon debugging becomes a major challenging for engineers. For post silicon debugging, the lacking of observability is the main problem, the engineers can’t know exactly what happened inside the chip, even with the limited debug registers. Here we propose a method, which could provide much more observability compared to traditional methods. And with this method, the engineers could know what happened inside the chip exactly, and could even reproduce the problem in RTL simulation, in most of the circumstances.

Abstract：As the development trend of integrated, intelligent and low power application in embedded systems, ARM series processors have been widely used in recent years. Virtualization extensions, security extensions, and large physical address extensions supported in ARMv7′s Virtual Memory System Architecture(VMSA),This allows for hardware system support for the operation of multiple virtual machines or the implementation of partitioned operating systems.Complex embedded devices can use the new generation of embedded processors to complete system integration and miniaturization.This article focuses on the changes in the application structure of the memory management of extened VMSA in the ARMv7-A architecture.And also analyzed the application on airborne equipment.This will provides technical support for ARM processor used in airborne equipment applications.

Abstract：In order to solve the problem of limited time synchronization accuracy of payloads on spacecraft. A time synchronization method based on packet switching is proposed. The method realizes time synchronization of multi-payload with high reliability and high precision by step time compensation and frequency correction. At the same time, the network scheduling control is carried out to ensure that the influence of synchronization period and network background traffic on the time synchronization accuracy is reduced to the minimum. In the experimental verification, the implementation of this method is combined with the fibre channel network on the Xilinx K7 FPGA board. A nanosecond time synchronization is achieved with a small amount of hardware and network resources.

Abstract：Asynchronous FIFO, a basic structure in digital system, is widely used in digital design. The reliability of asynchronous FIFO can affect the entire system. This paper presents a foldable fault-tolerance approach to improve the reliability of asynchronous FIFO. Asynchronous FIFO can tolerate more failures with this method. In addition, in order to reduce the impact of faults on the depth of the FIFO, spatial redundancy is introduced on the basis of the folding method. Compared with the solution that uses only spatial redundancy, the proposed approach can tolerate 12 more failures, while the area overhead for fault tolerance only increases by 4.19% at most.

Abstract：Based on the analysis of High-Definition Multimedia Interface(HDMI)2.0 specification, the circuit of codec was designed with the hardware description language, and the prototype verification was implemented on the FPGA platform. The circuit of codec not only occupied less logical resource, but also compatible with HDMI1.4 and HDMI2.0 specification. The verification results show that the design can realized the functions of codec and meet the requirements of HDMI2.0 interface. It could be applied to system design.

Abstract：An ultra-wideband metal antenna with low radar cross section is proposed. The antenna adopts the metal Vivaldi as the radiation element, and utilizes a new-type frequency selective surface structure to reduce the radar cross section within the working frequency band. The frequency selective surface structure is composed of three layers of periodic planar structure and double layer medium, which has the characteristics of ultra wide bandwidth and controllable bandwidth. By changing the slot width and the width of the middle layer metal structure, can adjust the working frequency bandwidth and center frequency of the frequency selective surface structure, so as to realize the frequency bandwidth is adjustable, which can be used to reduce radar cross section of the antenna in different working frequencies.

Abstract：In order to solve the problems that the wireless meter reading system based on ZigBee technology has short transmission distance and weak anti-interference ability, a new remote meter reading system based on LoRa is designed, which is composed of routing module, repeater and terminal module. These three modules all use low-power RF chip SX1278. In the process of meter reading, 4 path detection command frames are proposed for the system. The system can relay the wireless signals at the concentrator to the terminal to extend the wireless communication distance. The test results show that the meter reading system is stable and reliable, and has wide range of meter reading and short networking time. It has broad application prospects.

Abstract：Existing power collection network concentrator doesn′t have ability to detect abnormal power load, which leads the main station of abnormal electricity behavior analysis and response lag. An information collection gateway with anomaly detection function is designed. Installing the gateway on the concentrator side to conduct anomaly analysis of the energy consumption data of the concentrator is an effective solution to the problem. According to the characteristics of the information collection network and the relevant grid regulations, the gateway constructs its own address by constructing the data frame to query the address pool from the master station. The abnormal electric behavior is judged based on the density clustering DBSCAN algorithm and the decision tree C4.5 algorithm. Test results show that the gateway can quickly configure its own address and conduct anomaly analysis of energy consumption data.

