The mainstream terahertz imaging technology has mutual constraints in imaging speed, resolution, sharpness and manufacturing cost, etc. A fast terahertz wave imaging technology based on a light-controlled spatial modulator is proposed. This technology realizes the single-pixel terahertz detector by integrating the new silicon-based terahertz modulation material with the commercial Digital Light Processing(DLP) device. The influence of terahertz beam distribution characteristics and Gaussian background on imaging results is further analyzed. An effective optimization method is proposed. The rapid imaging results confirm that this new technology can significantly improve the imaging resolution and clarity.

In order to realize THz gas spectroscopy system, researches upon off-chip measurement of 245 GHz subharmonic receiver chip are conducted. An off-chip measurement system for 245 GHz subharmonic receiver chip is established and an off-chip demonstrator measurement system for gas spectroscopy sensor based on 245 GHz receiver and transmitter chips is fabricated to measure the conversion gain and bandwidth of the 245 GHz subharmonic receiver chip. 15 dB conversion gain and 15 GHz bandwidth are obtained in off-chip measurement system, and 9 dB conversion gain and 16 GHz bandwidth are obtained in off-chip measurement demonstrator system. Experimental results indicate that the results of off-chip measurement system and off-chip measurement demonstrator system of 245 GHz subharmonic receiver chip are basically consistent. A THz gas spectroscopy sensor is constructed when a gas chamber filled with some specific kind of gas is added into the off-chip demonstrator measurement system. Compared with available gas spectroscopy of the same type, the subharmonic receiver chip has advantages of high gain, wide bandwidth, being integrated with local oscillation signals, and low power consumption, and is very suitable for applications of low volume intelligent gas spectroscopy in consumer electronics.

In the THz frequency band, the low quality factor of passive device capacitance inductance, the influence of parasitic circuit parameters and cut-off frequency of Metal-Oxide- Semiconductor Field-Effect Transistor(MOSFET) make it difficult for THz oscillator circuit to achieve high power output. A 300 GHz tuning oscillator is proposed to achieve high power output. Firstly, the structure of cross coupled Push-Push oscillator is adopted to realize the second harmonic signal output of 300 GHz. Through the method of power synthesis, the proposed oscillator exports the second harmonic 300 GHz signal, which increases the output power of the oscillator and breaks through the cut-off frequency of MOSFET. The gate interconnection inductance is also adopted to increase the output power. Secondly, instead of tuning with a variable capacitor on the chip, the Push-Push oscillator modulates the substrate voltage of MOSFET to adjust the parasitic gate capacitors of MOSFET to achieve frequency tuning, further increasing the output power. The proposed push-push oscillator is fabricated in TSMC 40 nm CMOS process. By extracting the second harmonic signal and using the substrate voltage tuning technology, the proposed coupled Push-Push oscillator has a peak output power of -2.2 dBm at 309.0 GHz and achieves 303.5-315 GHz tuning range. The phase noise of the oscillator is -79.5 dBc/Hz at 1 MHz offset and the power consumption is 28.6 mW.

Terahertz wave has the characteristics of high transmission rate, large capacity, good directivity and strong anti-interference ability, but it is more suitable for short-distance wireless communication due to the large atmospheric loss and the lack of high-power sources. Based on the excellent characteristics of the 300 GHz band, a 310 GHz terahertz wireless communication system is designed. The communication link is based on Virginia Diodes Inc(VDI) harmonic mixing and superheterodyne transmission. The system is mainly composed of frequency multiplier, power amplifier, filter and VDI subharmonic mixer. The actual communication system is built according to the link budget. The simulation and experimental results show that the distortion of the demodulated signal increases with the increase of the communication rate. The system can achieve 10 Gbps error-free transmission with a communication distance of 1 m. When the rate reaches 11 Gbps, the bit error rate is 5×10-6.

