摘要:A self-tuning,adaptive 1.9GHz fractional-N/integer PLL based frequency synthesizer is proposed in the paper.A combined tuning technique of digital tuning and analog tuning is used to improve the phase noise of frequency synthesizer by decreasing the gain of VCO.The adaptive loop is introduced for automatic adjustment of the loop bandwidth,which can quicken the locking process.Two operation modes (integer/fractional-N) are achieved by switching on/off the output signal of ΣΔ modulator.Just a programmable counter is needed for the swallow pulse divider.The on-chip VCO achieves a low phase noise by utilizing a bias filter technique and a differential inductor,and a 1.7GHz~2.1GHz tuning range by a switched capacitor array.Based on 0.18 μ m 1.8V SMIC CMOS technology,SpectreVerilog simulation shows that the frequency synthesizer has a 100 kHz loop bandwidth,a<15 μ s settling time,and the phase noise is lower than -123dBc at 600kHz offset.
WANG Xiao-ying, ZHAO Hai, CHEN Ying-ge, YIN Zhen-yu
Vol. 34, Issue 5, Pages: 778-783(2006)
摘要:Limited power supply is the uppermost constraint factor to sensor network performance and lifetime.Based on processor task scheduling of sensor network node,this paper put forward voltage and sleep dual-effective scheduling algorithm in term of dynamic power management and dynamic voltage/frequency scaling,namely DV/FS-RM and DV/FS-EDF.On the premise of guarantee node real time performance,the policy can not only dynamically control idle task to enter sleep mode but also reduce processor frequency to achieve better power save effect through DV/FS-RM or DV/FS-EDF algorithm.Experiments show that the effect of the power aware task scheduling algorithm is remarkable which can extends the lifetime of sensor network node several times which mainly depend on batteries power supply.
关键词:sensor network;dynamic power management;dynamic voltage/frequency scaling;task scheduling;DV/FS-RM;DV/FS-EDF
摘要:It is conventionally assumed that the average current controlled Power Factor Correction (PFC) boost converter operates in Continuous Conduction Mode (CCM),and hence the average CCM model is used for analyzing the dynamics of the converter.In fact,the converter might work in Discontinuous Conduct Mode (DCM),especially,in the light load condition.Therefore,the average CCM model cannot predict the dynamical characteristics accurately,especially bifurcation and chaos phenomena.The state space model,which covers possible DCM operation,is employed to investigate the low frequency bifurcation of the PFC converter.Low frequency bifurcation diagram is derived,which makes the route from period doubling bifurcation to chaos clear and makes a better understanding of the dynamics of the converter.The stability of the converter is investigated from the viewpoint of bifurcation,whose results show that the output capacitance and the load make great contributions to the stability of the converter.One dimensional bifurcation diagram and two dimensional parameters stability map give the hint to better understand and design the converter.The consistence of the simulation results and experiment results shows that the proposed method is reasonable and the results are correct.
摘要:In Simultaneous Multithreading (SMT) processors,co-scheduled threads share the processor's resources,but at the same time compete for them.A thread missing in L2 cache may occupy most of available resources for a long time,causing other threads run slower than they could or even stall because of lack of resources.As a result,the overall performance of SMT processors is degraded.In this paper,we propose a novel fetch policy called MFP (Multiple Fetch Priorities) to prevent the negative effects caused by L2 cache misses.Results show that our policy outperforms previously proposed fetch policies for all types of workloads,especially for memory bounded workloads,whether using IPC as a metric or using the harmonic mean as a metric.Results also tell that our policy shows different degrees of improvement over other fetch policies.The increment over PDG is greatest,reaching 19.2% in IPC and 27.7% in Hmean on average.
