XIE Zi-chao, LU Jun-lin, TONG Dong, WANG Xiao-yin, CHENG Xu
Vol. 39, Issue 11, Pages: 2473-2479(2011)
摘要:Way selective technique could reduce the instruction cache energy consumption significantly.However,existing solutions usually bring extra fetch latency due to mispredictions or complicated updating mechanism,reducing the energy-efficiency.The paper presents an energy-efficient Combining Way Selective Cache for the instruction cache in superscalar processors (CWS-Cache).It combines the advantages of way prediction and way history techniques,and selects the best way selective mechanism for different situations.It not only reduces the instruction fetch energy effectively,but also improves performance by reducing the latency of misalignment fetch groups.Experimental results demonstrate that,on average,CWS-Cache reduces fetch energy consumption of the 8-way set-associative instruction cache in the baseline processor by 84.98%,and improves performance by 3.5%.Compared with three existing techniques,CWS-Cache improves the energy-delay product (EDP) by 15.48%,14.13%,and 8.76%,respectively.
关键词:superscalar microprocessors;way prediction;way history
摘要:In order to fast surround the moving target by hunting robots,a multi-robot hunting algorithm based on dynamic prediction for the trajectory of the moving target and hunting points is proposed.With moving of the target,sample points are updated dynamically to predict possible position of the target in a short period of time by using the fitting of a polynomial,and the safe domain of the moving target,which is a denied area for the hunting robots,is built to avoid the target's escape when it detects robots.The method of negotiation is adopted to allocate appropriate desired hunting point for each robot.Then robots approach desired hunting points rapidly by means of multi-scout ant cooperation algorithm so as to achieve the hunting of the moving target.The simulations show that hunting robots can surround the moving target of which the trajectory is unknown rapidly and accurately by the algorithm in the environment with complex obstacles and the results obtained is satisfactory.
关键词:multi-robot hunting;dynamic prediction;multi-scout ant cooperation;safe domain;desired hunting point
LIU Wen-song, ZHU En, WANG Jian, XU Long-tao, LIN Ye
Vol. 39, Issue 11, Pages: 2486-2491(2011)
摘要:The algorithm and the hardware implementation of Arithmetic Coder (AC) in JPEG2000 are studied.A new kind of the sequence structure of the Renormalization Procedure (RP) is proposed.With the independent total-shift prediction procedure added,the current context can be processed serially by AC without loop computation.Based on the proposed algorithm,the 3-stage pipeline architecture with a slave pipeline is designed,where the pipeline is used to process the common situation of no byte-output and the slave pipeline is used to detect and process the byte-output situation separately,in order to reduce the critical path of the sub circuits.The synthesis result by the technology library of TSMC 0.18μm shows that the system clock frequency is 578MHz and the throughput is about 520 Msymbols/s.Comparing with the published works by the same technology library at home and abroad,they are optimized by 40% and 26% at least respectively.
LI Jun-hong, LI Ping, ZHANG Guo-jun, ZHAI Ya-hong, XU Jian-bo
Vol. 39, Issue 11, Pages: 2492-2496(2011)
摘要:A push-pull PWM converter based on bipolar process is proposed and verified.The dual outputs status push-pull shifting is realized through the cooperation of a status maintainer and a status switcher.Utilizing the bi-stable race,the circuit is insensitive to the current gain degeneration caused by irradiation.The circuit is manufactured via HUAYUE 2μm bipolar process and irradiated by Co-60 radiation source at the 5.6rad(Si)/s dosage rate and 100krad(Si) total dose.The test results indicate that the circuit is functional correctly while the single NPN device has current gain degeneration of 55%~65%.The bipolar power device driven by converter is fully cut when turned off,and the turn on voltage drop sees 12%~22% increase only.