Aiming for the low accuracy and difficulty of predicting solar cell life by using soft failure mode, a new method is proposed to obtain solar panel spectrum by using terahertz spectrometer. Based on the cuckoo algorithm, the study predicts the cell's remaining life by applying Particle Swarm Optimization-Support Vector Regression(PSO-SVR) algorithm and finally employs the ultraviolet acceleration test to verify the prediction results. It turns out that the method is applicable to predict the remaining life of solar cells with different levels of loss. Compared with other algorithms, the technique works better on the life prediction of traditional silicon solar panels and GaAs solar cells, and the accuracies are up to 98.92% and 92.86% respectively.

Due to the multi-peak of Binary Offset Carrier(BOC) autocorrelation function, an unambiguous acquisition algorithm of BOC signal based on time-domain filter is designed. The frequency response of the time-domain filter is obtained, then the filter coefficients are deduced to construct a Finite Impulse Response(FIR) filter. The simulation of different algorithms shows that the time domain filter can completely eliminate the side peaks of BOC autocorrelation. The filter order is reduced and an achievable unambiguous acquisition algorithm of BOC signal based on time- domain filter and Multi-bank Matched Filter(MMF)-Fourier Transform(FFT) is designed. The simulation analysis shows that the algorithm can eliminate side peaks and capture the signal correctly.

Based on the combination of outfield test and mathematical simulation experiment, a method for evaluating blanket jamming distance of netted radar based on operational scenario simulation is proposed. Firstly, the radar network data fusion simulation model is proposed based on the spatial situation and target characteristics. Secondly, the network radar blanket jamming distance test evaluation method is established. Thirdly, the radar electronic warfare simulation evaluation reckoning view system named EwserView is given. Finally, combined with this software system, the test evaluation example and interface of substitute equivalent reckoning method are offered. This method can solve the test evaluation problem of netted radar based on test situation, and the feasibility and effectiveness of this method are verified in experiments.

Wind Profiler Radar(WPR) is widely applied to provide continuous atmosphere detection of wind field with high time-space resolution for airport, meteorological administration and army. Usually five beams are used for detection, and interference signals could exist because of different factors. This paper firstly introduces the fundamental principle of WPR. Then the resources of interference and countermeasures are discussed. Next, anti-interference algorithms in signal-design, data processing, and beam-control are discussed respectively. At last, an Adaptive Dynamic Beam Selection(ADBS) algorithm is proposed in detail. The simulation results show the effectiveness of this algorithm.

A Radio Frequency(RF) energy harvesting based Cognitive Radio Network(CRN) is proposed, where a Secondary User(SU) first harvests energy from the RF signals of Primary User(PU) and then transmits data using the harvested energy in one slot. The total consumed energy by the SU must be no more than the total harvested energy, in order to protect the PU from interference. Under the satisfaction of Quality-of-Service(QoS) of SU, the goal is to determine the optimal transmitting time and power allocation that maximizes its Energy Efficiency(EE) in the RF Energy Harvesting CRN(EH-CRN). In the process of maximizing energy efficiency, a balance is found between QoS and energy consumption. Less energy consumption allows the cognitive energy harvesting system to run more steadily and continuously, which is important while throughput constraint ensures the QoS of the system. To solve EE optimization as a nonlinear fractional optimization problem, it is firstly decomposed into two sub-problems by using Coordinate Ascendant, and then the nonconvex problem is transformed into an equivalent concave problem by using Charnes-Cooper Transformation method. Simulation results show the proposed scheme achieves effective EE.

Traditional Received Signal Strength Indication(RSSI)-based moving target localization and tracking generally employs tri-lateration/angulation techniques. Although this method is simple and easy to be implemented, it creates signi.cant errors in localization estimations due to nonlinear relationship between RSSI and distance. The Generalized Regression Neural Network (GRNN), a one-pass learning algorithm, is well known for its ability to train quickly on sparse data sets. Therefore, GRNN-based mobile Target Tracking(GMTT) is proposed in this paper. GMTT deals with high nonlinearity in RSSI’s target location relationship by using GRNN, then further refines these location estimates with the help of KF framework. Simulation results show that GMTT can effectively decrease the Root Mean Square Error(RMSE) of target localization compared with RSSI+KF.

