摘要:Resource constraint is a choke point of workflow system,and the efficient resource modeling and Verification methods are more important for workflow management system(WFMS).A new formal method used to model and analyze workflow system with resource constraint,resource workflow net,is presented based on Petri net and workflow techniques in this paper.Based on the definition of resource space(RS),resource subnet,resource workflow net(RWF-net),and soundness of the RWF-net,the resource scheduling views model formulated in WFMS are described and translated into Petri net,and a necessary and sufficient condition for the soundness of RWF-net is presented.The method integrate process model and resource model that can solve system analysis and verification problem for the process and resource modeled respectively and assure the consistency between process model and resource model. The usability of the research results in this paper is illustrated by an example.
摘要:For total variant image deblurring problem,it was proposed a total variant image deblurring algorithm based on split bregman method,which applied split bregman method to optimizing and solving the problem model.Firstly,taken advantage of auxiliary variable and quadratic penalty function,total variant image deblurring optimization problem was converted into a unconstraint optimization problem.Secondly,based on Bregman iterative,the problem was divided into two sub-problems and use the alternative minimization method to solving.Thirdly,according to the characteristic of subproblem structure,the Discrete Fourier Transform and Shrinkage technologies were used to implement the fast computation of sub-problems.The experimental results indicate that with different size blurry kernels,our algorithm can recover image effectively and steadily.Furthermore,compraring with FTVd and IRN deblurring methods,our algorithm can obtain better recovery results,and compute faster.
摘要:Image inpainting algorithms prefer to use the texture-based method to inpaint large scale missing regions,but the result is undesirable when the image contains structural information.Through introducing new priority value,mean value complement and penalty term in the traditional exemplar based inpainting method to improve the performance of it.After combining the linear and curve structural features,we propose an exemplar and structure based inpainting algorithm for large scale missing regions in this paper.The new algorithm is easy to implement and the inpainted experimental results compared with some inpainting algorithms are plausible.
关键词:image inpainting;texture;exemplar based method;structural information
摘要:Fuzzy relative entropy is capable of measuring the difference between two fuzzy sets.According to the indeterminacy of steganography communication,a fuzzy empirical matrix for n-th order Markov model is defined.New security measures in terms of fuzzy relative entropy and weighted fuzzy relative entropy are introduced for steganographic system.These new security measures are proved to be nonnegative,symmetric and uniform.Furthermore,some existing security measures under a deterministic data statistical distribution model can be derived from the proposed security measures.Simulation results show the new security measures have better evaluating ability than the existing deterministic security measures under the same modeling condition.In addition,the higher the order of the Markov model,the better the measuring ability of the proposed security measures.The proposed security measures may also provide more insights for designs of secure steganographic algorithms.
摘要:Indirect-branch prediction becomes a performance bottleneck for modern processors as object-oriented programs are more commonly used recently.This paper proposes a Hint-indexed Indirect-branch prediction (HIP) technique.It first uses the existing branch direction predictor to distinguish different indirect-branch occurrences,and treats the corresponding data as hints.Then,those hints are calculated with the indirect-branch addresses to generate the virtual addresses,which are used to access the BTB for the predicted indirect-branch targets.Our evaluation shows that HIP could achieve attractive performance improvement without large dedicated storages.Compared to that of a commonly-used BTB,it improves average performance by 20.38%.Compared with previously proposed hardware-based predictors,HIP improves performance by 8.66% over that of the VPC predictor,and it achieves the equivalent performance improvement provided by a 48KB Tagged Target Cache (TTC) predictor.The energy consumption is also reduced by 14.34% over the baseline.
摘要:Aiming at the segmentation-based method to estimate the spreading sequences of non-periodic long code(LPC) direct sequence spread spectrum(DS-SS) signal,this paper presents a Fast-ICA based method to estimate the spreading sequence of CDMA signals with optimize segmentation blindly,which on the base of the mathematical analysis of long-code CDMA signal.And then,the segmentation criterion different from current method,which as small as possible to cost the jumping probability,is given.The method is not only applicable for synchronous short-code single- or multi-user DSSS signals,but also adoptive to non-periodic long-code single- or multi-user DSSS signals.At the same time,an effective segmentation method is given in this paper,mathematic analysis and simulation results demonstrate the validity of these methods.
