摘要:A generalized frequency domain weak modulation transmission matrix H in optical fiber systems is presented in this paper.Simultaneous interaction of loss,nonlinearity and all order of dispersion in the fiber are taken into account.It is found that the odd-and even-order dispersion terms only change the amplitude-and phase-frequency response respectively,while the intensity and phase noise are only infected by the odd-order dispersion terms.Quantitative calculations of intensity,phase and optical field noise spectra are presented for intensity-detection,phase-detection and optical spectrum analyzer applications.
摘要:With the growth of Internet and storage capability in recent years,image has become a widespread information format in World Wide Web.However,it has become increasingly difficult to search for images of interest,So effective image search engine for the WWW needs to be developed.Image classification is an important module in image search engine.In this paper a novel approach for image classification based on feature element is proposed.Compared with feature vectors,feature elements can capture visual meanings of the image according to subjective perception of human beings.By using feature element,our approach for image classification is totally different with traditional image classification method.It does not calculate the distance between two vectors in the feature space,while trying to find associations between feature element and class attribute of the image.After the implementation,we have compared it with NFL ― a traditional image classification algorithm based on feature vector.Some improved results are presented.
GE Feng-xiang, WAN Qun, LIU Shen-jian, PENG Ying-ning
Vol. 30, Issue 9, Pages: 1266-1269(2002)
摘要:The problem concerned here is the frequency estimation for the sinusoidal signals with unknown lowpass envelope.Based on the eigenanalysis method applied to the parametric localization of distributed sources and quadratic programming,a super-resolution frequency estimator,called EQP method,is proposed.Here only single-experiment data are required to achieve the estimation.The EQP method,the nonlinear least-squares (NLS) estimator and the conventional MUSIC method are compared in terms of the super-resolution and computational complexity.The mean square errors (MSE) of frequency estimate of the EQP method,versus the sample numbers,signal-to-noise ratio (SNR) and distributed parameters are also presented,and these results are compared with the Cramer-Rao bound (CRB).The simulation results are shown to illustrate the good performance of the EQP method.
关键词:frequency estimation;eigenanalysis;quadratic programming;sinusoidal signals with unknown lowpass envelopes
摘要:An expandable BP on-chip learning neural network chip is designed and fabricated with a standard 0.6 μm CMOS technology.It includes 8 neurons and 64 synapses.A novel expanable topology is proposed so that no additional neuron error computation chip is needed to construct a whole neural network system.A neural network with L-layers can be composed by cascading L such chip.Analog circuits are used in this chip and capacitors are adopted to store weight values.The on-chip learning itself can be used as a refreshable tool to keep weight values right.The experiment of the parity check demonstrates the on-chip learning ability of the neural network chip.
LU Yong, ZHANG Wen-jun, ZHENG Qi-tong, LI Cheng, YANG Zhi-lian
Vol. 30, Issue 9, Pages: 1274-1277(2002)
摘要:Process Synthesis is a new top-down design method of semiconductor device and process.MOSPAD software is a kind of TCAD tool,based on the process synthes is methodology.This paper introduced the usage of the process synthesis and Response Surface Method (RSM) by MOSPAD,which will provide the process conditions for the improvement of a Fab..RSM speed up the synthesis,and will be used to analyse the sensitivity of the process window.
LIN Min, WANG Hai-yong, LI Yong-ming, CHEN Hong-yi
Vol. 30, Issue 9, Pages: 1278-1281(2002)
摘要:Design and test of a 2-GHz RF low noise amplifier(LNA)implemented with CMOS technology is presented here.Measurement results show this amplifier works well centered at 2.04GHz frequency,with a 22dB forward power gain(S21)and a noise figure(NF) less than 3.3dB.It has a fairly wide 3-dB bandwidth which is 110MHz,suitable for most portable wireless applications.
摘要:This paper proposes a new transconductor circuit which adopts CMOS double pairs.It has a transconductance linearly variable by a gate voltage,and suitable for application to highly linear continuous-time filters.Simulation results show that the total harmonic distortion (THD) of the output current is less than 0.5% for the differential input signal of 2.3V(peak-to-peak)when the supply voltages are 5V.
