Advances in Intelligent Computing

In this paper, we proposed a novel method for protein secondary structure prediction using sequence profile and conserved domain profile. Sequence profile generated from PSI-BLAST (position specific iterated BLAST) has been widely used in protein secondary structure prediction, because PSI-BLAST shows good performance in finding remote homology. Conserved domains kept functional and structural information of related proteins; therefore we could draw remote homology information in conserved domains using RPS-BLAST (reverse position specific BLAST). We combined sequence profile and conserved domain profile to get more remote homology information, and propose a method which used the combined profile to predict the protein secondary structures. In order to verify the effectiveness of our proposed method, we implemented a protein secondary structure prediction system. Overall prediction accuracy reached 75.9% on the RS126 data set. The improvement by incorporating conserved domain information exceeded 3%, and this result showed that our proposed method could improve significantly the accuracy of protein secondary structure prediction.

Genome-wide differential expression studies of human diseases using microarray technology usually produce long lists of genes with altered expression, therefore, the genes causally involved in a disease cannot be effectively separated from innocent bystanders. Existing methods for differential analysis of gene expression profiles seem unable to solve this problem successfully. In this paper, we present a systematic strategy that combines gene-wise and function-wise differential analysis of gene expression profiles to interrelate genes and functions with human diseases. The gene-wise analysis adopts a modified T-test to analyze the expression alteration of each single gene, and the alteration is represented by quantitative significant

p

-value. The function-wise analysis uses a new combined S-test to identify coordinate alterations of genes within each functional category. A computational tool, MageKey, is developed based on this strategy, and its utility is demonstrated by the analysis results of gene expression dataset of human Amyotrophic Lateral Sclerosis disease. MageKey is freely available upon request to authors.

Extended Continuous Event Graphs (ECEG) are a special class of Continuous Petri Nets. As the limiting form of Extended Timed Event Graphs (ETEG), ECEGs can be used not only to describe the discrete events approximately, but also to describe the continuous processes. In this note, we obtain some of the global properties of ECEGs. In the end, a simple example is given to illustrate the feedback control of CEGs with input.

Here we develop a new approach of combined literature mining and gene expression analysis (CLMGE) to model the complex neuro-endocrine-immune (NEI) interactions. By using NEI related PubMed abstracts and the Human Genome Organisation gene glossary for subject oriented literature mining (SOLM), it is found that the NEI model serves as a scale-free network and the degree of nodes follows a power-law distribution. Then we evaluate and eliminate the redundant of SOLM-based NEI model by multivariate statistic analysis basing on selected gene expression data. Each involving expression data is tested by cross validation with Leave One Out strategy. The results suggest that the performance of CLMGE approach is much better than that of SOLM alone. The integrated strategy of CLMGE can not only eliminate false positive relations obtained by SOLM, but also form a suitable solution space for analyzing gene expression data. The reasonable biological meanings of the CLMGE-based NEI model are also evaluated and demonstrated by classifying its sub-functions according to DAVID and SwissProt databases.

Image segmentation is supposed to be the most important step in microarray image analysis. In this work, we proposed a new template-based segmentation method for DNA microarray images. Different from the local-based segmentation techniques adopted by all the available analysis softwares, our algorithm segments images from global view of point. Based on mean shift filtering technique, we first segmented image into some different homogenous regions in which all the spots appeared as different local maximum regions. Then an initial spot segmentation template was extracted by morphological H- reconstruction. Finally, a refined spot segmentation template was obtained by histogram analysis. Experimental results showed that our algorithm is robust and can obtain accurate spot segmentation results. Especially, compared to all the available algorithms, our template-based spot segmentation scheme not only can facilitate downstream intensity extraction step but also can be very helpful to improve the accuracy of intensity extraction.

A comparative analysis of olignucleotide frequencies in two sets of introns of genes highly-transcribed and lowly-transcribed respectively has suggested that the existence of potential positive regulatory motifs of transcription in yeast introns. To further reveal the distribution feature of these motifs, we detected significant clusters of these motifs (mainly pentanucleotides) in introns by r-scan analysis. The results showed that there are more clusters of regulatory motifs in the introns of ribosomal protein genes (highly-transcribed genes) than in lowly-transcribed introns. Experimental studies show that the transcription factors function cooperatively in transcriptional activation and the corresponding binding sites for factors generally cluster in DNA. Accordingly, we speculated that the cluster distribution of regulatory motifs of transcription in yeast intron

s

could becorrelated with the cooperative action of transcription factors and the transcriptional rates of genes could be improved by the cooperativity.

This paper investigates the effects on neural classification performance of biological data by features selection. Where the Relief-F and Symmetrical Tau feature selection algorithms were employed on a set of high level features of DNA and structural profiles. It was observed that even with a small percentage of the features used in neural classifiers, the recognition rate of E.coli promoters was not degraded significantly.

DNA automaton is a simple molecular-scale automaton, in which the converting of information deploys in molecule-scale by DNA and DNA-manipulating enzymes autonomously. Finite automaton with two internal states has been applied to medical diagnosis. This paper analyses the computation ability of DNA automaton with different enzymes and the possibility of DNA finite automaton with three internal states which is more powerful than the two internal states finite automaton. Finally, we describe a DNA finite automaton with three internal states and proposal a scheme of DNA automaton model in which the tape head can move forward and backward, and symbols can be read from and write into the tape, thus extend the computation ability of DNA automaton and its application fields.

With the birth of DNA computing, Paun et al. proposed an elegant algorithm to this problem based on the sticky model proposed by Roweis. However, the drawback of this algorithm is that the “exponential curse” is hard to overcome, and therefore its application to large instance is limited. In this s paper, we present a DNA based evolutionary algorithm to solve this problem, which takes advantage of both the massive parallelism and the evolution strategy by traditional EAs. The fitness of individuals is defined as the negative value of their length. Both the crossover and mutation can be implemented in a reshuffle process respectively. We also present a short discussion about population size, mutation probability, crossover probability, and genetic operations over multiple points. In the end, we also present some problems needed to be further considered in the future.

An DNA computing model of solving the graph isomorphism problem with 3-D DNA structures is proposed in this paper. The k-armed branched junction molecules are used to encode k-degree vertices. Double stranded molecules are used to encode edges. These molecules are to be mixed in a test tube to be ligated. The reaction product can be detected by gel electrophoresis. The time complexity of the algorithm is

o

(

n

2

) , where

n

is the number of vertices of the graph.