HUANG Yi-hua, YANG Zhi, CHEN Zhi-rui, SONG Jun-jiang
Vol. 34, Issue 5, Pages: 800-803(2006)
摘要:When the electro-magnetic transformer is used in the cold cathode fluorescent lamp (CCFL)driver,the transformer is very easy to be broken,the EMI and the power lose is very serious,and the weight and volume is hard to be minimized.If the electro-magnetic transformer is replaced with the piezoelectric transformer (PT)in the CCFL driver,and using the Half-Bridge inverter as the input stage using,the LC circuit as the output stage of the PT,and to trace the frequency of the PT,the performance of the CCFL driver can be improved.
ZHOU Jun-sheng, DAI Xin-yu, YIN Cun-yan, CHEN Jia-jun
Vol. 34, Issue 5, Pages: 804-809(2006)
摘要:Automatic recognition of Chinese organization name is a very difficult problem in many NLP tasks.This paper presents a new algorithm of Chinese organization name recognition based on cascaded conditional random fields.In the proposed algorithm,the person name and location name are first recognized by the lower model.The result then is passed to the high model and supports the decision of high model for recognition of the complicated organization names.We experimentally evaluate the algorithm on large-scale corpus.In open test,its recalling rate achieves 90.05% and the precision rate 88.12%.The evaluation results show that the algorithm based on cascaded conditional random fields significantly outperforms previous methods.
关键词:named entity;Chinese organization name recognition;conditional random fields
摘要:To aim at video structure analysis,a new shot boundary coefficient model is proposed.Firstly,windowed frame differences are calculated using the special approaches,which have different characteristics for intra-shot frames and shot transition frames respectively.Secondly,shot boundary detection operators are used to calculate shot boundary similarity coefficients.If shot boundary similarity coefficients are bigger,corresponding frames are more probably shot boundaries.Thirdly,to improve shot boundary similarity coefficients,shot boundary coefficients are defined based on shot boundary similarity coefficients.Shot boundary coefficients have many good characteristics,and are used to not only detect shot boundaries independently,but also combine traditional shot boundary detection methods for shot boundary detection.Therefore,a robust shot boundary detection framework is provided.Experimental results show that shot boundary detection methods,based on the shot boundary coefficient model,can improve the performance of shot boundary detection.
摘要:To solve the trust problem during resource access in grid service and to avoid subjective notion during the execution of subjective trust,the trust mechanism based Beta function is proposed.The trust relations are set up upon recommendations among nodes from the reference of social network relation.Through judging the recommendation ability of medium nodes according to the former interaction result with the method of probability and statistics,the best recommendation ability nodes are selected as the recommenders to search trust links according to the limitation of the requester about link recommendation value and recommendation number,the trust values of each attribute for the resource node presented by the last nodes on the trust links are synthesized with Beta function to decide whether the access to resource node can be enforced.The experimental result expresses its validity.
PENG Yan-bing, GONG Jian, LIU Wei-jiang, YANG Wang
Vol. 34, Issue 5, Pages: 822-827(2006)
摘要:An original element reconstructing algorithm named Reconstruction with Semantically Enhanced Counting Bloom Filter (RSECBF) is proposed to timely recover the original element in set S from Counting Bloom Filter's hash space.The semantically enhanced hash function is based the ideas that,the independent hash space preserved for each different hash function eliminates the internal confliction among hash functions,and the hash function could be extended from the uniform distribution to any distribution;the overlapping of hash bit strings bring the ability to recover the original string by the uniqueness of hash mapping process and the hits amount balance of overlapped hash string.The recovery algorithm is greatly simplified for the Pareto distribution when only the principal component is analyzed.For Directly Bit String Selecting,the reproduced longest string just is the distribution character of the original strings.The simulation and the validation with published data trace suggested that the recovery result of RSECBF is acceptable.It can be used to find the network behavior characteristics when abnormal behavior bursts in the real networks.