XU Li-zhong, LI Min, SHI Ai-ye, TANG Min, HUANG Feng-chen
Vol. 39, Issue 11, Pages: 2497-2501(2011)
摘要:Inspired by the theory of interaction between spatial feature and temporal feature in insect visual system,beyond the traditional feature extraction method,this paper proposes a feature detector model for multi-spectral remote sensing image,which is based on the cross correlation of multi-band,and without the limitation of the global window size and parameters.By the analysis in frequency domain,it showed that this feature detector model can respond to the edge and texture by integrating spectral information and spatial information.Beside this,the proposed feature detector model was effective and robust to the random noise.This paper also brings the feature detector model into reconstruction algorithm of remote sensing image.Reconstructed results validate the effectiveness of the feature detector model,which contains rich high-frequency information.
摘要:A novel method with the property of law was for the first time put forward to improve the thermal stability of NiSi film.According to the way,adding a thin Tantalum interlayer within the nickel film was first reported to effectively improve the thermal stability of nickel monosilicide.After rapid thermal annealing (RTA) at temperatures ranging from 600℃ to 800℃,the sheet resistance of formed Ni(Ta)Si samples was about 2Ω/□,and its value is also lower than that of nickel monosilicide without the interlayer.X-ray diffraction (XRD),AES,RBS and AFM results both reveal that NiSi phase exists in these samples,but the high resistance NiSi2 phase does not exist because tantalum interlayer as the diffusion barrier for oxygen element moved to the top of the thin film after rapid thermal annealing.Fabricated Ni(Ta)Si/Si Schottky barrier diodes with the guard ring structure displayed good quality,with the barrier height being located generally about 0.64 eV and the ideality factor approaching unity.Therefore,It shows that Ni(Ta)Si is a satisfactory local connection and contact material.
关键词:nickel silicide;rapid thermal annealing (RTA);X-ray diffraction (XRD);auger electron spectroscopy(AES);rutherford backscattering spectrometry (RBS);atomic force microscopy(AFM)
摘要:A novel radiation-hard FPGA technology mapping method based on partial TMR and logic gate masking and a fast parallel fault injection and Monte Carlo simulation platform are presented.This platform and method have been used in the mapping module which is part of the CAD flow for self-developed FPGA by Fudan University named FDP4.The experimental results show that FDRMap can decrease the circuit fault criticality by 32.62% with the 14.06% area penalty.Comparing to the partial TMR,it decreases the criticality by 12.44% along with reducing the resources by 12.23%.
摘要:Type-based proxy re-encryption is an effective and efficient cryptographic solution to ciphertext delegation and distribution,which also enables the ciphertext owners to carry out fine-grained delegation control.In this paper,we propose a new type-based proxy re-encryption scheme without pairings and prove it against the chosen-ciphertext attack (CCA) in the random oracle model.Compared with the existing type-based proxy re-encryption schemes constructed with pairings,our scheme owns better efficiency as well as the master key security.
摘要:Three-way decision-theoretic rough set model is a probabilistic extension of the algebraic rough set model.The required parameters for defining probabilistic lower and upper approximations are calculated based on cost functions through Bayesian decision procedure.Through providing different cost functions,decision-theoretic rough set model can derive many other probabilistic rough set models,such as variable precision rough set model,etc.This paper constructs an optimum problem based on decision-theoretic rough set model.Through solving the optimum problem,one can get the proper cost functions and thresholds without any preliminary knowledge.An adaptive learning parameters algorithm is also proposed to solve the optimum problem.The search space of the algorithm is the set of all instances'probabilities.Under the three-way decision-theoretic rough set model which is based on the learned cost functions and thresholds,the decision cost is minimal and a better classification performance can be gotten from that.The experimental result on some data sets shows the efficiency of our algorithm.