A variable-shaped log-periodic dipole antenna with adjustable pitch angle is proposed. By changing the rotation angle of the antenna body and the oscillator, the maximum radiation direction of the log-periodic dipole antenna is continuously adjustable within a certain range. And the addition of the X-type reflector effectively improves the front-to-back ratio and increases the gain of antenna. This log-periodic dipole antenna, which can adjust the rotation angle of the oscillator, can realize a wide impedance bandwidth ranging from 1 GHz to 2 GHz with Voltage Standing Wave Ratio(VSWR)<2 and the pitch angle continuously adjustable from 25° to 90°.

In the field of microwave power transmission, when the polarization mismatch between receiving antenna and transmitting antenna occurs, the energy received by rectifier antenna will decrease sharply. A dual polarized slot receiving antenna with harmonic suppression function at 2.45 GHz is designed. The receiving antenna can receive energy even when it rotates at a certain angle. The horizontal and vertical polarization of the antenna can be realized by opening a suitable size and symmetrical structure slot on the patch, and the harmonic suppression function can be realized by using defected ground structure. The simulation results show that at 2.45 GHz, the isolation degree of the antenna is higher than 17.6 dB, the gain is 3.9 dBi and the impedance matching is good. The echo loss of the antenna at the second and third harmonics is -0.38 dB and -0.96 dB. The simulation results are in good agreement with the measured ones.

A microstrip antenna is proposed, which is with a thin-substrate and light-weight but considerable bandwidth and good directional radiation properties. This antenna works at 5.8 GHz and is printed on a thin substrate with a height of 0.5 mm, corresponding to 0.01λ0, where λ0 is the working wavelength of the antenna. To overcome the narrow bandwidth problem owing to the thin substrate, three parasitic patches of different sizes and with metalized conductive vias are placed close to the rectangular radiating patch. A prototype with 5×5 radiating elements is designed, fabricated and tested. Measurement results show that the proposed antenna possesses a |S11|<-10 dB impedance bandwidth up to 2.2%(from 5.866 GHz to 5.74GHz) and a gain of 8.2 dBi. The prototype is with a size of 164.5 mm×179 mm and a weight of 130.1 g, including the weight of 25 SMA connectors, 70.8 g, corresponding to 4.4 kg/m2.

A compact wide-bandwidth and dual-polarized antenna is presented in order to detect different radars effectively. The antenna can achieve VSWR≤2.5 and good radiation pattern which can meet the covering airspace requirements of the phased array. By analyzing the antenna array theory, the array polarization switching and beamforming are realized simultaneously. The measured results of the array reveal that it achieves multi-polarization at bandwidth of f0 to3f0, scanning angle of ±45° in the azimuth and ±20° in the elevation. The design has laid a technical foundation for the naval vessel RF integration system.

A C-band wideband and high speed frequency hopping source with low noise is presented. In order to improve the phase noise performance, the source contains single Phase Locked Loop(PLL) and down conversion circuit. The high frequency output moves to low frequency, which makes the working frequency of N divider reduced greatly, meanwhile the maximum output frequency of the frequency synthesizer is improved, and the frequency outputs at the same interval. This solves the contradiction of high frequency and low frequency interval, and realizes the output with low phase noise. The tests show that the phase noise is -123 dBc/Hz@10 kHz under output frequency of 4 460 MHz. The design uses controlled low dropout linear regulator as power supply for HMC704LP4 to avoid error mode selection. It effectively solve the unlock problem when powering up HMC704LP4 caused by error mode selection.