Abstract：In order to achieve high precision of the time and frequency synchronization of pseudolite automatic network，a pseudolite time and frequency synchronization system using the structure of Self-Calibrating Pseudolite Array(SCPA) is designed. The system takes FPGA+DSP as the core processor, and integrates high-speed A/D, D/A conversion circuit and up and down converter circuit. The design and implementation of pseudolite two-way pseudo distance measurement technology and the time and frequency synchronization scheme are introduced in detail. The test results show that the phase synchronization error of carrier phase is less than 0.1 Hz and the time synchronization accuracy is better than 2 ns, which provids a significant guarantee for the positioning and timing accuracy of the pseudolite automatic system.

Abstract：Obstacle detection, as an important part of intelligent auxiliary driving, has received extensive attention. Among them, manhole cover detection plays an irreplaceable role in subsequent manhole cover location and defect detection. This paper starts from a macro perspective, and proposes a classification method of multi feature fusion. It sets trapezoidal region as the region of interest with prior knowledge, locks candidate region through the connectivity of edge, and extracts well cover position from the background through multi feature decision quickly and accurately. The method is implemented on the TMS320DM6437 platform, and the feasibility of the scheme is verified.

Abstract：Aiming at the problem of IC card theft in the society, an IC card security defense system based on outlier detection is designed. The system collects the physical layer and application layer information of the IC card through RFID terminal and transmits the data to the cloud server. The cloud server uses the outlier detection algorithm based on integrated learning to process the data. If the IC card is abnormal, it prompts the user through the mobile APP. The test result shows that the accuracy of magnetic field signal anomaly detection sub-classifier and user habit anomaly detection sub-classifier is 0.95 and 0.65 respectively, indicating that the system can identify the pseudo card through effectively detecting the abnormal state of the IC card.

Abstract：In flight test, the test data captured is transported to ground system by telemetry link for real-time monitor, and the test data needs to be proceeded before telemetry transported. Using software defined radio technology, a flight test data relay management system was designed. The key parts of unit are fully detailed, including the FPGA,hardware circuit and the embedded software. The functions of processing and transmitting flying test data and management of wireless link are designed, meeting the application requirements by means of flight test.

Abstract：In this paper, according to the characteristics of the electrical components in the AC charging piles, the main fault types in AC charging piles are analyzed in order to diagnose and detect the possible faults. Extracting the relevant electrical characteristics of the main circuit and using the fault dictionary method，it detects and diagnoses the faults of AC charging piles. An experimental platform is built to simulate the various fault states and verify the effectiveness of the fault dictionary method for AC charging pile fault detection. The results show that the fault dictionary method has the good capability to detect the typical faults that may occur in AC charging piles.

Abstract：The conventional Real Time Kinematic(RTK) location technology utilizes radio modem to transmit differential data between base station and mobile station. In this case, the transmission distance is limited and the location precision can be easily affected by external factors. In this paper, a practical RTK location technology which adopts 3G wireless communication technology to establish data link between mobile station and Continusly Operating Reference Station(CORS) center is proposed. The technology is implemented in S5PV210 microprocessor and Linux embedded system. This paper introduces the system architecture and operating principle of the terminal, and elaborates the methods to implement the hardware module and software module. The system has characters of good reliability, high precision and strong practicability. Test results show that the terminal is reliable and robust with centimetre-level location precision.

Abstract：Interacting multiple model extended Kalman filter(IMM-EKF) algorithm is a sub-optimal algorithm which can solve the positioning problem in which the motion model is uncertain. But this method still gets sub-optimal solution and wastes computational resources when the carrier does the motion of which the model is certain. Aiming at this kind of defects of IMM-EKF, the off-line training probabilistic neural network model is adopted to judge the classification of current motion model in real time. We choose to operate with the single corresponding model when the motion model is in the state of certainty, and choose the IMM-EKF algorithm when the motion model is in the uncertain state. Thus it not only ensures the positioning accuracy, but also reduces the unnecessary computation burden. Simulation experiments verify the validity and accuracy of the algorithm, while the contrast test verifies the advantages in accuracy of the new algorithm compared with IMM-EKF algorithm.