关键词:transconductor;second order effects;CMOS double pairs
摘要:Passive capacitor error averaging is an inherently linear capacitor mismatch calibration technique for pipelined analog-to-digital conversion,but its conversion speed is halved because of double sampling and double transferring.An improved capacitor error averaging technique is presented to make the conversion faster.The conversion speed is improved through the reduction of the number of the clock phases required for one conversion and the time allocated for one clock phase.Circuits analysis and MATLAB simulation indicate that the speed can be typically improved by more than 52% or 64% when the OTA is switched-capacitor common-mode feed back or not.The improved technique is ideal for applications of high-speed,high-resolution,and continuous operation.
摘要:A NLOS (Non line of sight) high-resolution location algorithm is proposed based on the parameters TOA(Time of Arrival)、DOA(Direction of Arrival) and DOD (Direction Of Departure) of multiple propagation paths,which can be measured simultaneously under the scenarios where single scattering is dominant in the double-array mobile communication system.From the knowledge of DOD,the single scattering path can be reconstructed and so NLOS location problem becomes solvable.Error analysis disclosed the performance of this algorithm.Simulation shows its effectiveness when no line-of-sight propagation paths exist between the mobile station and the base station.
摘要:An algorithm to improve the robustness of elastic matching in face recognition is proposed when it is strongly disturbed by the background.Automatic human face recognition has various applications.Elastic matching is an important face recognition algorithm,and it has higher robustness compared to other algorithms.But it also cannot deal with the disturbance from background effectively.The basic point of our algorithm is combining local information and global information together and using the local information undisturbed by the background to improve elastic matching algorithm.Experimental results demonstrate that the proposed method has improved the robustness of elastic matching.
MA Xiao-hong, YU Jin-long, DAI Ju-feng, YANG En-ze
Vol. 30, Issue 9, Pages: 1295-1297(2002)
摘要:Abstract:The intrinsic parameters of 10Gb/s semiconductor laser are obtained from experimental measurement and simulation.These parameters will be used in solving single mode rate equation of the laser.Thus the output pulse waveform and chirp of the gain-switched laser are analyzed and experimented by transmitting in various fibers.It is verified that the simulation result is in agreement with the experimental result.
关键词:single mode rate equation of laser;dynamic character;chirp
CHEN Jian, ZHANG Wei, LIU Xiao-ming, PENG Jiang-de, ZHOU Bing-kun
Vol. 30, Issue 9, Pages: 1302-1304(2002)
摘要:Based on the experimental and theoretical investigations on the gain and noise characteristics,a broadband hybrid amplifier composed of distributed Raman amplifier and EDFA is realized.A method for optimal design of the hybrid amplifier is proposed.It is pointed especially that adjusting the gain specta of Raman amplifier and EDFA properly can realize flat noise figure spectrum.
摘要:This paper extends PEEC method for fast-accurate electromagnetic analysis of RF-MEMS spiral inductors.By introducing electrical boundary conditions into PEEC,the capacitance matrix is extended to include 3D-substrate coupling.With equivalent of charges in interface,Green functions in uniform dielectric regions can be used in calculation.By using proper mesh size and considering retardation times,the precision of PEEC can be maintained even in tough cases.These improvements reduce the analysis complexity and calculation time.The numeric results show good agreement with published measurement data.
关键词:radio frequency microelectromechanical systems (RF-MEMS);partial equivalent element circuit (PEEC);inductance matrix;capacitance matrix
摘要:Linear prediction (LP) HMM does not make the independent and identical distribution (IID) assumption as traditional HMM;however it often produces unsatisfactory results in practice.In this paper,a new combined model for speech recognition is proposed,based on a new analysis of both HMMs' modeling strengths and weaknesses.The new model works with LPHMM as the dynamic part and traditional IID-based HMM as the static part;in addition,easy implementation and low cost are preserved.Experiments on speaker-independent continuous speech recognition demonstrated that the combined model performed much better than both LPHMM and traditional HMM.Theoretically,a new closed-form parameter re-estimation formula is suggested for training LPHMM.