HUA Wei, BAO Jian-jun, YANG Xiao-qing, HUANG Ka-ma
Vol. 34, Issue 5, Pages: 828-832(2006)
摘要:The study on the effective permittivity of mixed solution is an important work to Microwave-Chemistry.Some new characteristics of effective permittivities of mixed solution have been found and can not be explained by the classic theory,such as Clausius-Mossotti’s,Onsager’s and Kirkwood’s formulas.The effective permittivity for biphase organic solution at microwave frequency has been obtained by the perturbation method.The results show that the real part (or the imaginary part) of the effective permittivity may has a peak value with respect to the fraction of volume.Where the effective permittivity is larger than any of that of the components while the real part of permittivity (or the imaginary part) of components is closed to each other and the imaginary part of permittivity (or the real part) is quite different.The analysis has been done to explain this new characteristics of effective permittivities by the symmetric Bruggrman's theory.
摘要:As a main video transmission mode for digital media networks,the capability to predict VBR video traffic can significantly improve the effectiveness of quality of services.Therefore,aiming at the complex characteristics of MPEG videos,a novel intelligent integrated traffic prediction model is proposed based on fuzzy and neural network.The prediction error is reduced by the fuzzy predictor,and the implementation of neural network is used to lower prediction computation for real time.Simulation results show that the proposed method is able to predict the original traffic more accurately than the normal AR method and can be easily applied into other methods.
关键词:variable bit rate of moving picture experts group (VBR MPEG);video traffic;fuzzy neural network (FNN);intelligent integrated;prediction model
ZHANG Wen-bo, ZHAO Hai, SU Wei-ji, XU Ye, WANG Jin-dong, HAN Guang-jie
Vol. 34, Issue 5, Pages: 837-844(2006)
摘要:In open environments such as the Internet agents from heterogeneous sources could exist and interact to accomplish joint activities.The general Contract Net Protocol model which works on bids invitation between a Manager agent and Contractor agents can successfully realize the cooperation among agents.But it also faces many problems such as the high network traffic and low adaptation to the changes of the contractors' capacity and load.So in this paper,the concepts of grade field and timing consumption are put forward and introduced into the general contract net protocol and a new model of grade field and timing consumption contract net model is made.By dividing three different grades subfields according to the capability and success rate in all the service agents,the intercooperation of the service agents is limited in grade field 1.The theoritic analysis and the experiments show that this new model can reduce the searching range and lighten the network.traffic.Furthermore it can balance the load of the contractors.
关键词:service agent;cooperative problem-solving;grade field;time consumption;contract-net;grade field and time consumption contract-net model
LU Jian-jiang, XU Bao-wen, KANG Da-zhou, LI Yan-hui
Vol. 34, Issue 5, Pages: 845-851(2006)
摘要:The approximate information retrieval approach finds least upper bounds of a concept and then uses them to get upper approximation of the concept to solve this problem of ontology heterogeneity.However,the current method considers the bounds only containing separate concepts,so it cannot get the least upper approximation of the concept.In this paper,disjunction of the concepts is introduced to define multielement least upper bounds,and the approximation based on them is proved the least upper approximation of a concept.In general,multielement least upper bounds may contain much redundancy,which will increase the expression complexity of the least upper approximation of a concept.We also define the simplified multielement least upper bounds and provide effective algorithm to find them last.
摘要:In analysis of nonlinear network responses,Volterra Series can be used to educe the nonlinear transfer function similar to the transfer function of linear systems.Via utilizing the nonlinear transfer function,it can make precision analysis for nonlinear system by using the methods of linearization and systematization.A set of continuous calculating formulae of nonlinear network responses was proposed based on the solutions of Volterra Series.In order to solve the problem of computing the continuous formulae,the block-pulse function transform was adopted to deal with the repeatedly computation of a set of generalized convolution integral which described the relationship between the nonlinear network response and its input by using Volterra series.This method could result in a set of discrete calculating formulae which could be used to obtain the response of nonlinear network based on the solutions of Volterra series.The results of simulation showed that the curve of nonlinear network response produced by this algorithm was very closed to the curve of the real model and proved the availability of the method.