关键词:three-way decision-theoretic rough set model;cost functions;thresholds;optimum problem
摘要:Aiming at recognition problem caused by variations in lighting,facial expression etc and small sample size problem which generally exist in face recognition,this paper proposes a novel recognition method,called RB2DLDA,which fuses bidirectional 2DLDA feature by resampling technique.In matrix-based LDA,2DLDA uses inner-class and intra-class covariance information in the vertical direction,and E2DLDA uses inner-class and intra-class covariance information in the horizontal direction,this paper proves in theory that class information in the two directions is complementary to each other to some extent,provides theoretical foundation for fusing class information in the two directions to design classifier and improves recognition performance of the classifier.Moreover,proposes an adaptive algorithm of setting feature reduction parameter.Our method is tested on AR and CAS-PEAL-R1 face databases,and the experimental results show that the proposed approach possesses higher recognition accuracy and robustness than other state-of-the art methods.
关键词:face recognition;resampling;bidirectional two-dimensional linear discriminant analysis (2DLDA)
QI Hong-zhi, SUN Chang-cheng, XU Min-peng, MING Dong, WAN Bai-kun, LIU Zhi-peng
Vol. 39, Issue 11, Pages: 2534-2539(2011)
摘要:P300 Speller is an important information transforming method in Brain-Computer Interface.However,because the signal-to-noise ratio is comparably low and the size of training sample set is too large,the recognition efficiency in P300 Speller is not ideal by using popular linear recognition method or non linear methods such as Support Vector Machine.This paper introduces a novel method based on Adaptive Boosting SVM using weighted resampling process.We use AdaBoost resampling method to construct a series of little size training sample sub sets on the large integer set and then train the SVMs on these sub sets,finally run the recognition by combining these SVM's output.Using 6 subjects' EEG features from P300 Speller character identifying experiment we find that this method improves character identifying accuracy significantly.We achieve a character identifying accuracy of 97.5% on combining 5 repetitive stimulus features.A further evaluation using international BCI Competition dataset has proved that this method achieves a 7.35% enhancement in character identifying accuracy and a 48.9% enhancement in information transforming velocity on combing 5 repetitive stimulus features.This study demonstrated that ABSVM has the ability of improve recognition accuracy and information transforming velocity in P300 Speller and it is worthy of a further study and development.
关键词:adaptive boosting support vector machine (ABSVM);event related potential (ERP);brain computer interface (BCI);weighted resampling
JIN Di, LIU Da-you, YANG Bo, LIU Jie, HE Dong-xiao, TIAN Ye
Vol. 39, Issue 11, Pages: 2540-2546(2011)
摘要:Recently,complex networks are always very huge and take on distributed nature.Therefore it is gradually becoming instant requirement to propose fast network clustering algorithms in the sight of local view.For the problem,this paper deduces a local objective function f aiming to each node in the network,which is based on the profound analysis on network modularity function Q,and proves that Q is monotone increasing with function f of any node,and then proposes a fast network clustering algorithm (FNCA) by using local optimization.In this algorithm,each node optimizes its own objective function f by only local information,and all the nodes collectively optimize function Q to detect network community structure.Both efficiency and effectiveness of algorithm FNCA are tested against computer-generated and real-world networks.Experimental result shows that this algorithm is better than some excellent network clustering algorithms in term of these two respects.
摘要:The rapid growth of the number of mobile services provides enormous potential for mobile users in different contexts to find mobile information of interest.Mobile user preferences elicitation has been used as a valid means to ease the "mobile information overload" problem.Inspired by the model of level of processing and the theory of distributed cognition,a cognitive psychology-based approach to user preferences elicitation for mobile network services is proposed.It uses a six-tuple to describe the data structure of user preferences information,analyzes the level of processing of services for users to elicit context-free user preferences,then identifies valid types of contexts as well as their influences on user preferences,and finally elicits converged user preferences.Experimental comparisons of this approach against some baseline methods with a synthetic data set have been conducted,showing improvements in performance.