摘要:This paper proposed a new feature extraction method for Chinese character recognition by using Gabor filters.Based on the theory of Gabor filters and the statistical information of Chinese character images,an effective method to design Gabor filters was developed.Moreover,to improve the performances for low quality images,non-linear functions were designed to regulate the outputs of Gabor filters adaptively.This paper also improved the feature extraction method to enhance the discriminability of histogram features.Experiments show that our method performs excellently for images with noises,backgrounds or stroke distortions and can be applied to printed or handwritten character recognition tasks in low quality greyscale or binary images.
摘要:The measurement of wave speed is always one of the key problems in the research of subsurface penetraing radar,particularly when the location and image of the targets in the ground are measured.On the basis of the signal features of subsurface penetrating radar,extraction of edge contours which uses zero-crossing points and others traditional methods are discarded.The method of extremum is given for extraction of edge contours.The edge contours of the targets can be extracted more accurately using the method of extremum and the new method generates less data for Hough transform.And according to the analysis of Hough transform and the deficiency of traditional Hough transform,an improved Hough transform is proposed in which a FIR low-pass filter is added to improve the accuracy of Hough transform.The improved Hough transform has better performance than the traditional Hough transform by the evaluation for the practical data.
摘要:The effects of dual-polarized square patch antenna physical parameters,such as the width of feedline,the relative dielectric constant and the depth of the substrate,on the antenna resonant frequency,reflection coefficient and isolation are analyzed by the FDTD method.The different effects of these parameters on the two-element array performance are compared with those of the single element.The variances of the array coupling with its parameters are also studied.
摘要:This paper presents a Constant False Alarm Rate (CFAR) detection approach of ship wakes for Synthetic Aperture Radar (SAR) images based on the technique of "adjacent pixels summing (APS)".The key of this method is using the APS strategy to perform Radon transform over the subimage approximately.The CFAR detection algorithm,which uses statistical hypothesis testing on probability model in transform space,is then developed.The effectiveness of this algorithm is demonstrated by the test results of simulating wake images and actual SAR images of ship wakes.
摘要:Incremental length diffraction coefficients (ILDC) is often used to calculate the diffraction by edges.However,ILDC displays infinities for certain combinations of observation and incidence directions.To eliminate such infinities,Michaeli derived a set of expressions for the fringe current components of equivalent edge currents,those expressions are finite for all aspects of illumination and observation,except for the Ufimtsev singularity.However,for certain non-singular aspects of ILDC,maybe there is a wide divergence between these two methods' results.In this paper,new exppressions for fringe current components of equivalent edge currents are presented.Compared with those given by Michaeli,new expressions can not only eliminate the infinities in LDC effectively,but also keep good agreement with ILDC.So it is very valuable in practical applications.
关键词:electromagnetic scattering;method of equivalent currents;incremental length diffraction coefficients
摘要:In this paper a new blind estimation algorithm is proposed which uses Singular value decomposition of signal self-correlation matrix to estimate Signal-Noise-Ratio (SNR) of non-constant envelope signal over AWGN channel.The theoretic analysis is also given.Computer simulation results show that the performance of this algorithm is robust and the estimation error is less than 0.5dB when the actual SNR ranges from 1 to 20dB for MPSK and MQAM signals.
关键词:signal-noise-ratio estimation;singular value decomposition (SVD);non-constant envelope signal
摘要:This paper presents a sandwich structure of photonic bandgap (PBG),in which a periodic pattern of square metal pads is printed in the midmost of the substrate.It is compatible with the conventional microstrip circuits.Neither drilling nor suspending the substrate is required,and the integrality of the ground plane is also kept.Several circuits with different pattern of periodic lattice are compared,and potential applying way is demonstrated.
关键词:photonic bandgap;finite difference time domain method;microstrip;filter
摘要:In this paper,we discuss a new method for rule extraction based on hyper-box representation.The method is deducted from the MRCC model,used for pattern classification.The method partitions the pattern space with multi-dimensional fuzzy hyper-boxes,and assigns a class label with certainty degree for each variable fuzzy region.These rules are extracted from numerical data through a recursive learning procedure.Experiment shows the efficiency of rule extraction procedure.