关键词:Volterra series;nonlinear network;block-pulse function transform
WANG Ben-nian, GAO Yang, CHEN Zhao-qian, XIE Jun-yuan, CHEN Shi-fu
Vol. 34, Issue 5, Pages: 856-860(2006)
摘要:RLGA,an algorithm which implements mechanism of reinforcement learning under the framework of genetic algorithm is described.by using gene space division the algorithm maps the gene space of genetic algorithm into the strategy spcaces of multi-agent.The convergence theorems for the algorithm are presented,and the time and the space efficiency of the algorithm as well as the relation between them and the division granularity are discussed.The experimental results show that RLGA has well global convergence performance,and the further experiments provide the guide range of the size of gene space division in RLGA.
摘要:As the traditional algorithms for local histogram equalization is used to enhance the contrast of an image,the output brightness is always regardless of the input image.However,this is not the desirable property in many applications where preserving the original brightness is essential,so a new algorithm of local histogram equalization is proposed.Based on the algorithm of partially overlapped sub-block histogram equalization (POSHE),we change the way of equalizing sub-block:based on the mean,each sub-block is separated recursively into a series of sub-images with different gray ranges,then the histograms are equalized over their respective gray ranges,and finally the output of the sub-block is composed with the resulting equalized sub-images.The simulation results show that,with the new algorithm,the local details on the image can be enhanced effectively,and the original brightness can also be preserved in the meantime.Besides,scalable degree of brightness preservation may be achieved by choosing the number of recursive separation.The proposed algorithm can enhance the image in a more natural manner and hence,is very useful for many commercial purposes such as consumer electronics where brightness preservation is required.
摘要:A network collaboration framework based on High Level Architecture and Multi Agent System is introduced in this paper.Some required properties about network collaboration are presented.A network collaboration model and its algorithm HOAM are proposed based on HLA ownership and agent multicast.The shortage of being not satisfied with message causal order in the current implementation RTI of HLA is solved.Causality constraint supported by HOAM algorithm is proved.A comparative analysis is made in terms of the performance of HOAM and LICRA algorithm.It shows that HOAM,as a simple pessimistic algorithm,has the approximate same performance with LICRA,as a complex optimistic algorithm.
摘要:A new type of authenticated encryption schemes is proposed.It first combines signature with message recovery schemes and symmetric encryption schemes together.This type of schemes has the following two characteristics:the commitment value of the signature is only recoverable by an intended receiver,so the value also can be used as a shared symmetric key of symmetric encryption and acts as "two birds one stone".The validity of the received message is verified by comparing the recovered message from the signature with the decrypted message instead of using Hash or Redundancy functions in traditional method.Therefore,the proposed scheme is an authenticated encryption without using Hash functions or Redundancy functions.
关键词:cryptography;signature with message recovery;authenticated encryption;Hash function;Redundancy function
摘要:ID-based public key cryptography can be a good alternative for certificate-based public key setting,especially when efficient key management and moderate security are required.This paper proposes a primitive called mixed verifiably encrypted signature scheme (MVESS) and presents an efficient MVESS based on the ID-based signature scheme due to Cheon,et.al.Then,we provide some theoretical discussions for the security model for MVESSs,and show that our new scheme can be proven to secure with the hardness assumption of the computational Diffie-Hellman problem of pairings in the random oracle model.Our new scheme can be used as blocks to build efficient ID-based optimistic fair signature exchange protocols,which can be widely used in signing digital contracts,e-payment and other electronic commerce.
摘要:Drawing idea from the quantum secure direct communication (QSDC) and the dense coding,we propose a novel quantum secret sharing (QSS) scheme with high efficiency based on the GHZ states.Alice shares her secret message with Bob and Charlie directly using quantum correlations and a public string K,rather than shares a joint key with them and uses the joint key to transmit her message.In our scheme,a GHZ state can be used to share two bits message.We analyze the security of our scheme in the realistic and lossy channel respectively,and show it is necessary to use a public K due to the difference between quantum direct secret sharing and QSDC.