关键词:mobile network services;user preferences;level of processing;distributed cognition;context
摘要:A novel approach to license plate locating and identifying based on the Curved Field Space is proposed in this paper.Firstly,by detecting the brightness information in the color vehicle image,the gradual gradient vectors are fitted,and then the dot matrix depiction with curved quantum is used to describe the gradual gradient vectors and the color information is eliminated.Then,the sampling curved quantum groups are merged into sub-curved spaces which include car's body,plate and chars.The sub-curved spaces of plate and chars which are formed of curved quantum groups constitute the curved field space by linking up with each edge curved quantum smoothly.Finally,recognition results are obtained by depth and dimension both of which are extracted from Curved Field Space.Experiments show that the proposed approach has overcame the problem of feature rotating,varied angle,wear and fuzzy chars,and achieves good locating and identifying results.
关键词:curved field space;curved license plate;curved quantum;field order dimension;vehicle license plate locating and identifying
摘要:We proposed two methods to measure the similarity of normal cloud models.One uses the expectation curves to reflect the overall feature of cloud models and to calculate the similarity by the expectation curves. The other uses the maximum boundary curve to compute the similarity between different clouds.The two methods can obtain a qualitative result,which overcomes the traditional deficiencies of the high time complexity,unstable result and excessively remarkable expectation character.The experimental results demonstrate that our methods can calculate the similarity of cloud models objectively and improve the efficiency of the algorithms in collaborative filtering recommendation and time series classification.
摘要:Targeting the problem of weak time awareness,low service utilization and poor composition reliability in the process of web services composition,we propose to bring quantitative timed properties into web services interaction adaptation framework and make efforts on researching on time-aware web services interaction modeling,automated interaction mismatch detection and automated adaptor protocol generation.First,we propose timed service protocol (TSP) model to represent interaction behavior of single time-aware web service and timed service protocol network (TSPN) model to represent concurrent composition of multiple time-aware web services;Second,we transform the interaction mismatch detection problem into the determination problem of compatibility of TSP models,define the mapping rules from time-aware web service descriptions to TSP models and from TSP models to UPPAAL specifications,give the CTL representations of compatibility properties and implement automatically interaction mismatch detection of time-aware web services based on model checker UPPAAL,whose result determines whether it is necessary to build adaptor for the next step.Finally,experiments demonstrate feasibility and validity of above idea.
关键词:time-aware web services;timed service protocol;timed service interaction;model checking;mismatch checking
WANG You-ren, HUANG Yuan-yuan, FENG Ran, ZHANG Zhai
Vol. 39, Issue 11, Pages: 2576-2582(2011)
摘要:The synthesis method of quantum reversible logic circuits based on the genetic algorithm was studied.The multi-objective optimal design of reversible logic circuit's function,the number of quantum gates,the number of garbage bits and quantum cost was achieved by new synthesis method.The mathematical model for quantum reversible logic circuit synthesis was established,the matrix encoding method of quantum reversible logic circuits was proposed,the evolution operations and the multi-objective evolutionary algorithm for quantum reversible logic circuit design were developed.The 8 bits reversible multiplier was designed as an example,the experimental results show that the proposed multi-objective evolutionary design method for quantum reversible logic circuit synthesis is correct and effective.
摘要:Optimal resource allocation is the core problem in wireless multicast.An adaptive Multicast Control algorithm (AMC) is proposed.It uses segmented bandwidth allocation to reduce iterations for medium-speed and high-speed mobile Ad Hoc network.It overcomes the disadvantage of average resource allocation by adopting marginal utility and adaptive access decision mechanism.And a bidding mechanism is used to improve the quality of heterogeneous bandwidth demanding for multicast traffic.Simulation results show that AMC policy has excellent performance for mobile network,especially for layered heterogeneous multicast in dynamic network condition.