This paper presents an implementation of Croitoru’s genetic algorithm for graph coloring problem, and some necessary modification and simplifying are made by using DNA operations. In this algorithm, each vertex and edge is encoded with a series of encodings incorporating position information, and the initial diverse candidate population is generated using POA. One crossover operator, two mutation operators, evaluation and selection operators are all implemented using basic operations on DNA. It is shown that the algorithm can be implemented with space complexity much decreased and time complexity O(mn

2

) to get a new generation, where n is the number of vertices and m is the number of edges. Moreover, borrowing ideas from the above implementation, an algorithm for Maximal Clique problem is also presented.

For the minimum initial marking (MIM) problem is one of minimum resource allocation problems, it is significant to study the MIM problem for a class of hybrid timed Petri nets, called a hybrid timed event graph (HTEG). An HTEG has additional continuous places and continuous transitions than a timed event graph (TEG). By the construction of a new dioid endowed with the pointwise minimum as addition and the composition of functions as multiplication, a linear min-plus algebraic model of HTEG was derived. Based on the min-plus algebra and its properties, the MIM problem for HTEG was studied in the text.

A voting-mechanism-based fuzzy neural network system is proposed in this paper. When constructing the network structure, a generalized class cover problem is presented and its two solving algorithm, an improved greedy algorithm and a binary particle swarm optimization algorithm, are proposed to get the class covers with relatively even radii, which are used to partition fuzzy input space and extract fewer robust fuzzy IF-THEN rules. Meanwhile, a weighted Mamdani inference mechanism is adopted to improve the efficiency of the system output and a real-valued particle swarm optimization-based algorithm is used to refine the system parameters. Experimental results show that the system is feasible and effective.

We present a simulation-based algorithm to compute the average reward of singulary perturbed Markov Reward Processes (SPMRPs) with large scale state spaces, which depend on some sets of parameters. Compared with the original algorithm applied on these problems of general Markov Reward Processes (MRPs), our algorithm aims to obtain a faster pace in singularly perturbed cases. This algorithm relies on the special structure of singularly perturbed Markov processes, evolves along a single sample path, and hence can be applied on-line.

In this paper, an application of hybrid intelligent control algorithm to semi-active Control of the MR Damper is presented for engineering structures. The control signal is optimized directly by the

μ

GA approach to obtain the numerical relation between the control signal and the system output. And then, this relation is stored in the weight value of a trained artificial neural network, which can be available for another structure subjected to other seismic inputs. The results of numerical example indicate that the semi-active control of MR damper based on the hybrid algorithm can efficiently reduce the structural responses induced by earthquake.

The chaotic phenomenon in the brushless DC motor is revisited in this paper. For a specific real physical plant (i.e., the brushless DC motor), the main drawback of some existing chaos control methods is first analyzed. Then, a piecewise quadratic state feedback method is proposed for controlling chaos in the brushless DC motor model. In the proposed method, the direct-axis (or quadrature-axis stator voltage) is used as the control variable, and the piecewise quadratic state feedback is used as the control law. The control mechanism is illustrated and then the principle of parameters selection is discussed. Chaos in the brushless DC motor model can be satisfactorily eliminated by the proposed piecewise quadratic state feedback, and the speed of the motor can be stabilized to a constant value. This control method is simple and can be easily implemented. Simulation results verify the effectiveness of the method.

Support vector machine is a new and promising technique for pattern classification and regression estimation. The training of support vector machine is characterized by a convex optimization problem, which involves the determination of a few additional tuning parameters. Moreover, the model complexity follows from that of this convex optimization problem. In this paper we introduce the support vector machine adaptive control by Lyapunov function derivative estimation. The support vector machine is trained by particle filter. The support vector machine is applied to estimate the Lyapunov function derivative for affine nonlinear system, whose nonlinearities are assumed to be unknown. In order to demonstrate the availability of this new method of Lyapunov function derivative estimation, we give a simple example in the form of affine nonlinear system. The result of simulation demonstrates that the sequential training algorithm of support vector machine is effective and support vector machine adaptive control can achieve a satisfactory performance.

Tomita devised a method of generalized LR(GLR) parsing to parse ambiguous grammars efficiently. A GLR parser uses linear-time LR parsing techniques as long as possible, falling back on more expensive general techniques when necessary. In this paper, the motivation of adopting the GLR parsing algorithm to construct parsers for programming languages is presented. We create a multi-level scheme to fasten the GLR parser. We introduce runtime control mechanisms to the GLR parser to invoke semantic actions attached to grammar rules. The algorithm has been implemented in Development Expert Tools (DET), a compiler which is designed by Institute of Intelligent Machines, Chinese Academy of Sciences, at Hefei. Experiments show that the speed of our GLR parser is comparable to LALR(1) parsers when parsing deterministic programming languages.

Based the character of Modular Self-Reconfigurable (MSR) robots, a homogeneous lattice robot called M-Cubes which owe to the property and structure of agent was build, feature vector matrix of modules describe completely the connection relation among modules. Motion mode and action of modules were analyzed. Emergent control is the most suited system control of MSR by comparing control way. Due to MSR robot is similar with cellular automata, emergent control model based CA was proposed. A two layers NN which has 7 inputs and single output simulate the nonlinear rules function, the feature vector of module is input of CA’s rules, action of module is output of CA’s rules. The simulation results show that emergent control based CA is great significance to enhance robustness and scale extensibility of MSR.

This paper points out some drawbacks and proposes some modifications to the conventional layer-by-layer BP algorithm. In particular, we present a new perspective to the learning rate, which is to use a heuristic rule to define the learning rate so as to update the weights. Meanwhile, to pull the algorithm out of saturation area and prevent it from converging to a local minimum, a momentum term is introduced to the former algorithm. And finally the effectiveness and efficiency of the proposed method are demonstrated by two benchmark examples.

This paper describes the architecture of a computer system conceived as an intelligent assistant for public transport management. The goal of the system is to help operators of a control center in making strategic decisions about how to solve problems of a fleet of buses in an urban network. The system uses artificial intelligence techniques to simulate the decision processes. In particular, a complex knowledge model has been designed by using advanced knowledge engineering methods that integrates three main tasks: diagnosis, prediction and planning. Finally, the paper describes two particular applications developed following this architecture for the cities of Torino (Italy) and Vitoria (Spain).