关键词:quantum secret sharing;quantum secure direct communication;dense coding;quantum direct secret sharing
摘要:A new blind restoration model based on PDE was proposed for image restoration,which improved the Chan and Wong's alternating minimization (AM) algorithm.The diffusion coefficient was not obtained from the gradient magnitude of the image directly,but the edge information restored from the gradient magnitude based on the model of the image.Avoided the error caused by the gradient magnitude of the denominator.Existence,uniqueness and stability of the model's viscosity solution were provided.The experimental results are given and compared with the Chan-Wong's model.The improvement by the new model is visible.
摘要:In order to fix the problem of ill-posedness,some a priori information should be incorporated in the process of optical tomography reconstruction.In this paper,a Gibbs distribution with binary line process is introduced as a prior image model,which can result in a global smoothness with sharp edges.Because of the coexistence of the binary and continuous variables in the objective function,traditional optimization algorithms are not valid.Therefore,a coupled gradient neural network is proposed.In the process of optimization,the gradient computation of the energy function with respect to optical parameters is critical,for which,an algorithm based on the gradient tree is put forward.The reconstruction images corresponding to both the absorption and scattering coefficients proved that the proposed algorithm can be implemented effectively with high quality results by the introduction of the binary line process.
关键词:optical tomography;line process;Bayesian method;radiative transfer equation
ZHOU Shu, PAN Wei, LUO Bin, ZHANG Wei-li, DING Ying
Vol. 34, Issue 5, Pages: 897-901(2006)
摘要:This paper proposes a novel quantum genetic algorithm (QGA)——PSQGA,which uses particle swarm optimization method instead of quantum gate to update the state of quantum bit.It has the advantages of particle swarm optimization and quantum genetic algorithm.A variance function is introduced to estimate the stability of the algorithm.Though the experiments of four continuous functions and combination optimization problems,as well as its application to image sparse decomposition.Compared with the improved algorithm which involved quantum gate (IQGA),the ability of finding the best solution and the stability of PSQGA are greatly improved.PSQGA has better convergent property and ability of searching more extensive space.It is fit for the solution of complex optimization problems.
摘要:A realizable equivalent γ d n model for RLC interconnect driving point admittance is proposed in this paper.By introducing two formative RLC cells,the RLC interconnect which is modeled as a lossy transmission line can be accurately constructed and extended with these basic cells to meet the demand of given frequency and precision without the problems of instability or passivity.Simulation results show that this model can obtain a higher accuracy compared with the common used model open p .
关键词:transmission line;driving point admittance;RLC interconnect
摘要:A two dimensional dynamic cellular automata (CA) model is presented for simulation of photoresist etching process for the first time.In the dynamic model,only etching front cells are processed in relevant etching steps,so the dynamic model is stable and fast.The model has been successfully tested using some well-known etch-rate distribution test functions.
摘要:This paper presented a knowledge based genetic algorithm.This algorithm used the specialized genetic operator and adjusted adaptively the parameter.In addition,it incorporated the domain knowledge into its initial population.This method increases greatly the efficiency of the algorithm and overcomes the problem of premature convergence of the simple genetic algorithm.Experimental results show the feasibility and effectiveness of the algorithm in path planning for mobile robot.
摘要:This paper presents a simple and efficient approach to action scene detection in martial arts movies.Quick tempo is an important movie-level character for action scene.A shot pace function defined by shot length and MPEG-7 motion activity is used to measure the tempo.From the shot pace change curve,it is easy to locate the rough position of action scene.According to the character of action scene development,action scene boundary is detected at the shot level by analyzing the visual contents of shots within and around the above rough region.Two clues from the movie-level and shot-level make the method simple;working on compressed video directly makes the method very fast.Experimental results based on real-world movies verify its efficiency.