关键词:Ad Hoc network;heterogeneous multicast;nonlinear optimal
摘要:An improved deadlock and livelock free resource co-allocation protocol called OODP3 (Optimal ODP3) is proposed.OODP3 utilizes the safe state method in the framework of ODP3 to avoid deadlock and livelock.However,OODP3 reduces its time complexity to polynomial level and theoretical proof is conducted to verify the validity of OODP3.Experiment results show that OODP3 not only executes faster but also achieves a better performance than the existing deadlock and livelock avoidance protocols.At last,how back-off protocol and resource allocation policy affect the performance of OODP3 is further discussed.
摘要:Attribute reduction is a key studying point of rough set theory.It has been proven that computing minimal reduction of decision tables is a NP-hard problem.An efficient attribute reduction algorithm based on the quantum frog-leaping co-evolution is proposed.A dynamic multi-cluster frog structure is designed,individuals are represented by multi-state qubits.The self-adaptive adjustment of quantum rotation angle,quantum mutation and quantum entanglement strategies are applied to accelerate the convergence.Cooperative searching information of different clusters during attribute reduction is shared by adopting the bidirectional co-evolutionary method.Experiments on some benchmark problems indicate the proposed algorithm has outstanding ability to balance the global exploitation and local exploration on condition of the good convergence,and results on some UCI data sets validate it is more competitive on the attribute reduction accuracy and efficiency,compared to the traditional evolutionary algorithms.
YANG Yang, WANG Yong-gang, CHEN Wei, WANG Shuang, ZHANG Li-jun, HUANG Da-jun, ZHANG Tao, HAN Zheng-fu
Vol. 39, Issue 11, Pages: 2604-2609(2011)
摘要:Phase coding is frequently used in current quantum key distribution(QKD) systems.It has posed many technical challenges on the design of the control electronics.After the description to the overall design of our phase coding QKD system,the paper focuses on the special requirements and implementations on the synchronizing mechanism and automatic phase drift scanning and compensation.The electronics designed in this paper combining with the optical devices can be an independent node device of a QKD network.We have successfully used these node devices to build a practical encrypted communication system through the commercial communication optical fiber system of Wuhu Telecom in Anhui Province.Encrypted communication between multiple nodes by quantum cryptography is implemented in this system.The QKD system in the network actualizes BB84 phase coding protocol at 20MHz repetition rate of light pulses.It achieves quantum bit error rate(QBER)2.95% and secure key generation rate 4.91kbps with a 20km transmission distance.
ZHOU Yan-ping, FU Sen, YU Si-yuan, WANG Jian, SONG Yi-wei
Vol. 39, Issue 11, Pages: 2610-2614(2011)
摘要:Due to the optical system is not in the rotating mechanism which can greatly reduce the weight of system,Periscopic Structure has been widely used in recent years in photoelectric terminal.But its periscope asymmetric structure always faces the phenomenon of cold welding,non-linear friction torque which caused by the low-temperature vacuum in space environment.In order to solve these problems and also to improve the control precision,a "dynamical lubricate" conception is proposed in this dissertation,which based on the dual-loop PID control,it includes position-loop and velocity-loop controller and a high-frequency pulse controller.The controller can lubricate the bearing and make the system in a dynamically status by using "soft" algorithm.The friction torque is compensated and the range of the controller linearity is ensured.At last a space temperature simulation experiment was held to test the effect.The temperature range was set from-30 degree to 50 degree.The positioning accuracy in the experiment is less than 10μrad,sinusoidal tracking accuracy is less than 28μrad.This result shows that by using the new control algorithm,an high-precision control needs is achieved during the great range of temperature and the tracking errors are distributed evenly.
摘要:This paper deals with the optimization of subarray partition operated in large array radar.The theoretical basis of the optimal subarray partition is analyzed based on the principle of minimizing the "residual error" of excitation coefficients.The judgment basis about the contiguous characteristic of the optimal subarray partition is presented,and an innovative approach for subarray partition is proposed.Comparing with the traditional method based on clustering algorithm,the new method can decrease the "residual error" father more,and hold better beam performance.Numerical experiments are reported when the construct of the array and the feed architecture of sum and difference beams are given.Comparing research with two old methods is also presented,and the results demonstrate the efficacy of the proposed methods.