Theoretically, magnetic resonance wireless power transmission can achieve high efficiency. However, due to the influence of various media, a large deviation exists between the measured efficiency and the theoretical efficiency. A magnetic resonant wireless power transmission system in non-ferromagnetic medium is proposed based on the equivalent circuit model. The influences of different media on the coil impedance and resonance frequency are analyzed by calculation and simulation, and the results of calculation and simulation are in good agreement. The design steps of the high efficiency system in the medium are proposed, and the simulation results show that the efficiency can be improved by more than 30% after optimization.

A three-dimensional microstrip electromagnetic material with shielding effect is designed, which can be used for debugging RF communication system. Compared with the traditional metal isolation strip, the presented material has the advantages of being light and thin, small in size and frequency adjustable. Those properties are advantageous for miniaturization and light weight design of the product.The presented material is composed of a horizontal，Electromagnetic Band Gap(EBG) structure and a vertical period “7” type structure. The EBG structure has the function of suppressing crosstalk signals in specific frequency bands. The “7” type structure can introduce the coupling current into the GND to further enhance crosstalk suppression and purify the electromagnetic environment inside the cavity. By adjusting the value of the inductance embedded in the material, the suppression of high frequency crosstalk signals at different frequencies can be achieved. Experimental results show that the presented material can achieve crosstalk signal suppression greater than 35 dBc in 15-19.25 GHz, 19.4-21 GHz, and 21.2-24 GHz bands, respectively.

Nowadays, the issue of electromagnetic reliability due to the non-linear effect of the components in electronic products becomes more serious. In this paper, a simple and effective test and evaluation method based on narrow-band filter and microstrip line is designed to accurately demonstrate the characteristic of different Electrically Conductive Adhesives(ECA) under Sinusoidal Excitation. A filter is adopted to eliminate the intrinsic harmonics in the test system. The high-order harmonics components are generated due to the non-linear effect. The power of the third-order harmonics is used as a measurement index for non-linear effect of electrically conductive adhesives in this method. The comparison test can be conducted by measuring the power of the third-order harmonics of different electrically conductive adhesives. This work provides strong technology supports for the evaluation and design of Electrically Conductive Adhesives.

A novel magnetron memristor model based on the cubic smooth memristor is proposed, a new chaotic oscillation circuit of five-order is designed by using two magnetically smooth memristors, common capacitors and inductors. It is found that the dual-memristor system has the equilibrium point set. The phase diagram, Lyapunov exponent and bifurcation diagram are analyzed by using the conventional dynamics method. The dual-memristor chaotic circuit has complex dynamic behaviors. The motion trajectory is not only dependent on the circuit parameters but also affected by the initial state of the circuit. Strange attractors are explored in terms of energy, resulting in the coexistence of different periodic motions and different chaotic attractors with different topologies. The correctness of Matlab simulation and the realization of the circuit design are verified by PSpice.

At present, many image fusion algorithms mainly use image energy information to fuse image coefficients, ignoring the texture information of the image, which brings the defects of Gibbs and block phenomenon to the fusion results. This paper designs a multi-focus image fusion algorithm based on the second generation Curvelet transform coupled with texture information adjustment. Firstly, the second generation Curvelet transform is utilized to obtain different sub-band images from the input image. Then, the texture information factor is constructed by using R,G and B values of the image, and the texture information of the image is measured. By combining the information entropy characteristics of the image and the R, G and B values of the image, the fusion results have more texture information. The average gradient feature of the image is adopted to compute high frequency coefficient fusion, which makes it more capable of describing details such as edges. Finally, the image fusion test of this algorithm shows that compared with current fusion algorithms, the fusion image of this proposed algorithm is clearer, without the defects of Gibbs and block phenomenon, and with larger values of mutual information and standard deviation.