关键词:action scene;tempo;pace function;compressed video
摘要:How to describe and standardize the digital forensic requirement is an important part in digital forensic methodology.The paper propose the definition method of digital forensic requirement and an abstract model,named Environment Request Description (ERD).ERD definite three methods to describe digital forensic environment and requirement:Components,Environment Profile and Environment Target.ERD don't depend on a special digital forensic environment.Customer,developer and others depending on active usage could decide requirement.It will be useful to definite a forensic requirement in different and complex digital forensic environment.
摘要:Guided by the switch-signal theory,the operations of current transmission switch used to describe the action of threshold-controllable switch in multiple-valued Schmitt trigger were established,and novel current-mode CMOS ternary and quaternary Schmitt triggers with output current and voltage signal were presented.The hysteresis current values of multiple-valued Schmitt triggers were controlled by only sizing the transistor aspect ratios.The main properties of the circuit are its simple structure,fully adjustable hysteresis and suitable for low-voltage applications.The designed circuits are simulated using HSPICE program and the model parameters of a TSMC 0.25 μ m CMOS process,with 1.5V power supply.From the simulation results,the effectiveness of the proposed approach and the ideal characteristic of multiple-valued Schmitt circuit are validated.
关键词:switch signal theory;current-mode CMOS circuit;multiple-valued logic;Schmitt trigger
摘要:In the European DVB-T systems,the conventional synchronization algorithms used to adopting pilot and the guard interval.The mobile speed of the DVB-T receiver is within 100 km/h.The novel frame algorithm,which employs the superimposed Barker code with much lower power and the guard interval,is proposed for DVB-T receiver.Simulation shows that the algorithm can attain frame synchronization for DVB-T systems with mobile speed 150 km/h.Furthermore,the effect from the superimposed Barker code can be ignored.
摘要:We propose a image segmentation method which couples support vector machine and mumford-shah active contour model,where the advantages of supervised learning classification and the global region distribution information can be exploited to enhance the performance.A new region-based image energy term in curve evolution based on the output of support vector machine classifier is presented.It is more robust than classical active contour because it takes into account the image segmentation knowledge of human being and interactive operation as well.In order to improve the segmentation speed,support vector machine also be used to obtain initial contour firstly.Experimental results have demonstrated the flexibility and better performance of this new image segmentation method.
关键词:support vector machine (SVM);mumford-shah active contour model (ACM);image segmentation
摘要:The dependence of PHEMT's drain current and gain on temperature were analyzed.It was found that the current and gain decreased with temperature increase.Two methods were presented for automatically compensating the temperature dependence and their principals were analyzed.A series source resistor could keep the PHEMT's drain nearly unchangeable and decrease the temperature effect on the amplifier gain to some content.However,the automatic gate voltage temperature compensation was a strong compensation.It increased the gate voltages and PHEMT's transconductance with the temperature increase,thus greatly decreased the temperature effect on the gain.The two methods were then combined and applied to a broadband low noise amplifier.Good experimental results were obtained.When the temperature increased from -55℃ to +85℃ and from -55℃ to +125℃,the amplifier gain drop decreased by 60% and 51% at 6 GHz,respectively,which greatly improved the gain-temperature feature of the amplifier.
摘要:The recent studies show that the network traffic is self-similar.This nature has a great impact on flow control and queuing performance in networks,therefore it has received significant attention.By taking advantage of self-similarity of network traffic,we propose a new algorithm,named DMPIC-AQM (Dynamic Matrix PI Control -Active Queue Management).This algorithm has two significant features.On one hand,it makes use of linearity regression theory to predict the characters of future traffic.On the other hand,it is capable to control the active queue based on the principle of predictability PI controller in control theory.Thus,the algorithm can both restrict the oscillations and stabilize the length of queue in expectation.The simulation experimental results show that the integrated performance of the proposed algorithm is obviously superior to RED and PI algorithms,especially when network traffics change dramatically and the length of queue is short.