关键词:array signal processing;phased array antenna;subarray partition;sum and difference beams;monopulse;clustering algorithm
GAO Feng, ZHAO Zhu-xin, WEN Gong-jian, LU Huan-zhang
Vol. 39, Issue 11, Pages: 2622-2626(2011)
摘要:For the influences of noises and feature extraction method,uncertainties are exiting in the extracted features,which decrease the degree of similarity between corresponding features,and enlarge the matching illegibility.If the uncertainties are not analyzed,they may decrease the accuracy of matching and even lead to a failure.Hence,the paper analyses and models the uncertainties of the features,and then proposes a method which bases on Vague sets to compute the similarity between features with uncertainties.The study result can be used in the fields of image registration and target recognition which relate to feature matching.
摘要:Based on the in-depth analysis of the massive real AS-level topology data,we abstracted the multi characteristics describing AS-level topology's hierarchy and AS relationships,and gived material modeling ways about network's hierarchical partition and business relationships between ASs.Then,we proposed a novel algorithm for generating synthetic graphs annotated with AS relationships named ASRTM based on the k-core decomposition.Finally,the numerical simulations and analysis results compared with the real AS-level topology data show that most of macroscopic statistics closed to real networks',including power-law degree distribution and AS relationships properties.Otherwise,ASRTM,as a conferring algorithm about AS relationships,can give a graph annotated with AS relationships for given undirected topology graph.
关键词:modeling on Internet AS-level topology;AS topology annotated with AS relationships;Annotation-degree distributions;network hierarchy;the k-core decomposition
LI Jian-jiang, CUI Jian, WANG Dan, YAN Lin, HUANG Yi-shuang
Vol. 39, Issue 11, Pages: 2635-2642(2011)
摘要:Through well-defined interfaces and runtime support library,MapReduce parallel programming model can automatically perform the large-scale computing tasks in parallel,hide the underlying implementation details,and reduce the difficulty of parallel programming.This paper reviews the domestic and overseas research of the MapReduce,describes and analyzes the characteristics and lack of the typical research achievements about MapReduce at home and abroad.Then this paper focus on the in-depth analysis of the key technologies about MapReduce (including:model optimization,model implementation according to the different platforms,task scheduling,load balancing,and fault tolerance).Finally,this paper prospects the MapReduce for the future trend.
关键词:MapReduce;parallel programming model;runtime support library;massive data processing
摘要:In distributed cognitive radio networks (CRNs),a common fusion centre is always not available to collect the sensing results from all cooperative users,and the sensing cooperation is likely to be disrupted by spectrum sensing data falsification (SSDF) attacks.Allowing for these factors,this paper proposes a modified consensus-based scheme for decentralized cooperative spectrum sensing(CSS).Utilizing the iteration rule based on Metropolis weight matrix,each secondary user (SU) can maintain cooperation with others only through the local information exchange in the neighborhood,without requiring any a priori knowledge of the whole network.To counter three potential attack models of SSDF,some anti-attack strategies are also introduced,by which the authentic SUs can identify and reject the malicious users quickly.As represented by the extensive simulation results,the proposed scheme can generally guarantee most authentic SUs to reach a consensus and make right decisions individually; also,it is proved to be much more robust against all potential SSDF attacks,compared with the existing consensus-based scheme.
关键词:cognitive radio networks;cooperative spectrum sensing;sensing security;consensus;spectrum sensing data falsification attack
ZHU Si-feng, LIU Fang, QI Yu-tao, CHAI Zheng-yi, WU Jian-she
Vol. 39, Issue 11, Pages: 2648-2653(2011)
摘要:To solve the joint call admission control problem in heterogeneous network environment of single operator,a novel admission control algorithm based on immune computing was proposed.The mathematical model of joint call admission control problem was expounded,the framework of admission control solution based on immune multi-objective optimization algorithm was given,and simulation experiments were done to validate proposed solution.Experimental result shows that the proposed solution,compared with other solutions,obtains better performance tradeoffs between recent utility and frequency spectrum,balances preferably each radio access network of the same operator,and has the advantage of good application value.