In order to achieve synchronous encryption of multiple images and improve the security of ciphertext and the ability to resist geometric transformation attacks, the multi-image optical encryption algorithm based on quick response and cascaded fractional transform is proposed in this paper. By introducing complex number theory, the first plaintext is processed with template image to form complex image. The complex image is decomposed based on the phase–amplitude truncation mechanism to obtain the phase distribution. With the help of Logistic map, the random masks are obtained, and the modulation masks are constructed by combining Fresnel band and Hilbert models. Combining the phase truncation method with fractional order Fourier transform, a cascaded Fast Fourier Transform(FFT) encryption mechanism is designed, and the remaining plaintext is processed by frequency domain transformation based on modulating mask and phase distribution to output the amplitude image, then it is converted into corresponding Quick Response(QR) code. The Gyrator transform is introduced, and the QR code is interfered by laser beam through the corresponding photoelectric device to get the cipher. The test data show that compared with the existing multi-image optical encryption technology, the proposed algorithm has higher security and anti-attack ability, and the distortion of the restored image is smaller under the clipping transformation.

In order to improve the fusion quality of remote sensing image, a remote sensing image fusion algorithm based on non-downsampling shear wave transform coupled information quantity restriction rule is proposed. Multi-Spectral(MS) images are decomposed by using Intensity-Hue-Saturation(IHS) transformation to extract the brightness (I) component of MS images. Then, based on non-downsampling shear wave transform, I component and PANchromatic(PAN) image are decomposed to obtain high and low frequency coefficients. Finally, using the mean value of low-frequency coefficients and regional energy information, the information quantity restriction rule is established, and the influence degree of different low-frequency coefficients on fusion coefficients is analyzed. Different fusion methods are designed to fuse these low-frequency coefficients. The improved average gradient measurement model is adopted to finish the fusion of high frequency coefficients. The fused high and low frequency coefficients are computed by the inverse transformation of Non-Subsampled Shearlet Transform(NSST), and new brightness components are output. The fused remote sensing images are formed by combining them with the original Hue and Saturation components through the inverse transformation of IHS. The experimental results show that compared with the existing fusion scheme, the fusion image of the proposed algorithm contains more information and less spectral distortion.

In view of the fact that most image restoration algorithms rely on fixed size sample blocks to search for the best matching blocks, ignoring the structure information of the sample blocks, resulting in the discontinuity and ringing phenomenon of the repaired images. Based on the approximation between the sample block and its neighborhood block, this paper designs an image restoration algorithm by using the structure information constraint rule and matching model. The information entropy feature of the image is introduced into the priority calculation of the blocks to be repaired, and the priority repaired blocks are obtained. Structural information measurement model is constructed by the number of known pixels in the sample block to measure the structural information of the sample block. Constraint rules of structural information are established according to the measured values to adaptively adjust the size of the sample block. The matching model is constructed by using the color and gray features of the image, in order to find the best matching block in the known region after adjusting the size of the sample block, and repair the block needing repairing. The experimental results show that the restored image obtained by the proposed algorithm has high structural similarity, good texture continuity, and no information discontinuity.

At present, many image tampering detection methods mainly measure the distance between image features to complete feature matching, ignoring the color information of the image, resulting in more false detection and missed detection in the detection results. In this paper, color information is introduced into the process of image feature matching, and a tamper detection algorithm based on color constraint model is designed. The Laplacian operator and Harris operator are utilized to extract the image features. The R, G and B primary color information of the pixels, as well as the image feature descriptor are adopted to form a color restriction model for measuring the color information between feature points. Then the image feature matching is accomplished by using the distance measurement between the measure value and the feature points to fully eliminate the mismatch phenomenon and effectively improve the matching accuracy. Additionally, the distance penalty model is constructed according to the distance variance between feature points to cluster the matched image features for identifying the tampered content accurately. The experimental results show that compared with other tamper detection algorithms, the proposed algorithm not only has higher detection accuracy for forgery content, but also has better adaptability for content operations such as blur and rotation.