This paper proposes a general formalism for evaluating hybrid Bayesian networks. The formalism approximates a hybrid Bayesian network into the form, called fuzzy partial least-squares Bayesian network (FPBN). The form replaces each continuous variable whose descendants include discrete variables by a partner discrete variable and adding a directed link from that partner discrete variable to the continuous one. The partner discrete variable is acquired by the discretization of the original continuous variable with a fuzzification algorithm based on the structure adaptive-tuning neural network model. In addition, the dependence between the partner discrete variable and the original continuous variable is approximated by fuzzy sets, and the dependence between a continuous variable and its continuous and discrete parents is approximated by a conditional Gaussian regression (CGR) distribution in which partial least-squares (PLS) is proposed as an alternative method for computing the vector of regression parameter. The experimental results are included to demonstrate the performances of the new approach.

Multi-agent systems (MAS) are studied from the point of view of software architecture. As the existing architecture description languages (ADLs) are difficult to describe the semantics of MAS, a novel architecture description language for MAS (

π

-net ADL) rooted in BDI model is proposed, which adopts

π

-calculus and Object-Oriented Petri nets presented in this paper as a formal basis.

π

-net ADL stresses the description of dynamic MAS architecture, and it is brought directly into the design phase and served as the high-level design for MAS implementation.

π

-net ADL can visually and intuitively depict a formal framework from the agent level and society level, and analyze, simulate and validate MAS and interactions among agents. Finally, to illustrate the favorable representation and analysis capability of

π

-net ADL, an example of multi-agent systems in electronic commerce is provided.

As the information in the internet proliferates, the methods for effectively providing the information have been exploited, especially in conversational agents. Bayesian network is applied to infer the intention of user’s query. Since the construction of Bayesian network requires large efforts and much time, an automatic method for it might be useful for applying conversational agents to several applications. In order to improve the scalability of the agent, in this paper, we propose a method of automatically generating Bayesian networks from scripts composing knowledge base of the conversational agent. It constructs the structure of hierarchically composing nodes and learns the conditional probability distribution table using Noisy-OR gate. The experimental results with subjects confirm the usefulness of the proposed method.

In this paper we reveal that the environmental noise will not only suppress a potential population explosion in the stochastic Lotka-Volterra competitive systems with variable delay, but also make the solutions to be stochastically ultimately bounded. To reveal these interesting facts, we stochastically perturb the Lotka-Volterra competitive systems with variable delay ẋ(

t

) = 

diag

(

x

1

(

t

),...,

x

n

(

t

))[

b

 + 

Ax

(

t

 − 

δ

(

t

))] into the Itô form

dx

(

t

) = 

diag

(

x

1

(

t

),...,

x

n

(

t

))[

b

 + 

Ax

(

t

 − 

δ

(

t

))]

dt

 + 

σx

(

t

)

dw

(

t

) and show that although the solution to the original delay systems may explode to infinity in a finite time, with probability one that of the associated stochastic delay systems do not. We also show that the stochastic systems will be stochastically ultimately bounded without any additional conditions on the matrix

A

.

This paper proposes a new neural network approach specifically designed for solving two dimensional binary image recognition problems. The proposed neural network is an extension of the Hausdorff ARTMAP introduced by Thammano and Rungruang [1]. The objectives of this research are to improve the accuracy and correct the drawbacks of the original network. The performance of this proposed model has been compared with that of the original Hausdorff ARTMAP. The experimental results on two benchmark databases, the ORL and Yale face databases, show that the proposed network surpasses the original Hausdorff ARTMAP in both performance and processing time.

For solving the dynamic condition problem of vendor selection, the analytic hierarchy process method is modified to a dynamic approach in the period of analytic cycle. The balanced scorecard is used to define the 4 major frameworks of supplier selection including customers, financial, internal business processes, and innovation and learning. The 16 attributes are extended from major frameworks. The main character of proposed method is the scores of attributes and alternatives from the estimation of commander’s trade-off can be changed in time axis under the changeable and conjecturable business environments. In case study, the advantage and limit of the model are illustrated.

In this paper, we demonstrate an example of using artificial intelligent in solving problems with complex and uncertain features in communication networks. The concept of Fuzzy Expert System is used in the design of an Active Queue Management (AQM) algorithm. Expert System and Fuzzy Logic are commonly used methods in solving various kinds of uncertain problems. Network congestion control is a problem with large scale and complexity, where no accurate and reliable model has been proposed so far. We believe Fuzzy Expert System methods have the potential to be applied to congestion control and solve those problems with uncertainties. This research demonstrates the possibility of using Fuzzy Expert System in the network congestion control. In this paper, a fuzzy-expert-system-based structure is proposed for network congestion control and a novel AQM algorithm is introduced. Simulation experiments are designed to show that the fuzzy-expert-system-based AQM algorithm exhibits a better performance than conventional approaches.

The mathematical model of underground water flow is introduced as basis to identify the permeability coefficients of rock foundation by observing the water heads of the underground water flow. The artificial neural network is applied to estimate the permeability coefficients. The weights of neural network are trained by using BFGS optimization algorithm and the Levenberg-Marquardt approximation which have a fast convergent ability. The parameter identification results illustrate that the proposed neural network has not only higher computing efficiency but also better identification accuracy. According to identified permeability coefficients of the rock foundation, the seepage field of gravity dam and its rock foundation is computed by using finite element method. The numerically computational results with finite element method show that the forecasted water heads at observing points according to identified parameters can precisely agree with the observed water heads.

An intelligent predictive control with locally linear neurofuzzy identifier and numerically optimization procedure has been proposed for temperature control of electrically heated micro heat exchanger. To this end, first the dynamics of the micro heat exchanger is identified using Locally Linear Model Tree (LOLIMOT) algorithm. Then, the predictive control strategy based on the LOLIMOT model of the plant is applied to provide set point tracking of the output of the plant. Some computer simulation is provided to show the effectiveness of the proposed controller. Simulation results show better performance for the proposed controller in comparison with PID controller.

The flood of data is occurred in ITS(intelligent transportation system) according to the progress of wireless telecommunication and sensor network. To deal with large data and provide suitable services smoothly, it is necessary for an index technique to store and search bulk object data rapidly. However the existing indices require a lot of costs to insert a huge amount of data because they store every position data into the index directly. To solve this problem in this paper, we propose a buffer node operation and design a GU-tree(Group Update tree). The proposed buffer node manner reduces the input cost effectively since it stores the moving object location in a group. And then we confirm the effect of the buffer node manner which reduces the insert cost and increases the search performance in a time slice query from the experiment to compare the operation with some existing indices.