关键词:blind estimation of spreading sequence;blind source separation;non-periodic long code;multi-user
摘要:Based on the ordinary singular value factorization,a quantitative correspondence of the singular value and singular vectors between the universal extended matrix and its mother matrix is derived,and the perturbation analysis and image analysis for the universal extended matrix is provided accordingly.In addition,the formula of the singular value factorization and generalized inverse of universal extended matrix are given,which makes calculation easier and accurate.The results of Zou's SVD for extended Matrix are generalized.
摘要:The paper proposes a group of new distributions for interferometric magnitude and their corresponding parameter estimators.Firstly,under the frame of multiplicative model,based on the complex Wishart distribution,and bringing in the theory of clutter classification brought forward by Frery et al.,the new interferometric magnitude’s statistical models are deduced according to utilizing the reciprocal of a square root of Gamma and the square root of generalized inverse Gaussian law.The group of new models can precisely model the interferometric magnitude corresponding to different clutter areas with broad degrees of homogeneity,and they achieve the unification of the statistical distributions for single-channel SAR images and multi-channel SAR interferometric magnitude.Secondly,The new models' estimators are acquired based on the Mellin transform which can estimate the contained parameters quickly and accurately,guaranteeing the new distributions' fitting precision.According to the experiments performed on the real spaceborne SAR image,using the Kullback-Leibler (KL) distance,mean square error (MSE) and Kolmogorov-Smirnov (KS) test as similarity measurements,the experimental results show the effectiveness of the proposed models and estimators.
关键词:interferogram;statistical model;E-Gamma distribution;E-G0 distribution;E-G distribution
摘要:Power-saving is significant in wireless networks.To remedy the shortcoming in the power-saving class II of the IEEE 802.16e standard that energy is unnecessarily expended by the idle mobile stations (MSs) that are prevented from sleeping by the fixed length of the listening window,a two-stage reliable multicasting scheme (TRMS) is proposed,which lets a base station (BS) multicast during the first stage and remulticast,in the second stage,the network coded packets that contains the lost packets in the first stage.The TRMS enables MSs to sleep as soon as they become idle and realizes reliable packet delivery under delay constraint.Simulation shows that TRMS achieves power saving and outperforms the traditional retransmission and acknowledgement mechanism in terms of duty cycle,energy consumption,throughput,and packet loss rate.
摘要:A new method to identify active decoys from ballistic targets in the course of ballistic missile penetration is proposed based on the time invariant of orbital elements,while the effects of radar placement upon the identification are analyzed as well.And a novel approach named Inverse Track Filter (ITF) based on the principle of kinetic track reciprocity of two bodies is put forward to decrease the initial error and shorten the convergence time.Simulations are presented to verify the effectiveness and feasibility of the proposed methods.
摘要:Illumination synchronization is the key technology for distributed interferometric SAR systems.A novel illumination synchronization method based on the maximum coherence is proposed.Then,the Doppler centroid frequency and the image SNR loss of bistatic mode SAR are analyzed theoretically.The advantages of the method compared with the traditional illumination synchronization method are presented by analyzing the effect of the interferometric image pair coherence、the number of independent looks for interferometric phase estimation and the error of interferometric coregistration time.A simulation which contains full process of InSAS system is carried out,The simulation results indicate that this method can improve the interferometric performance of the distributed InSAR system effectively.
摘要:Context awaress is one of the essential characteristics of the ubiquitous computing environment.Through the division of context space,this paper proposes a hierarchical distributed context awareness model which describes the layered abstraction from sensor data to context information.On the basis of the hierarchical model,this paper gives the detailed specification of the diversified roles of a context awareness system and the relationship between those roles.At the end of this paper,the simulation experiment is built to compare the proposed model with typical Context Toolkit model.The experiment shows that the proposed model has an advantage of lower runtime payload over the Context Toolkit model.Through the analysis on the activeness of the model,it is proved that the aware activeness is proportional to receiving data payload and inversely proportional to sending data payload.The hierarchical model which will play an important role in the configuration of a distributed context awareness system.
摘要:A hybrid construction is a formal method of combining asymmetric encryption and symmetric encryption.By introducing key encapsulation mechanism into certificate-based encryption scheme,we propose a formal definition and security model of the certificate-based key encapsulation mechanism.Furthermore,we present a construction of the certificate-based key encapsulation scheme.Based on decision bilinear Diffie-Hellman difficulty assumption,the proposed scheme is provably secure against indistinguishability under adaptive chosen-ciphertext attack in the standard model.