During each iteration of embedded Field Programmable Gate Array(eFPGA) customization process, router has to be run on a new Route Resource Graph(RRG) for the new architecture in order to meet design constraints in many aspects, such as area and timing. Conventional eFPGA RRG modeling method regenerates the whole chip’s architecture description and builds routing nodes and edges based on it in every evaluation iteration. However, the efficiency of this method suffers from the rising scale of the chip being evaluated. A module-based RRG modeling method is proposed to address this problem. It firstly builds RRG for every module type and models interconnect relations among them, after that, stitches them together according to the device resource arrangement. It depends on a relatively small database, reduces the modeling runtime and memory peak footprint by around 84% and 64% respectively and thereby improves the eFPGA evaluation efficiency.

Ensemble Clustering(EC) is one of the key means to solve data mining problems, but the existing EC methods rarely consider the various noises that may damage the clustering structure and reduce the clustering performance. To solve this problem, an Improved Spectral Ensemble Clustering(ISEC) method is proposed. Firstly, the clustering problem is modeled as a graph partitioning problem of coincidence matrices derived from inputting multiple Basic Partitions(BPs). Then, The ISEC method learns to obtain the low rank representation of the covariance matrix, and carries on the spectral clustering to improve the clustering performance. Finally, the optimization solution is carried out by the enhanced Lagrange multiplier method, so as to obtain the final clustering result. The simulation results on several real data sets show that the clustering performance of ISEC method is better than that of most existing clustering methods.

In order to improve the accuracy of sensor measuring soil moisture, Improved Particle Swarm Optimization(IPSO) algorithm is proposed. Firstly, Gauss transform is utilized to improve the local search ability, and Cauchy transform is adopted to attract other particles to better search space area, which improves the global search ability. Secondly, Chaotic function is adjusted the inertia weight dynamically, it has larger value in the initial iteration stage and smaller value in the later iteration stage, and the searching speed is slowed down in the later iteration. The simulation results show that the Mean Square Error(MSE) and Pearson correlation coefficients of IPSO algorithm are better than that of other algorithms for measuring gravel dehumidification and moisture absorption data, the MSE of the measured data of dehumidification for IPSO at the substrate potential of 1 000 cm is 16.62×10-6, which is 75.59%, 66.67%, 63.53%, 53.73% and 57.53% lower than that for LSM, FOA, HSA, PSO and SAA respectively. For the measured data of moisture absorption at the substrate potential of 1 000 cm, MSE of IPSO is 10.21×10-6, which is 81.42%, 75.29%, 72.00%, 65.57% and 67.69% lower than that of LSM, FOA, HSA, PSO and SAA respectively.

The Spatio-Temporal Context(STC) tracking algorithm has defects in feature representation and scale adaptive strategy. Undesired conditions, i.e. abrupt deformations, partial occlusions or scale variations of the object appearance, would severely degrade the performance of the tracker. In this paper, based on the improvement of the STC algorithm, an algorithm is proposed for merging template response according to the STC model and color histogram response to locate the target object. The color statistics-based model has a good complementary nature to the STC model. The STC tracker combining the color histogram response can be inherently robust to both motion blur and deformations. Moreover, another scale search strategy which is based on a multi-scale pyramid model is adopted to replace the scale module in STC tracker, and makes scale estimation more accurately and adaptively. Extensive experimental results on large-scale benchmark sequences show that the proposed algorithm exhibits better tracking performance and adaptability under the complex environment of different influencing factors while running at 134.2 frames/s on average.

Automatic crowd counting has attracted widespread concern in the field of video surveillance. In recent years, the Convolutional Neural Network(CNN) has achieved miraculous results in crowd counting. However, current research based on deep learning mainly concentrates on high-performance PC to count the people with a single still picture. The network model has huge computational resources consuming due to its large amount of parameters and complex network structure, which is difficult to deploy in actual surveillance video crowd counting system. Therefore, the deep learning method is adopted to realize the real-time crowd counting on the embedded platform by pruning and compressing the network model and using TensorRT to accelerate the model inference. The proposed crowd counting algorithm achieves a balance between accuracy and speed with Mean Absoulte Error(MAE) of 21.6 and average Frames Per Second(FPS) of 22. Its performance on the embedded platform can approach the real-time result.