In pervasive computing community, there is a high interest on context-aware computing. Much work focuses on context reasoning, which deals with high-level abstraction and inference of pervasive contextual information and several prototype systems have been proposed. However, overwhelming contextual information in pervasive computing makes these systems inefficient, even useless. In this paper, we propose an application-oriented context filtering system (ACMR) to deal with above problem. To prevent the pervasive applications from being distracted by trashy contexts, ACMR system only deals with application-related contextual information rather than all the available contextual information. Experiments about ACMR system demonstrate its higher performance than those of previous systems.

With the development of e-manufacturing, the flexible intelligent maintenance is necessary. Wireless sensor network is competent for a flexible maintenance system. Client/Server model adopted in traditional intelligent maintenance which requires transmitting vast data via network is unsuitable in wireless sensor network with limited bandwidth and unstable connection. Emerging mobile agent technology can reduce the network traffic and overcome the network latency, and is an eligible substitute for Client/Server model. This paper presents a mobile agent based wireless sensor network for intelligent maintenance. The flexible maintenance system integrates soft-computing, wireless sensor network and mobile agent technology. The system is used to maintain Mori Seiki Co. SV500 numerical control machining center and the results illustrate the effectiveness and efficiency of the system.

A scheme for integrated guidance/autopilot design for missiles based on terminal sliding-mode control is proposed. Firstly, the terminal sliding mode control is introduced, based on which and the backstepping idea the guidance/control law is designed when an integrated guidance/autopilot model of the yaw plane is formulated. Secondly, an estimating method is given for the unavailable information of the maneuvering target, and an auxiliary control based on a sliding mode estimator is used to offset the estimation error. Finally, a simulation of some missile against high maneuvering targets on the yaw plane was made to verify the effectiveness and rightness of the scheme, and the simulation results have shown that high accuracy of hitting target can be got when the scheme is adopted.

The control system design problem for unmanned blimps has not only theoretical significance but also practical signification because the moving behaviors of blimps are complex nonlinear with coupling, unstructured, imprecise models and disturbance. In this paper the sliding mode control method is applied to control the blimp maneuvering due to its strong robustness. The simulation results show that sliding mode control method is suitable for unmanned blimps through comparing to the simulation curves of sliding mode controllers and PD controllers.

In present study, the complex real-time autoregressive (AR) modeling based on the Levision-Durbin recursive algorithm and the maximum entropy spectral estimation is developed to calculate the spectrograms of Doppler blood flow signals for producing more comprehensive information about the components of the blood flow profile and increasing the display quality. A portable Doppler blood flow analysis device, which is based on a digital signal processor (DSP) TMS320VC549 (Texas Instruments) and contains a 240 * 320 LCD color graphic display module (Hantronix), has been used to implement the spectral estimation algorithm and display spectrograms on the LCD in real-time. Directional Doppler spectrograms are computed directly from the in-phase and quardrature components of the Doppler signal. The results are applied to the development of a compact, economic, versatile bi-directional and battery- or line- operated Doppler device, which can be conveniently used to perform basic vascular testing.

FTIR is a great improvement in IR spectroscopy, and two-dimensional infrared (2D IR) correlation spectroscopy well advances its capabilities. But for complicated mixture systems, such as traditional Chinese medicines, the spectra are rather similar and these methods fall short. To improve the resolution of 2D IR spectrum, and make it possible to distinguish complicated mixture systems, the application of wavelet transform to 2D IR was explored in this paper. After performing decomposition, processing, and reconstruction of the set of dynamic spectra, the resolution of the synchronous 2D IR correlation spectrum was improved obviously. More peaks appeared and the peaks became quite clear and separate. Four

Coptis

samples of aged 1-4 could be distinguished with this approach. Using wavelet transform, 2D IR would become more powerful in analysis and discrimination.

This paper deals with applying Euclidian data distance based GA-PSO (Genetic-Particle Swarm Optimization) algorithm (EU-GA-PSO) to PID controller tuning of AVR (Automatic Voltage Regulator). Through this approach, global and local optimal solution can simultaneously achieved for tuning of controller parameter.

As their scales and complexities increase, the computer-based network systems suffer from increasing probability of being intruded or crashed and decreasing dependability. Such a problem can be solved by extending the traditional security research to survivability research. This paper concentrates on the design and the implementation of the architecture and the applications of network systems to achieve its survivability. For the architecture, a double-

barrier

secure structure is constructed, i.e. the outer barrier defends and detects intrusions and the inner one tolerates intrusions and faults. The CORBA-based applications tolerate and detect intrusions and faults and recover from the adverse environment. Then the applications of the network system will be performed successfully despite the intrusions and faults, that is, the survivability of the network system will be achieved.

The objective of this work is to develop an optimization methodology to design adaptive truss structures with multiple optimally placed active members. The finite element model of truss structures with piezoelectric members is presented. The performance function is built for optimal design of active members at discrete locations in the output feedback control system by using the method proposed by K. Xu et al. Genetic algorithm (GA) is used to search the optimal locations of active members for vibration suppression of adaptive truss structure. A numerical example of the planar truss structure with two piezoelectric active members is performed, and the corresponding experimental set-up is designed for active vibration control. The experimental results demonstrate the effectiveness of optimal placement of active members for adaptive truss structures using genetic algorithm.

Stream Control Transmission Protocol (SCTP) is the third transport layer protocol next to TCP and UDP. The SCTP provides some distinctive features over the TCP. This paper is purposed to compare SCTP and TCP in the performance perspective. We compare the throughput of SCTP and TCP for the three different test scenarios: the performance comparison of SCTP and TCP for the different size of the user input data for the socket system call, the analysis of the fairness under competition of SCTP and TCP traffic, and the performance comparison of the SCTP multi-homing and single-homing cases. From the results, it is shown that the SCTP provides better throughput over TCP for a larger user input data. We also see that the SCTP traffic tends to compete fairly with TCP and that the multi-homing SCTP provides better performance than the single-homing case.

Fusion of the states of a nonlinear dynamic multiscale system (DMS) on the basis of available noisy measurements is one of the well-known key problems in modern control theory. To the best of our knowledge, all of the previous work focused attention on linear DMS. However, nonlinear DMS has never been investigated. In this paper, modeling and fusion estimation of dynamic multiscale system subject to nonlinear measurement equation is proposed. Haar wavelet is used to link the scales. Monte Carlo simulation results demonstrate that the proposed algorithm is effective and powerful in this kind of nonlinear dynamic multiscale system estimation problem.