摘要:Routine nonlinear estimation algorithms have problem of long calculating time and slow convergence speed,when they deal with linear system Huber estimation with great amount of data and high dimension of parameter.Firstly this paper proposes the orthogonal searching algorithm according to the character of Huber estimation,then deduces the algorithm and process to calculate Huber estimation by using orthogonal searching algorithm,finally compares orthogonal searching algorithm with classic furthest falling algorithm through simulation,and the result shows that orthogonal searching algorithm has great advantage to Huber estimation when the amount of data is great and the dimension of parameter is high.
摘要:Taking full account of the high reliability of nowaday network technology,the instance of unsuccessful service can be ignored.This paper re-defined the system heavy load condition at embedding points,under which the performance of double-star is re-evaluated through the mathematical modeling.As the result,the customer transfer probability Pij and the mean detained time of the customers are given,and then the performance evaluation is done through simulation.Compared with previous performance evaluation,with the concise mathematical analysis and conclusion,this investigation has not only theoretical value,but also practical value.
YOU Zhi-qiang, PENG Fu-hui, KUANG Ji-shun, ZHANG Da-fang
Vol. 39, Issue 11, Pages: 2663-2669(2011)
摘要:As the advances in integrated circuit technology,what multi-core and many-core are used on system-on-chip will be a trend.The shortcomings such as inefficient communication,high power consumption and long delay in the traditional two-dimensional grid (2D-mesh) topology will become clearly.In this paper,firstly we analyze and compare the performance about several commonly used topologies in multi-core and many-core.After that,we employ butterfly-fat tree (BFT) topology which has lower routing complexity and better performance among the topologies to solve problems.A multicast routing protocol and design of the node encoding are proposed to realize multicast test for the same kinds of cores.This paper also implements a comparing logic in the routers for test response data comparison.This method significantly reduces test application time,storage of automatic test equipment (ATE) and total communication energy consumption.Simulation experimental results verify these conclusions.
关键词:design for testability (DFT);system-on-chip (SOC);network-on-chip (NOC);multicast testing;low-cost test
ZHANG Jian-wei, FANG Lin, CHEN Yun-jie, ZHAN Tian-ming, LI Xiao-tian
Vol. 39, Issue 11, Pages: 2670-2673(2011)
摘要:Left ventricle MRI segmentation active contour model based on local and global features was presented in this paper.The local fitting term contained means and variances of local region,overcame the influence of intensity inhomogeneities.The global intensity fitting term,which conquered the unexpected local minimum stemming from image local intensity and deals with weak.The experiment results demonstrate that the algorithm is effective for segmenting the left ventricle MR images.
PENG Bao-hua, ZHOU Jing-lun, FENG Jing, LIU Xue-min
Vol. 39, Issue 11, Pages: 2674-2679(2011)
摘要:The metallized film capacitor is one of the most important types of components of the laser driver in inertial fusion facility.The residual lifetime of the capacitor is of great value in maintenance and spare parts decision making.We used the Wiener process to model the degradation process based on an analysis of the capacitance degradation mechanism of the metallized film capacitor.Taking the inherent difference of each capacitor into consideration,we proposed to use the Wiener process with random effect,where the drift and diffusion parameters of the Wiener process were treated as random variables,whose distribution were estimated using the degradation data from other capacitors from the same batch.When predicting the residual lifetime of a particular capacitor,a Bayesian method was used to integrate its degradation data and the batch information.Posterior distribution of the residual lifetime parameters were deduced.The proposed method can improve the precision of prediction when its degradation data are few.