摘要:In this paper,the use of GSM Short Message method incorporating GPS system vehicle guidance is discussed and a mathematical model is founded to represent the acceleration and main error which affects the GPS/DR integrated system position accuracy.A data fusion algorithm of integrated system is introduced to improve the accuracy of the navigation system.Since the dead reckoning(DR) technique is being used in the system,this algorithm is not only improving the GPS position accuracy but also preventing the error of the DR to be diversified.The method has made the integrated system to be robust and reliable in an environment with a lot of tall buildings,tunnels,bridges and mountains like Hong Kong.
摘要:Both the electrically small multiple dipoles model with embedded cascade inductor network and the network synthetic method,are offered proposals for expanding the impedance bandwidth of electrically small antenna.Because of the results of sort deduced from the impedance of computing model of small antenna,to guarantee the precision of synthesizing components of cascaded inductor network,the random optimizing method based on the measuring data of the impedance on multiple port has been employed for synthesizing the parameters of embedded match network.Good agreements are obtained between synthesizing and experimental results,and the proposed method is verified
关键词:multiple small dipole;optimizing;synthesis;matched network
摘要:The perfectly matched layer (PML) is a highly accurate approach to the finite difference time-domain (FDTD) mesh truncation.However,the PML does enhance increase the computational complexity and the memory requirement for the FDTD simulation.In this paper,we propose a new PML construction based on the properties of the split fields in the PML region.The construction only contains eight PML subsections without corners.Also,we transform a three-dimensional split field component to a two-dimensional array in order to save computer memory.Verifications of the PML construction are provided with numerical examples.
SUN Mei-xiang, CHEN Quan-lin, LU Guo-wei, LI Tian-li
Vol. 30, Issue 9, Pages: 1400-1401(2002)
摘要:Virtual Instrument completely changes the style of traditional instrument construction which is defined by producers and,with the rapid development of computer and network technology,Virtual Instrument successfully runs to the remote Virtual Instrument.This article focuses on the implementation method of remote Virtual Instrument,and discusses how to use the available network resources and technologies for the implementation of such a remote virtual digital oscilloscope in detail.
关键词:Virtual Instrument;DSP(digital signal process);C/S model.
Wu Xian-liang, TANG Jing, ZHAO Jin, CHEN Dong-fang, LI Shi-xiong
Vol. 30, Issue 9, Pages: 1402-1404(2002)
摘要:Solution on electromagnetic wave propagation in a two-dimensional reflector antenna by Symplectic Geometrical Theory has been gotten.At first,the new vectors having the same numbers with these original physical vectors are introduced.The new vectors combine with those original physical vectors to form a symplectic space.The propagating problem in the physical space is promoted to Langrange manifold in the symplectic space,with better result obtained.
SONG Guang-nong, ZHAO Yong-xiang, XING Xin-yu, CHEN Chang-jia
Vol. 30, Issue 9, Pages: 1409-1412(2002)
摘要:Class Guided Random Early Discarding (CGRED),a novel scheme to provide services differentiation in today's Internet,is proposed.It can precisely maintain the assigned priority and pre-allocated bandwidth for each service class.As a single queue management mechanism,CGRED possesses all advantages of the simplicity,robustness and scalability of schemes like RIO and WRED.And it has the capability to maintain the pre-allocated bandwidth for each class of service precisely just as CBQ.The key enforcement of CGRED to standard RED is to discard packets of different service classes adaptively with different probabilities according to the bandwidth actually occupied by different classes and the pre-allocated bandwidth.To accomplish this,a couple of class-specific guidelines,the 'adjustment direction'and 'guiding probability',are introduced to conduct the adaptive determination of class-differential discarding probabilities for each class of service.The guidelines for each class of service are updated periodically based on online measured bandwidth utilization.When discarding packets,CGRED sets the class-specific guidelines as the boundary of the packet dropping probability of standard RED.In this way,the unbalance of bandwidth consumption among different service classes is successfully transferred to differential operations in the buffer management among different service classes.Extensive simulation work has been conducted to verify the performance of CGRED and robustness of parameters in its algorithm.The simulation results show that CGRED is an effective scheme to provide Differential Services in networks with pre-allocated bandwidth and pre-assigned priority hierarchy among service classes.