We present a method for a template matching and an efficient cascaded object detection. The proposed method belongs to wide criteria which can regard to the “feature-centric”. Furthermore, the proposed cascade method has some merits to the face changes. The proposed method for an object detection uses to find the object to most approach better than to find the object to correspond completely. Therefore, this method can use to detect the many faces mixed with different objects. We expect that the result of this paper can be contributed to develop more detection methods and recognition system algorithm.

Robot arms with redundant degrees of freedom (DOF) are crucial in service robots where smooth trajectories and obstacle avoidance in the working area are needed. This paper presents a closed form analysis on the joint limits, and elbow movement minimization using the redundancy of a seven DOF manipulator based on a closed form inverse kinematics. Using the redundancy circle of the redundant arm, the NULL space, the motion planning has been performed. The solution provided has the advantage of being exact with low computational cost. Consequently, this method eliminates the need for trial-and-error which takes times and may not result in a desirable solution. Experimental results have been provided showing the advantages of closed form inverse kinematics over the iterative methods.

This paper introduces the combined fuzzy-based approaches to detect the anomalous network traffic such as DoS/DDoS or probing attacks, which include Adaptive Neuro-Fuzzy Inference System (

ANFIS

) and Fuzzy C-Means (

FCM

) clustering. The basic idea of the algorithm is: at first using

ANFIS

the original multi-dimensional (

M-D

) feature space of network connections is transformed to a compact one-dimensional (

1-D

) feature space, secondly

FCM

clustering is used to classify the

1-D

feature space into the anomalous and the normal.

PCA

is also used for dimensional reduction of the original feature space during feature extraction. This algorithm combines the advantages of high accuracy in supervised learning technique and high speed in unsupervised learning technique. A publicly available DRAPA/KDD99 dataset is used to demonstrate the approaches and the results show their accuracy in detecting anomalies of the network connections.

In autonomous mobile robot industry, the landmark-based localization method is widely used in which the landmark recognition plays an important role. The landmark recognition using visual sensors relies heavily on the quality of the image segmentation. In this paper, we use seat numbers as the landmarks, and it is of great importance to the seat number recognition that correctly segment the number regions from images. To perform this assignment, the support vector machine method is adopted to solve the color image segmentation problems because of its good generalization ability. The proposed method has been used for the mobile robot localization problems, and experimental results show that the proposed method can bring robust performance in practice.

The case where multiple control tasks share one embedded CPU is considered. For various reasons, both execution times of these tasks and CPU workload are uncertain and imprecise. To attack this issue, a fuzzy logic based feedback scheduling approach is suggested. The sampling periods of control tasks are periodically adjusted with respect to uncertain resource availability. A simple period rescaling algorithm is employed, and the available CPU resource is dynamically allocated in an intelligent fashion. Thanks to the inherent capacity of fuzzy logic to formalize control algorithms that can tolerate imprecision and uncertainty, the proposed approach provides runtime flexibility to quality of control (QoC) management. Preliminary simulations highlight the benefits of the fuzzy logic based feedback scheduler.

The cable compensation system is an experiment system that performs simulations of partial or microgravity environments on earth. It is a highly nonlinear and complex system. In this paper, a network based on the theory of the Fuzzy Cerebellum Model Articulation Controller (FCMAC) is proposed to control this cable compensation system. In FCMAC, without appropriate learning rate, the control system based on FCMAC will become unstable or its convergence speed will become slow. In order to guarantee the convergence of tracking error, we present a new kind of optimization based on adaptive GA for selecting learning rate. Furthermore, this approach is evaluated and its performance is discussed. The simulation results shows that performance of the FCMAC based the proposed method is stable and more effective.

In respect that, in practical systems, we usually merely consider oscillation while strong oscillation is sometimes ignored which is also of wide applied background, this paper presents some results of the oscillation and strong oscillation of impulsive neutral parabolic differential systems with delays. Some criteria on the oscillation and strong oscillation are established by using analytical techniques. It is shown that, for impulsive parabolic differential systems with delays, although strong oscillation has more restriction than oscillation, the result of strong oscillation can be parallel to that of oscillation under certain conditions.

In this paper Monte Carlo Method (MCM) is used for tracking slow or fast fading in one antenna communication channel and in multi-antenna channels with space-time block coding (STBC), we compare it with Kalman filter tracking method, discuss its tracking ability when the system has carrier frequency offset, Simulation shows that MCM can be used as a blind method for channel tracking, many research hot spots of MCM are given in the end.

In this paper, a new eye detection method is presented. The method consists of three steps: (1) extraction of binary edge image (BEI) based on the multi-resolution analysis of wavelet transform; (2) extraction of eye region and segments from BEI, and (3) eye localization using light dot or intensity information. An improved face region extraction algorithm and a light dot detection method are proposed to improve eye detection performance. Experimental results show that our approach can achieve a correct eye detection rate of 98.7% on 150 Bern images with variations in view and gaze direction and a rate of 96.6% on 564 AR images with different facial expressions and lighting conditions.

Firstly, a method is introduced which uses Volterra series deploying technique to construct a nonlinear model based on OFS model. Then an improved novel incremental mode multiple steps adaptive predictive control strategy is brought forward, which can import more information about the system’s dynamical characteristics. Experiments of constant water pressure equipment’s control prove that this proposed algorithm can effectively alleviate system’s oscillation when used to control a plant with severe nonlinearity, and that this algorithm shows good robustness for outer disturbances. So it is suitable to be generalized to the design of complex industrial process controller.

As people’s web usage is growing bigger, personalized support for web browsing is in great demand. Furthermore, the diversity of a user’s interests demands an appropriate methodology for profiling multiple domains of user interests. To comply with such demands, we propose one feasible design approach to support personalized web usage, in which a

web user agent

takes over the task of learning and profiling the multiplicity and the changeability of user interests. To evaluate the advantages of this approach, we have constructed a personalized web supporting system, in which an autonomous agent, namely the

web guide agent

, utilizes the information gathered by the web user agent for adaptation, e.g., selective retrieval and re-ranking of web links, and automatic delivery of specific web pages. Compared to other design alternatives, the proposed scheme is operationally simple, while producing acceptably reliable outcome.