It is necessary to carry out the reasoning to realize effective judgement of target maneuver recognition based on information of target position and motion state in order to meet the needs of target behavior prediction in the multi-purpose application such as battlefield identification and civilian target surveillance. Aiming at the problem that the original geographical location cannot provide semantic information, this paper adopts fuzzy membership theory to construct road grid model to extract semantic feature of target position, and the possible derivation error of location semantic is avoided by using K Nearest Neighbor(KNN) method. On the basis of positional semantic modeling, the target maneuver action is inferred with Hidden Markov Model(HMM). Combining with the application of airport surface surveillance, the simulation results show that the accuracy of the proposed method based on model of location sematic feature and KNN is improved compared with the general HMM.

In the analysis of nuclear energy spectrum, the single energy peak is affected by low energy tail and high energy tail. Therefore, a method is proposed in which the data fitting window is determined by the full width at half maxima values of single energy peak, and then the Gaussian function fitting method is applied by using fast iterative algorithm. In this paper, the Gaussian signal fitting experiment with noise is carried out, then, the single energy peaks with different contents in the actual measured X-ray fluorescence spectrum are fitted by fast Gaussian method. The results show that this method has a good fitting effect on the single energy peak of high content nuclides, and the single energy peak of high background and low content nuclides should be smoothed before fitting.

Micro-Electro-Mechanical System(MEMS) packaging stress is a residual stress produced by MEMS packaging process, which has significant influence on the thermal stability and long-term stability of MEMS devices. An accurate assessment of stress is useful for the studying of packaging stress. Because the package stress is too low to be directly measured by the existing techniques, a novel measurement method based on stress magnifying structure and micro-Raman spectroscopy is proposed to obtain the average package residual stress in MEMS devices. Based on the theoretical analysis, the magnification ratio between the original package model and the stress magnifying structure is obtained, and a design rule of the stress magnifying structure is given. Then, a Finite Element Modeling(FEM) of a kind of high precise MEMS micro-accelerometer is conducted to prove the theoretical analysis result. Finally, for the package residual stress measurement of the micro-accelerometer, fabrication and package are successfully performed. Then stress measurement experiment is conducted to measure the magnified package stress and to calculate the actual average package residual stress in the micro-accelerometer, showing an identical result with previous simulation, which proves the measurement method much reliable.

The micro-scale hemispherical resonator gyroscopes based on Micro-Electro-Mechanical Systems(MEMS) manufacturing technology are characterized by small size, light weight, low cost and suitable for volume production while absorbing the advantages of full-angle measurement of traditional hemispherical resonator gyroscopes. In the process of fabricating hemispherical shell structure by sacrificial layer process, whether the molds of hemispherical resonator on single crystal silicon have smooth surface and structural symmetry has decisive influence on the performance of hemispherical resonator. In this paper, the principle of isotropic etching of HNA(Hydrofluoric, Nitric, Acetic) for making micro-scale hemispherical resonator molds is described. Then the research progress of micro-scale hemispheric resonator gyro based on isotropic etching of HNA solution is introduced. Finally, the main technical problems of isotropic etching process are pointed out, and some possible solutions are put forward.

Based on the threshold-arithmetic algebra system and guided by HE map, the threshold controllable switch of Schmitt inverter is analyzed, and the ternary current-mode CMOS Schmitt inverter is designed. The simulation results of Hspice show that the designed circuit has correct logic function and good transient characteristics. This circuit design not only makes the design of the threshold-arithmetic algebraic system further improved, but also makes the design of the ternary Schmitt inverter more simple and intuitive.