摘要:Based on the microstructure and the domain reversal behavior of ferroelectric thin films,the ferroelectric domains are abstracted as a cluster of dipoles.Each dipole has self-independent coercive field,and the complicated mechanism of their polarization switching under random signal is considerably facilitated by a geometric interpretation.Then,we introduce the concept of polarization switching function to represent the relation between the macroscopical polarization and electric field.The polarization switching function also help to simply the computation and avoid the enhanced errors from the differentiations of experimental data.A model based on polarization reversal function is developed,which has explicit physical background and concise mathematical formalism.
关键词:ferroelectric film capacitors;model;dipole;hysteresis
摘要:To the qualitative problem of capability idle,request overload and function invalidation of the semantic web service,we proposed a method to model the relationships between multi-attributes for evaluating the quality of the semantic web service.We chose a set of evaluation indexes which are both on the aspects of technology and society,and defined their attributes’ computing models,made the analysis on potential relationships between the attributes and proved consistency,and utilized an existing swatch to do experiment for the capability analysis and effect validation of this proposed method.
关键词:semantic web service;quality evaluation of service;multi-attribute relationships
摘要:To solve the problem of multi-target tracking model with the time-varying number of targets,a novel Gaussian mixture PHD filter is proposed for the nonlinear Gaussian system.A third-degree Spherical-Radial rule is applied to calculate the prediction and update distributions of target states for nonlinear multi-target models.The pruning method is optimized by using a fuzzy threshold to avoid the exponential increasing of the Gaussian components.The measurements are used to generate the density of new targets that appear randomly anywhere in the observation space.The performance of the four nonlinear Gaussian Mixture PHD filters is compared.The simulation results demonstrated the efficiency of the proposed algorithm.
摘要:A key challenge of complicated decision problem understanding is to make the information of decision problem,i.e.domain,goal,status and structure,readable and understandable for decision support system.This paper introduces ontology to describe semantics,and then problem semantic relationship and semantic computation are presented.Based on the rule of keeping the understanding results unchanged,semantic iteration method is addressed in order to transform the decision problem semantics,which cannot be analyzed through ontology,into fully understandable problem semantic.Further,semantic refining method is presented so that each complicated decision problem space can be refined into a most optimized closed problem space with a minimum complexity,which can be calculated from two facets:structure complexity and content complexity of problem space.Experiments show that the method of complicated decision problem semantic analysis is effective and feasible.
关键词:ontology;semantics understanding;semantics iteration;semantics refining;most optimized closed problem space
摘要:The chaotic performance of small-time scale network traffic was covered by noise,which made the traffic unpredictable.This paper introduces the local projection to denosie network traffic;a chaotic and predictable traffic trend is obtained.As the network traffic series is long-period and time-varying,a new method named optimal training subset online fuzzy least squares support vector machines (OTSOF-LSSVM) is proposed.Samples temporal and distance nearest to prediction sample are chosen as optimal training subset,and the subset are fuzzified.On this basis,the prediction model is established by fuzzy LSSVM.The model update computational complexity is reduced by partitioned matrix calculation.The noise reduction and trend prediction on network traffic shows the proposed method can predict the trend quickly and exactly.
关键词:network traffic;trend prediction;chaotic theory;optimal training subset;least squares support vector machine (LSSVM)
摘要:This paper presents a novel Regex matching algorithm with Smart Finite Automaton (SFA),where branching transitions of the XFA are augmented with adding extra check instruments,so that back-off transitions between states are eliminated,avoiding unnecessary state transitions.Experimental results show that compared with the XFA,the SFA significantly improves the time/space efficiency,separately reducing 44.1% and 69.1% in terms of the memory consumption and memory accesses of state transitions.
摘要:The paper presents a lexical analysis algorithm based on combining senses of words for computing subsumption relations between entities in ontology matching.It firstly finds the suitable sense of each word and extends it;then formally defines the representation of an entity notion based on semantic elements;finally,infers subsumption relations between all possible pair of entities,one from each of two ontologies.The experiments,over four real in use ontologies,show that the algorithm helps to increase the recall of the system.
摘要:A method that improving the resolution of PWM(Pulse-Width Modulation) mode DAC(Digital-To-Analog Converter) by interpolator was proposed for the problem that PWM mode DAC in embedded applications and how to choose DAC chip.The PWM signal spectrum was analyzed.The design principle and method of analog filter was proposed.The design of High-Resolution DAC using MCU’s(Micro Control Unit) PWM proved that this method is stable and reliable by analysis of experiments.