关键词:metallized film pulse capacitor;residual lifetime prediction;Bayesian method;Wiener process;Information fusion
摘要:We propose a transmit power optimization method for distributed transmit antenna space-time block coded orthogonal frequency division multiplexing (STBC-OFDM) systems in multipath Rayleigh fading channels with multiple residual frequency offsets (RFOs).We assume that the transmitter is equipped with two distributed transmit antennas and the receiver uses the decision-feedback detector.The proposed method is designed to minimize the lower bound of the average bit error ratio (BER) and the optimal transmit power for each transmit antenna is calculated according to the average channel power gains,the RFOs variance,and the noise variance.Simulation results show that the proposed method can save the transmit power while enhancing the BER performance compared with the conventional full power transmit method.
关键词:wireless communication;space-time block code(STBC);orthogonal frequency division multiplexing(OFDM);transmit power optimization;residual frequency offsets
摘要:An optimal mismatched filter design method,based on the second order cone programming (SOCP),is proposed for the suppression of sidelobes of the quasi-step frequency waveform.On the basis of derivation of the SNR loss expression,practical application consideration is given to the optimal mismatched processing criteria.A conditioned Min-Max criterion is developed,i.e.,to minimize the maximum sidelobe level with an upper limitation of SNR loss.The optimal mismatched filter design problem is then converted to the dual problem in SOCP.The filter coefficients are obtained by using efficient primal-dual interior-point method for solving SOCP.Thus,sidelobes suppression with constant loss in process gain is implemented.
关键词:step frequency waveform;second-order cone;optimal mismatch;sidelobe suppression
摘要:A method of measuring light source angles is presented based on the optical line transmission characteristic.The sensors on the polyhedron flats to measure light source angles are applied.The light source angles measuring is changed into the geometry angles computing in the space geometry by means of constructing equivalent vector model,so that the measurement is simplified.The azimuth and the pitching angle between light source and the measured point are included in the measuring method.At the same time,this method is suitable for light source angles measurement in the sky or on the ground.In this way,the prior art measurement range insufficient is resolved.The results of the measurement for solar angle prove that the method is correct and feasible.
关键词:light source angles;measuring method;polyhedron;vector model
摘要:A versatile class-AB analog multiplier was presented in this paper.The current squarer/current follower,based on class-AB current mirror and MOS trans-linear (TL) principle,is the basic building block in realization scheme.Since the multiplier can be driven by current,voltage or both voltage and current without changing the internal circuit structure and element parameters,so it is a versatile analog multiplier.The major advantages of this multiplier are wide band-width,low power consumption,and good linearity.As a result,it can be considered as a useful building block in current-mode and voltage-mode circuits.PSPICE simulator with TSMC 0.35 μm CMOS technology parameters is employed to verify the circuit.
摘要:Considering the frequency spectrum of the piezoelectric ceramic transducer,this study uses a bi-phase coding signal modulated by a sine carrier as the excitation mode.Through using Barker code signal with each chip modulated by different periods of carrier,the transducer effect on the exciting signal and compression result is analyzed,modulation parameter is optimized,and the best matching mode between the exciting signal and the transducer is discussed.The echo of four target wires is processed by pulse compression with the optimum signal,the result shows that this excitation mode has a good performance on the transducer response,and it can enhance the SNR of the ultrasonic signal significantly.
摘要:Neurons with complex-valued weights have a stronger performance due to the multi-threshold logic.In the paper,the robustness for this kind of neurons is presented in accordance with its mathematical model and the binary perceptron's definitions of robustness.The robust design for basic digital logics of multiple variables or XOR logic is proposed after that and the robust algorithm for any Boolean functions base on the neurons is given,also the correctness of the algorithm is proved.The final use of specific examples not only shows that the algorithm can realize an arbitrary Boolean function.It also shows that the neurons used to achieve the flexibility and advantage of realizing any digital logic function,illustrates its powerful logical processing capabilities.