Stream Control Transmission Protocol (SCTP) is a new end-to-end transport protocol, which can be used to support the mobility of mobile terminals. This paper describes a framework of SCTP handover and analyzes the handover latency for the single-homing mobile terminals. We then show the experimental results of the SCTP handover in terms of the handover latency and throughput for the two different movement patterns: linear and crossover patterns. For the linear movement pattern, it is shown that the SCTP handover latency may severely depend on the handover delay at the underlying link layer. In the case of the crossover movement pattern, we see that the throughput of the data transmission could be degraded, as the crossover movements occur more frequently.

This paper proposes a time series prediction method for the nonlinear system using the fuzzy system and the genetic algorithm. At first, we obtain the optimal fuzzy membership function using the genetic algorithm. With the optimal fuzzy rules and the input differences, a better time prediction series system may be obtained. In addition, we may obtain the optimal fuzzy membership functions in terms of the evolutionary strategy and we obtain the time series prediction methods using the optimal fuzzy rules. We compare the time series prediction method using the genetic algorithm with that using the evolutionary strategy.

We propose a new sub-term relation to specify syntax of messages with MAC (Message Authentication Code) payload for the cryptographic protocols in the strand space model. The sub-term relation was introduced to formal analysis of cryptographic protocols based on theorem proving, but some defects have been found in it. In the present paper, first, the

operatorf

is defined to the extend sub-term relation, which is used to amend its original flaws. Second, a new ideal is constructed, and is used to expand the bounds on the penetrator’s abilities. Third, the decidable theorem for honesty of ideals holds as it is described under the extended sub-term relation is proved. Fourth, we propose the theorem of the satisfiability for decidable conditions of honest ideals and annotate how invariant-sets generate, which is used to verify security properties of cryptographic protocols.

This article proposes a neural network for building Chinese and English semantic resources connection. Abundant monolingual semantic information is stored into its bipartite graph structure respectively. Two hidden layers are also set in every part, word layer and concept layer. Every word associates with different concepts separately; every concept includes different vocabularies; and these two layers also independently connect to their counterparts through bipartite graph. These distributed characteristics in hidden layers meet the need of parallel network computing. The unsupervised method is used to train the network, and samples are translation lexicons, results of the bilingual word-level alignment algorithm. The training principle comes from the inspiration of bilingual semantic asymmetry. Every translational equivalent contains the unambiguous information by comparison between source and target languages. These translation lexicons are viewed as a kind of special context. They almost have definite meaning. Every input will activate and suppress various kinds of potential connections by the interaction of hidden layers, and modify their connective weights. Finally a demo test presents.

Data gathering is a broad research area in wireless sensor network. In this paper, we consider the problem of routing between the base station and remote data sources via intermediate sensor nodes in a homogeneous sensor network. Sensor nodes have limited and unreplenishable power resources, both path energy cost and path length are important metrics affecting sensor lifetime. In this paper, we first explore the fundamental limits of sensor network lifetime that all algorithms can possibly achieve. Different from previous work, we explicitly consider the constraints of the limited energy and the limited end-to-end latency. We then model the formation of length and energy constrained paths and define the new composite metrics for energy-latency-optimal routing. We also design a distributed data gathering protocol called ELAG (Energy and Latency Aware data Gathering). This protocol balances energy consumption across the network by periodically determining a new optimal path consistent with associated energy distributions. Simulation results testify to the effectiveness of the protocol in producing a longer network lifetime.

In this paper, we propose a scalable energy efficient medium access control protocol (SEMAC) based on time division multiple access (TDMA) technique for wireless sensor networks (WSNs), which uses the local information in scheduling, eliminates most collisions, is more energy efficient and is scalable to the number of sensor nodes in WSN. SEMAC uses the concept of periodic listen and sleep in order to avoid idle listening and overhearing. To balance the energy used in the whole network, SEMAC lets the node with lower energy be a winner in an election procedure based on their energy levels and the winner has more chances to sleep to save energy. We also use a clustering algorithm to form clusters so as to increase the scalability of SEMAC. The performance of SEMAC is evaluated by simulations, and the results show the gain in energy efficiency and scalability.

A multitude of applications of wireless sensor networks require that the sensor nodes be location-aware. Range-based localization schemes are sometimes not feasible due to hardware cost and resource restriction of the sensor nodes. As cost-efficient solutions, range-free localization schemes are more attractive for large-scale networks. This paper presents Weighted Centriod (W-Centriod), a novel range-free localization scheme extended on the basis of Centroid scheme, which takes received signal strength indicator (RSSI) metric into account besides connectivity metric used in Centroid scheme. It’s shown that our W-Centriod method outperforms Centriod scheme significantly in terms of both the average localization error and the uniformity of error distribution across different positions, which decrease by 49.3% and by 37.7%, respectively, under the best circumstance. Moreover, a two-phase localization approach consisting of a field data collection phase and an off-line parameter optimization phase is proposed for localization in wireless sensor networks.

Sensor networks are composed of a large number of densely deployed sensors. The sensor nodes are self-organized and form an ad hoc network. As the energy supply to sensor nodes is limited and cannot be replenished, energy efficiency is an important design consideration for sensor networks. We design and evaluate an energy efficient routing algorithm for data querying sensor networks that propagates routing instructions and build data paths by considering both the hop count to the sink node and the minimum residual energy of that path. The proposed Dynamic Energy Aware Routing (DEAR) algorithm can effectively choose a data path with low energy consumption and high residual energy. The simulation results show that DEAR can prolong the lifetime of networks compared with Directed Diffusion, Minimum Transmission Energy routing and Energy Aware Routing.

To increase the life of the sensor networks, each sensor node has to conserve energy. Due to the fact that each sensor node has one battery energy to remain alive for long times, energy management is a one of critical issues in wireless sensor networks. In this paper we propose a power efficient and low-delay MAC protocol (MT-MAC, Modified T-MAC) for wireless sensor networks. We first address the protocol design problem and suggest a novel solution based on media access control protocol. The MT-MAC uses network traffics to determine active and sleep periods to save energy and enhances the packet transmission latency. The MT-MAC shows better energy saving than the previous proposed protocols, S-MAC and T-MAC.

Based on the classical control system, the framework of information fusion in stochastic optimal control system and the architecture of multisensor information fusion system are developed. The fusion estimation algorithm for control system is applied to state estimation. A structure of sensor management(SM) method and the design method are developed. The practical example shows the performance of state estimation is improved by applying information fusion. But if the model error is very large, the measurement model of each sensor must be selected rightly to ensure the performance of state estimation.