摘要:Microbial Fuel Cells (MFCs) are considered to be one of the most promising alternative renewable power sources.But MFC cannot drive almost all the present commercial electron devices because of its low output voltage and current.Power Management System (PMS) is designed to accumulate power from MFCs and then discharge it to drive the load intermittently.Super-capacitor is usually selected as the power storage device in PMS.As MFC provides limited power output,maximizing the power transferred to the super-capacitor is of great importance.A theoretical method of determining the optimal capacitance that can maximize the average power generation in a single charging cycle was presented based on the generally accepted equivalent circuit model of MFC.A single chamber air-cathode MFC experiment system was built and experiments were implemented.Simulation and experimental results show the validity of the proposed method.
摘要:As is well known,the limited ability of data acquisition and the high-resolution reconstruction (required by many applications) is an inherent contradiction in modern signal processing.How to reconstruct the signal with high-resolution from a limited number of data is one unsettled problem in the research areas of signal processing,communications,applied mathematics etc.By reviewing the achievements of data acquisition approaches by incomplete sampling and high-resolution reconstruction,an optimization framework for signal reconstruction based on degree-of-freedom of information is abstracted.There are three key aspects in this framework,the degree of freedom of the signal information directly influences the sampling rate,the constraints are determined by the sampling method and the object function is decided by the feature of signal.This paper focuses on the effectiveness of the means used to analyze these data acquisition approaches and the discussion of their advantages and disadvantages.In the end,we propose the research prospect of data acquisition by the incomplete sampling and high-resolution reconstruction and show our new research and result.
关键词:degree of freedom of information;incomplete sampling;high-resolution reconstruction by optimization;compressed sensing;low-rank matrix completion
摘要:Two types of low-power asynchronous comparators named asynchronous ripple comparator and pre-stop asynchronous comparator are proposed based on the statistical characteristic of input data in the paper.The asynchronous ripple comparator stops computing at the first unequal bit,but it has to deliver the result to the LSB.The pre-stop asynchronous comparator is proposed by revising the truth table based on the new 2-bit comparison unit and stop judgment circuit.It can stop comparing at the first unequal bit and obtain the result immediately.The proposed and contrastive comparators (BCL comparator and clock-gating comparator) are implemented with SMIC 0.18 μm process.Simulation results show that the proposed pre-stop asynchronous comparator features the lowest power.It saves 87.1%,84.5% and 37.5%,28.6% power respectively compared to the synchronous BCL comparator and clock-gating comparator with random data and data from LDPC decoder.
关键词:low-power;statistical characteristics of input data;comparator;LDPC (low density parity check) decoder
摘要:Owing to digital television image diversification,there is no better enhancement method to enhance the image.An improved video quality enhancement algorithm is presented and the hardware circuit is designed in this paper.Unlike traditional histogram equalization approaches,the input images are divided into different classes by analyzing the information of input images and the contrast is adjusted in YUV color space firstly.Then the adjusted images would be processed as follows:the dynamic range is adjusted in RGB color space,the luminance is amended and the saturation is compensated in HSV color space.The VLSI(Very Large Scale Integrated circuit) hareware circuit of the algorithm is designed with Verilog,and it is verified based on the FPGA(Field-Programmable Gate Array) video verification platform.The experimental results show that the presented algorithm can adapt the pictures of different scenes and the enhanced images are more clear and vivid.
关键词:color space;video quality enhancement;dynamic range adjusting
摘要:To address the instability problems of level set and region growth,for example,they are sensitive to noises and initial boundaries as well as they will easily leak from the weak boundaries,an improved image segmentation method based on level set is proposed.Our model consists of an external energy term that involves the image gray-scale statistical information and gradient information.And we use region growth method to solve the problem that level set method is sensitive to initial boundaries.we contrast our improved method with region growth method,threshold method,GAC model,C-V model,Snake model to segment livers from abdominal CT images.The experiment results show that our method can not only be efficient for image segmentation,but also greatly improve the quality of segmentation.
摘要:In the study of IC design for manufacturing,models on the relationship of yield and reliability deserve much attention.The impact of defects on the yield and reliability is associated not only with the particle size of it but also with the location of it on the chip.In this paper,it is analyzed that the defects at the same size in interconnect different locations affect the effective width of interconnect wires,by discussing the impact of the loss of metal in the interconnect wire on the interconnect wire itself.The average effective width of interconnect wires is given based on the uniform distribution of defects.In addition,the model on the relationship of the IC manufacturing yield and reliability is presented based on the location of defects,by referring on the existing model of estimating yield and reliability.If the processing line is stable,the rate of the product failure can be estimated effectively by the expression and the yield of the processing line,which can shorten the development period of the new products.