Abstract：Mechanical model of magnetorheological(MR) fluid dampers is the basis of structure control of MR fluid dampers. For the Bingham mechanical model of MR fluid dampers, the shear strain rate is processed by fractional differential, a fractional order Bingham mechanical model of MR fluid dampers is proposed and used for image denoising, and a fractional order Bingham image denosing model is derived with definitions of fractional order edge detection operator and numerical implementation. Finally, the algorithm is used to denoise the synthetic image and the Lena image. MATLAB experimental results show that the algorithm can remove noise and preserve edges of object simultaneously.

Abstract：This paper proposed a scheme design of intelligent road card recognition system based on machine vision and deep learning. ARM9 is used as the part of front-end collection system,the image processing algorithm is used to complete the area extraction and segmentation of the road card text from the front-end acquisition image, the deep learning with a convolutional neural network is used to recognize the text in the cloud server, and the speech of the text information is presented to the user. Using the front-end hardware to get the road card images on the high way, the function of our system under CAFFE framework on the server is tested and the results show that the system can broadcast the road card information effectively to the user in real time.

Abstract：Aiming at the inconvenience and discontinuity of traditional cuff sphygmomanometer, a portable pulse wave sphygmomanometer is designed. The sphygmomanometer collects the pulse wave signal of the finger end through the photoelectric sensor，preprocesses the pulse wave signal and the blood pressure in a micro controller, and predicts and calculates the blood pressure value by training and optimizing a circulating neural network prediction model. Blood pressure values will be displayed on the OLED display, ultimately realizing real-time continuous blood pressure monitoring. After tested, portable pulse wave sphygmomanometer blood pressure prediction error is within plus or minus 5 mmHg, in line with international sphygmomanometer measurement error range. The portable RNN-based pulse wave sphygmomanometer designed in this paper provides a high reference value for the development and design of intelligent wear health monitoring equipment.

Abstract：In order to promote the devices belonging to different users to participate in cooperative communication in D2D networks, a time slot sharing cooperation strategy based on game theory is proposed. At first, the cooperative communication between devices which help each other to relay data is established based on sharing time slots. Secondly, the time slot sharing problem is analyzed by using Nash bargaining method of game theory. Meanwhile, the length of time slot which is used to relay the other devices is obtained by adopting Lagrange multiplier method. Thus, the problem that the devices how to cooperate is solved. Finally, the cut-off conditions of devices cooperation is derived, and the problem that devices when to cooperate is solved. Simulation results show that，compared with the direct transmission mode, the proposed strategy can improve energy efficiency of the devices effectively.

Abstract：A difference acquisition algorithm for Beidou B1 MEO/IGSO satallite is proposed by improving the acquisition algorithm of GPS signal.It can be used to solve the problem that NH code in Beidou B1 MEO/IGSO satallite makes the acquisition more difficult. Firstly,a new pseudo-random code is generated by modulating the Neumann-Hoffman(NH) code onto the C/A code to lengthen the coherent time. Then,the result of coherent integration is calculated by the differential accumulation method.Finally,according to the maximum criterion of the cumulative result, the sign of the differential term can be determined and the best combination can be obtained. Simulation result shows the gain can be improved 2 dB～3 dB by the algorithm.

Abstract：In this paper, the problem of Adaptive Modulation and Coding(AMC) is investigated in the Long-Term Evolution for Railway(LTE-R) communication system. An AMC strategy is proposed by introducing the Moore State Machine(MSM) model. According to the Modulation and Coding Schemes(MCS) adopted in the LTE-R system, the finite state set of the MSM is designed. Aiming at different modulation modes, the relationship between Bit Error Rate(BER) and Signal-to-Noise Ratio(SNR) is obtained. The SNR threshold of different modulation modes are derived by a given target BER. On the basis of these thresholds, the SNR thresholds corresponding to different MCS are achieved by plus and minus certain SNR constants. In order to overcome the frequent switching of the MCS in the vicinity of SNR threshold, the SNR threshold range of state transition is obtained by setting upper and lower bounds of SNR threshold. Based on SNR threshold bounds, the MSM is designed to dynamically adjust the MCS. The simulation results show that the AMC method presented in this paper has more stable spectrum efficiency and throughput than the traditional AMC method based on piecewise function. In addition, compared with the fixed modulation method, the proposed AMC method has better BER performance.