Min-max systems are discrete event systems whose timing involves the maximum, minimum, and addition operations. In recent years, progresses have been made in such topics as cycle time computation, ultimate periodicity of trajectories, structural properties, cycle time assignment, etc. They are surveyed in this paper. Furthermore, as an attempt to open new directions for further research we propose two new models referred to as

and-or net

and

and-or event graph

. It is found that timed and-or event graphs can be algebraically described by min-max systems. We hope that these new models would accommodate more new results.

Recently the study of workflow diagrams has received considerable attention in business process modelling. Formal methods such as Petri nets have been used to analyze and verify of logical properties. However, to our best knowledge, due to the complexity caused by the extreme flexible nature of workflow processes, little work has been done on the performance analysis for workflow diagrams except intensive simulations or approximation analysis based on Stochastic Petri net (SPN) or queueing theory. In this paper, timed workflow diagrams with both AND and OR logic will be modelled and analyzed as stochastic min-max systems. We will provide provable bounds on average tournaround time. The OR logic (known also as the Discriminator [1] ) requires that a downstream event happens whenever one of the upstream events happens. This is different from the AND logic modelling synchronization which requires that the output event happens when all input events happen.

This paper is the first to propose a hybrid quantum-inspired genetic algorithm (HQGA) for flow shop scheduling problems. In the HQGA, Q-bit based representation is employed for exploration in discrete 0-1 hyperspace by using updating operator of quantum gate as well as genetic operators of Q-bit. Then, the Q-bit representation is converted to random key representation. Furthermore, job permutation is formed according to the random key to construct scheduling solution. Moreover, as a supplementary search, a permutation-based genetic algorithm is applied after the solutions are constructed. The HQGA can be viewed as a fusion of micro-space based search (Q-bit based search) and macro-space based search (permutation based search). Simulation results and comparisons based on benchmarks demonstrate the effectiveness of the HQGA. The search quality of HQGA is much better than that of the pure classic GA, pure QGA and famous NEH heuristic.

Many practical systems in physics, biology, engineering, and information science exhibit impulsive dynamical behaviors due to abrupt changes at certain instants during the dynamical processes. In this paper, stability analysis and stabilization synthesis problems are investigated for a class of hybrid dynamical systems which consisting of a family of linear constant subsystems and a rule that orchestrates the switching between them. Furthermore, there exist impulses at the switching instants. A switched quadratic Lyapunov function is introduced to check asymptotic stability of such systems. Two equivalent necessary and sufficient conditions for the existence of such a Lyapunov function are established, respectively. The conditions are in linear matrix inequality form and can be used to solve stabilization synthesis problem.

The fault tolerant supervisory problem for discrete event systems is addressed in this paper. The proposed approach is based on the state avoidance control theory and observer-based control for Petri net. The key idea of the authors is to use a simple linear algebraic formalism to estimate system states and generate diagnostic information. Hence, the state explosion problem is avoided and the observer-based fault diagnosis algorithm can be made on-line.

In this paper, the effects of the interference environments on the performance of the time hopping (TH) binary PPM impulse radio (IR) system are presented. Based on the monocycle pulses available within the frequency of 3.1~10.6 GHz permitted for application by FCC, a PPM modulated TH IR system simulator was designed and followed by the analysis of the monocycle pulse characteristics as well as the system performance. Particularly for the evaluation of the system performance, the multiple access interference and the narrowband system interference signals were considered as the interference signals. Since the narrowband system interference signal has very narrow bandwidth and very large amplitude compared with those of IR system, the analysis of the IR system performance was implemented by considering the interference power and band fraction ratio of the narrowband interference signal.

In this paper, we proposed a channel estimation method by impulse signal train in OFDM. In order to estimate the channel response, 4 impulse signals are generated and transmitted during one OFDM (Orthogonal Frequency Division Multiplexing) symbol. The intervals between the impulse signals are all equal in time domain. At the receiver, the impulse response signals are summed and averaged. And then, the averaged impulse response signal is zero padded and fast Fourier transformed to obtain the channel estimation. The BER performance of the proposed method is compared with those of conventional channel estimation method using the long training sequence in fast fading environments. The simulation results show that the proposed method improves by 3 dB in terms of Eb/No, compared with the conventional method.

As results of study, the coding techniques provide more efficient improvement than a diversity technique, but coding techniques are required the adding bandwidth as many coding rate. Also, when the system is combined MRC diversity technique with coding techniques, the amount of improvement is dramatically increased.

In this paper, the performance of a non-binary convolutional turbo codes is evaluated through computer simulations. Especially, The influence of the various frame sizes and code rate are discussed. Also, in this paper the symbol-by-symbol MAX-Log-MAP algorithm is derived for this coding scheme. We present simulation results for the performance of the non-binary convolutional turbo coded system with QPSK, 16QAM, and 64QAM over an AWGN and the ITU-R channel models.

In this paper, we present an adaptive modulation/demodulation methods based on each load fluctuation after measuring and modeling of noise pattern due to the various load fluctuations of PLC (power line communication) transmission channel. Additionally, we proposed adaptive modulation for PLC using Decision Making Algorithm that could select the optimum modulation/demodulation methods adaptively according to the various load fluctuation. And we certified the availability of adaptive modulation for PLC using the computer simulation and hardware implementation. The newly proposed adaptive modulation can be widely used to attain the high performance PLC implementation in the severe noise environment due to the load fluctuations.

The transmission channel of Home network PLC (power line communication) are characterized by various noise components and delayed waves generated by load fluctuation and multi-path transmission. In this paper, a novel Interference-cancelled Home Network CDMA-PLC system based on the binary ZCD (zero correlation duration) spreading code are proposed as one solution to overcome the previous problems. The properties of the proposed ZCD-PLC systems are effective for MPI(multi-path interference) and MAI (multiple access interference) cancellation in the CDMA-PLC (code division multiple access-PLC) systems. By BER performance simulation, we certified the availability of proposed ZCD-CDMA-PLC system.

The large-scale implementation and deployment of biometric systems demand the concentration on the security holes, by which a reliable system can loose its integrity and acceptance. Like the passwords or PIN codes, biometric systems also suffer from inherent security threats and it is important to pay attention on the security issues before deploying a biometric system. To solve these problems, this paper proposes a novel chaotic encryption method to protect and secure biometric templates. To enhance the security of the templates, this research uses two chaotic maps for the encryption/decryption process. One chaotic map generates a pseudorandom sequence, which is used as private key. While on the other hand, another chaotic map encrypts the biometric data. Experimental results show that the proposed method is secure, fast, and easy to implement for achieving the security of biometric templates.