摘要:Bounds of undetected error probability (UEP) for binary constant weight codes (BCWCs) are studied in this paper.Firstly,by using the dual distance distribution and its properties for BCWC,we obtain a new lower bound of UEP for BCWC which improves the best known corresponding results by Fu-Kl ve-Wei.We point out that there are mistakes in results of Fu-Kl ve-Wei in 2003 which discussed the upper bounds,and a new upper bound of UEP for BCWC are obtained.
关键词:binary constant weight codes;undetected error probability (UEP);distance distribution;dual distance distribution
摘要:A new concept on the continuously distributed directional image/field(CDDF) and the method to compute it in the fingerprint images are proposed,which exhibits not only good continuity,well gradualness,and excellent robustness to the noises,but very high precision,as well.Then,the classical formula to compute the Poincaré Index and the algorithm for the singularity detection are improved,so that the modified version of Poincaré Index can present not only the rotation degrees,but also the rotation direction of the vector in the vector field,exactly.Therefore,it is able to locate the singularities (core points,and delta points)at pixel level with an accuracy of only one pixel.Based on these,a novel fingerprint classification algorithm based on both the continuously distributed directional image and the modified version of Poincaré Index is developed finally,which classifies input fingerprints into 5 categories:arch,tented arch,left loop,right loop,and whorl.The experimental results obtained on the fingerprint database of Jiangsu University of Science and Technology demonstrate that this algorithm is invariant to image rotation of any degrees,and successfully solves the problem of image rotation,translation,and transformation in fingerprint classification.For the 4,000 images in this database,a classification accuracy of 97.05% for the five-class problem has been achieved.So it has better classification performance than previously reported in the literature.
关键词:AFIS;fingerprint classification;continuously distributed directional image;singularity detection;Poincaré Index
摘要:A method of sensor-based path planning for mobile robot in unknown environment is presented.Two assumptions of the Bug algorithms are not used in the planning.Both the data error of the sensors and the physical size of the robot are taken into account.Only the necessary sensing data instead of the analytical expression of the obstacles are calculated in planning so as to save memory and improve the planning efficiency,thus the planning in real-time is guaranteed.And the proof of convergence is given.At last the simulation test is made to prove the feasibility.
ZHENG Ming-chun, YANG Shou-bao, YU Xiao-mei, SUN Wei-feng
Vol. 34, Issue 5, Pages: 957-961(2006)
摘要:In order to solve the problem that performance of TCP is reduced due to higher bit error rate and bandwidth asymmetry in wireless networks,we put forward a Bi-direction Flow Control mechanism (BFC).It controls traffics in two directions.In the forward channel,according an explicit ACK notification BFC adjusts data rate by cooperating multi-layer protocols.In the reverse one,we developed an ACK flow control algorithm that adjusts self-adaptively ACK rate according to congestion status of two channels.The results of the simulation experiments show that BFC has dramatically improved the performance of TCP over wireless/wired hybrid links.Particularly,in the networks with higher asymmetric factor,BFC has significant contribution.
摘要:The medical image registration method based on mutual information has been widely used recently.Despite of its advantages over other algorithms,it will introduce artifacts in calculating joint histogram with conventional partial volume (PV) interpolation.In this paper,an expand PV interpolation algorithm with B-spline is proposed to reduce the interpolation artifacts.A hybrid search strategy combined genetic algorithm with Powell's method is used to find the optimum value.Experimental results indicate the proposed algorithm can achieve better accuracy and robustness.
摘要:Architecture style is an important guideline for the architecture design.It constructs a common term space for the communication of the stakeholders and benefits the design reuse as well as code reuse.This paper attempts to provide a general formal framework for the style specification and make the variability of software architecture explicit.This framework can support modeling structure constraint,topology constraint and behavior constraint of the style.Through the introduction of the framework into ABC method,the paper offers a style-based architecture modeling method and supplies a graphical architecture modeling tool.