Abstract：For solving the problem that traditional routing protocols for mobile ad hoc network does not give consideration to data′s completeness and confidentiality, a secure routing protocol is proposed by combining data′s completeness protection with confidentiality protection. On the aspect of protecting data′s completeness, one way hash function and key are used to generate MAC, which is used to authenticate data′s completeness. On the aspect of protecting data′s confidentiality,μTESLA protocol is used to transmit cryptographic packets, deliver different key with different time interval, and execute a delayed disclosure of symmetric keys to simulate asymmetric mechanism, for reducing the resource consumption of authenticated broadcast. Simulation shows that the new method wins out over traditional routing protocols not only on the aspect of security, but also on the evaluating indicators including packer delivery ratio, routing overhead and end-to-end mean latency.

Abstract：Polar code is an error correction coding technique, which can prove from the mathematical point of view to reach Shannon′s limit. However, the traditional decoding algorithm, such as the successive cancellation(SC) and the successive cancellation list (SCL) decoding algorithms, have high complexity and latency in the decoding ends. This paper studies the principle and structure of the decoding algorithm, which proves that the decoding operation of some nodes is redundant. Therefore, simplified SC decoding and SCL decoding algorithm is proposed. It is proved that the simplified decoding algorithm can reduce the complexity of the decoding algorithm and maintain the original error rate performance at the same time.

Abstract：The zone routing protocol based on bus backbone networks can reduce network congestion caused by routing protocols such as common used table-driven routing protocols for urban traffic applications. However, there are a lot of redundant communication in the external cluster discovery phase of the routing protocol, which increases routing overhead. In order to reduce routing overhead, an improved regional routing protocol based on bus backbone networks is proposed. The improved routing protocol adds two fields to the link state update packet of the traditional zone routing protocol for storing the location information of the cluster head node and the destination node. In the external cluster discovery phase, the position constraint equation is constructed according to the position information of the cluster head node and the destination node. It is used to select the reasonable peripheral nodes to carry out data forwarding task and eliminate the redundant communication of other peripheral nodes. The simulation results show that the packet loss ratio, end-to-end average delay and routing overhead of the improved routing protocol are better than the zone routing protocol based on bus backbone network. The comprehensive performance metrics are also superior to two common used routing protocols such as AODV and DSDV.

Abstract：According to the wireless channel of typical hall environment in the direct and non direct situation, this paper puts forward an improvement of geometric scattering model reference model based on the assumption of scattering propagation environment. The probability density function of angle of arrival, the Doppler power spectrum density and the expression of time autocorrelation function are derived, and the impact of function variables on the statistical properties of the channel model is analyzed. In addition, according to the reference model, this paper designs an efficient SOC simulation model and two kinds of methods to calculate the parameters of channel simulation model, the performance of the two methods is compared with the Doppler power spectral density and time autocorrelation function. The result shows that the two methods for calculating the model parameters are accurate and efficient, and the statistical characteristics of the simulation model can be well matched with the statistical characteristics of the corresponding reference model.

Abstract：This paper proposes a convolution neural network（CNN） for image classification, which uses overlap pooling and dropout technology to solve overfitting problem. Compared with traditional CNN，the proposal obtains better results on the CIFAR-10 dataset，where the accuracy on testing data set is about 9 percent higher than that on the training data set.

Abstract：Missile-borne computer is a key component to realize the guidance and control of weapon system as a core unit of precision guidance weapon. With the development of weapons and equipment level, higher requirements were put forward to the integration of missile control computer in addition to complex control interface and the calculation of high performance. This paper introduces one design method of a new type of integrated control computer. The design based on the embedded real-time partition operating system AcoreOS653 adopts the integrated design idea of hardware and software and the PowerPC processor to provide high performance calculation ability. The design realizes the integration of the missile control computer with the effective integration of original multi-cabin computer′s complex cross-linked control. The hardware and software design and characteristics of the integrated control computer are described in detail. The control computer meets the needs of a certain type of aero weapons.