关键词:yield;reliability;defect;the size distribution of defect
摘要:Based on the thought of Extended Kalman Filter (EKF),the nonlinear filter fusing pose information was proposed by the analysis of relationship between pose and motion characteristic in 2D and 3D,with pose and velocity modulus as its control,the position of target was estimated through state and measurement equations,pose information and position information are fused effectively.The simulation results,which under the condition that pose with different noise,validate the accuracy and real time of the algorithm.
摘要:This paper proposes a pico-second precision programmable mixed-signal CMOS square-wave delayer which achieves the function of linearity,original input waveform duplication,variable resolution and low power consumption.And it can be integrated in the small size chip under low cost.The principle of the system is analyzed.The system is modeled and simulated with Tanner EDA tools.Simulation results validate that a highest resolution of 20ps/LSB and maximum delay range of 28ns can be achieved using 0.6μm mixed-signal CMOS fabrication process,an 8-bit digital delay control signal and a 2-bit digital range selection signal.
摘要:Software pipelining exploits instruction level parallelism by reconstructing loops,while modulo scheduling is a kind of widely used scheduling algorithms for software pipelining.Traditional modulo scheduling algorithms usually cause overlapping register lifetimes and increase register pressure,and thus are not applicable to embedded processors.This paper presents the NOn-Overlapped Iterative (NOOI) modulo scheduling algorithm based on an optimized backtracking model for embedded processors.NOOI algorithm avoids the register lifetime overlap by adding loop-carried anti-dependence and resolves the scheduling conflicts using dependence-constrained and resource-constrained backtracking model.Our evaluations show that NOOI can significantly improve the success ratio and loop initial interval,which leads to better program performance.
关键词:embedded processor;software pipelining;modulo scheduling;backtracking model
摘要:Most of the e-business recommender systems are based upon collaborative filtering (CF) algorithms.Since such systems have been shown to be vulnerable to shilling attacks in which malicious user profiles are inserted into the system in order to push or nuke the predictions of some targeted items,shilling attack detection has recently become a hot research topic in recommender systems.Firstly,the effectiveness of five types of attacks against different CF algorithms is analyzed.Secondly,a feature selection algorithm is presented.Two kinds of shilling attack detection algorithms based on supervised learning are then proposed:the first one is based on naÏve Bayesian classifier,and the second one is based on k nearest neighbor (kNN) classifier.At last,experimental results show the effectiveness of the feature selection algorithm and the sensitivity and specificity of these two kinds of detection algorithms.
摘要:In order to improve the convergence speed and the environmental adaptability of the path planning algorithm,a robot path planning algorithm based on multi-artificial fish-swarm is proposed.We present also a dynamic obstacle avoidance algorithm based on the rule-base of collision avoidance in dynamic environment to avoid collisions with the moving obstacles.In our approach,the distance between a fish and a goal is defined as food concentration and the distance between two neighbor grids is defined as step length.The preying behavior of fishes is regard as default behavior and perform clusters act or rear-end act is activated in some certain condition.Then the optimal path in static environment is planned by the search mechanism of bi-directional fish-swarms.After that,the effective collision avoidance behavior is obtained,from the obtained dynamic obstacle information through sensors.Many simulation experiments have shown that the algorithm has a fast convergence speed and strong search capability.Even in complex environments which have static and dynamic obstacles,it can avoid collision safely and plan an optimal path rapidly at the same time.
关键词:dynamic environments;robot path planning;multi-artificial fish-swarm;the rule-base of collision avoidance
摘要:In this paper,a method for adjusting parameters of Time-Difference Fluxgate system is proposed,where feedback module is applied.Based on the observation of time-difference output of simulated Langevin equation,research is conducted on the influence of both linear feedback and cubic feedback on system parameters and stochastic resonance effect,and conclusions that linear feedback lowers coercivity,and cubic feedback adjusts optimal stochastic resonance noise are drawn.Numerical simulations of the models with MATLAB shows that addition of feedback module makes it more convenient for the Time-Difference Fluxgate system to adjust its internal parameters and control the stochastic resonance effect,which improves the performance of the system considerably.Compared with no feedback module,the sensitivity is enhanced by 50%.