Abstract：In order to realize the high real-time, miniaturization and low power consumption of Beidou satellite navigation receiver, a design scheme of carrier tracking loop based on SoC FPGA is proposed. Through the analysis of FLL and PLL, and using SOPC technology, the carrier tracking loop based on SoC FPGA is realized, and the carrier can be completely stripped inside the FPGA. The test results show that the scheme can realize fast and accurate tracking of carrier signals, and has good real-time and application value.

Abstract：This paper analyzes the risk of using smart phones as U disk to complete data transmission. Combined with encryption technology and remote control technology,a data protection scheme is proposed, which integrates the functions of encryption storage, remote control destruction, location and tracking and so on. Finally，the programming realization has been carried on in order to solve the problem of data security for using smart phones as U disk.

Abstract：In this paper, based on the requirement of military integrated information systems and the characteristic of cloud computing, we design a model of platform. This model combines the cloud platform OpenStack, the large data processing framework Spark, in-memory-database and other latest tools, and then the ablity of high scalability and computing of the model is proved, by the testing of data service permornace and scalability. Expermimental results show that the model can meet the requriments of large scale data storage, processing and service in the future battlefield.

Abstract：In order to improve the efficiency and accuracy of network large data transmission and reduce the network data transmission load, a network large data fusion algorithm based on multilayer probabilistic network model and joint decision making is studied. Firstly, based on the data acquisition and caching of complex heterogeneous multi-layer networks, a multi-level probabilistic joint decision model is designed, which takes real-time sensing data and its accurate processing as the optimization objective. Then, through the main layer stratification and the signal strength of multidimensional network big data description, combined with the three step decomposition and three fusion, driven by transform denoising, the network data collaboration data fusion algorithm is proposed. Finally, the experimental and simulation results show that the proposed algorithm has obvious advantages in terms of data fusion accuracy and efficiency compared with experimental statistics.

Abstract：Taking into consideration the disadvantages of the serial structure controller-slow operation speed, poor grid current quality for the present grid-connected invert, the authors proposed a new grid-connected invert architecture, which is based on SOPC control technology on FPGA platform. It adopts the direct current PI control strategy with grid voltage feed-forward and uses hardware to realize algorithm, which improved the running speed and the overall performance of system. Meanwhile, designed regulator parameters for the control strategy and constructed grid-connected control system IP hard core for the invert in this paper. Experimental results of 1 kW prototype show that this method has good grid current quality and strong synchronous tracking capability.

Abstract：Based on the traditional average current control voltage loop, this paper proposed an optimized feedforward control scheme. The voltage control loop introduced the load feedforward control and the grid voltage feedforward control to reduce the voltage overshoot in DC side. The current control loop designed as a proportional controller using error iteration algorithm, which realize floating control and optimize the dynamic response of three-phase VIENNA rectifier. The proposed load feedforward terms updated when the load step and remains constant at steady state. In addition, this paper analyzed the frequency response characteristics and validates them by simulation.

Abstract：According to power supply requirements of power electronic converter with high-voltage and wide-range input，a double-tube flyback auxiliary switching power supply was designed. The operation principle of the double-tube flyback converter was analyzed. In order to solve the problem of current loop subharmonic oscillation from the wide-range input voltage，a slope compensation circuit was designed. A low-loss start-up circuit of current-mode double-tube flyback converter was presented. Experiments show that the design of the auxiliary switching power supply for power electronic converter is reliable and feasible.

Abstract：Designing novel harvesting system is necessary for collecting low-frequency renewable energy due to their environment dependent and intermittent characteristics. A strategy is proposed of using the multilevel feedback rectifying and BOOST/BUCK circuits, low power control chip and software control algorithm to tracking maximum power of input low-frequency energy(MPPT),meanwhile, supercapacitors are used as energy storage devices to replacing the batteries. Finally , use current transformer ,consequently, high efficiency to reach the effect of energy harvest, rectify and output. The results show that the maximum collecting efficiency is 91.16%, and that is 53.18% after integrating the input/output circuits for enhancing capability of driving loads.