An intelligent surveillance system can judge and handle a situation automatically within a wide monitoring area and unattended environment that has no certain human supervisor. In this paper, we propose a way to track persons through non-overlapping cameras that are connected over a network with a server. To track persons with a camera and send the tracking data to other cameras, the proposed system uses a human model that comprises a head, a torso, and legs. Also, with a trajectory model, the proposed system can predict the probability which an exited person from one camera is incoming to other cameras. The system is updated online during the lifetime of the system. These enable the proposed to keep tracking the recognized person in a wide area, to provide a guide for monitoring multiple cameras, and to adapt changes with time.

This paper presents a new wavelet domain look-up table (LUT) watermarking algorithm that leads to the sub-optimal embedding of watermarks in the sense of minimizing distortion. The algorithm provides a joint distortion-robustness design of the LUT based watermark. There are two key features in the algorithm: (1) a near minimum-distortion LUT with the maximum run of 2 is designed based on the statistical properties of the wavelet coefficients; (2) an expectation-maximization (EM) algorithm based method is employed to model the statistical distribution of wavelet coefficients and select significant coefficients (coefficients with large magnitude) for watermark embedding. The experimental results show that images watermarked by the proposed algorithm have about 1.5-2.5dB peak-signal-to-noise-ratio (PSNR) gain over the conventional odd-even embedding method, while the system presents great robustness. In the case of 0.2bpp (1/40) JPEG2000 compression, the proposed scheme can ensure a reasonably low bit error rate (BER).

Motivated by the need for designing secure and user-friendly authentication method for mobile devices, we present a novel image-based authentication (IBA) scheme in this paper. Its mnemonics efficacy rests on the human cognitive ability of association-based memorization. To tackle the shoulder-surfing attack issue, an interactive authentication process is presented. System performance analysis and comparisons with other schemes are presented to support our proposals.

Digital seal can be used to assure the authenticity, integrity and undeniability of electronic documents. In this paper, we implement a new digital seal system called as ASS (AssureSeal System) based on digital watermarking and digital signature technologies. ASS can protect not only the electronic documents, but also their paper copies through digital signature as well as robust and reliable digital watermarking technology. First we introduce the framework and the function modules of ASS, then propose a new technology that adds seal image into electronic documents, which combines the patented binary digital watermarking algorithm with normal digital signature technology, integrated PKI digital certificate. At last, we introduce the specifications and applications of ASS.

In this paper, we present ThresPassport (Threshold scheme-based Passport), a web-based, distributed Single Sign-On (SSO) system which utilizes a threshold-based secret sharing scheme to split a service provider’s authentication key into partial shares distributed to authentication servers. Each authentication server generates a partial authentication token upon request by a legitimate user after proper authentication. Those partial authentication tokens are combined to compute an authentication token to sign the user on to a service provider. ThresPassport depends on neither Public Key Infrastructure (PKI) nor existence of a trustworthy authority. The sign-on process is as transparent to users as Microsoft’s .NET Passport. ThresPassport offers many significant advantages over .NET Passport and other SSOs on security, portability, intrusion and fault tolerance, scalability, reliability, and availability.

With the progress of portable appliances such as cell phone and handheld PC, it can be foreseen the popularization of Personal Area Network (PAN) and the diversification of services that are based on Personal Mobile Gateway (PMG). Although inter-operability among home appliances reached a service stage, researches about virtual home services and middlewares for a new small scale network such as PAN are at an early stage. Network service/connection methods, terminal control schemes, and middlewares which are used in the traditional home networks must be enhanced to accommodate PMG-based PAN. In this paper, we propose an integrated virtual home network platform that guarantees seamless connections between home network and PAN. We also analyze the proposed indispensable functions and presents functions that should be added to the existent home gateway or the home server.

This research demonstrates an efficient traffic information provision system for mitigating traffic congestion in the street networks based on multi-type traffic detectors which include inductive loop, image recognition, beacon, and personal communication service (PCS). A methodology is demonstrated in this study, which combines and processes data including vehicle location and speed from the wireless communication network, i.e. PCS in order to validate real time traffic information through PCS and internet.

In this paper we first propose an I-frame encryption scheme for encryption of moving image video data. Second, we propose a licensing agent which provides automatic user authentication and data decoding when multimedia data encrypted in the system server is executed in the client system by the user. The licensing agent performs user authentication based on Public Key Infrastructure (PKI) using a shared key pool and encryption/decryption of multimedia data. After designing and implementing the proposed system, performance tests were then performed using video data files of various sizes for performance evaluation. We verified that the proposed system significantly reduces delay time, including decryption time, when playing back video data files in the client system compared with existing systems.

This paper proposes an improved Traffic-Controlled Rate Monotonic (TCRM) priority scheduling algorithm as a scheduling method in order to transmit real-time multimedia data in a wireless ATM networks. A real-time multimedia data transmission scheduling policy is applied using a different method to uplink or downlink states according to the wireless communication environment, and guarantees the requirement of QoS for both real-time and non-real-time data. In addition, it deals the issue of fairness to share and distribute insufficient wireless resources. Moreover, an issue of inefficiency for non-real-time data, which is a demerit of a TCRM, can be solved using an arbitrary transmission speed by configuring a Virtual Control (VC) in a Base Station (BS).

This paper describes an automatic change detection of roads using aerial photos and digital maps. The task is based on the idea that one can derive information about the changes strictly from its imagery once the geometric relationship among data sets is correctly recovered. The goal of research is achieved by using the Modified Iterated Hough Transform (MIHT) algorithm, the result of which not only solves the orientation parameters of the aerial camera but also filters out blunders from all possible combination of their entities. To examine the effectiveness of the MIHT algorithm, a digital road map and an aerial photo are used to detect changes. Experimental results demonstrate the potential of the MIHT algorithm for detecting changes of the Geospatial Information System (GIS) data.

In a Ubiquitous environment many individual devices which have the communication ability with the computation ability are connected with one another. For this change we must be preceded study of Radio Frequency Identification (RFID) afterward research of a sensor network. And we must grasp the threat to protect the RFID system. In this paper we propose safe RFID system modeling based on shared key pool in ubiquitous environment. This proposed modeling is secured to RFID system from many security threat and privacy problem, and it has shorter access delay time. We expect this proposed modeling to be used in ubiquitous environment for the RFID system being safe.

In this paper, we present a robust method to tackle the ambiguities in 3D arm tracking, especially those introduced by depth change (distance of the arm from the camera), and arm rotation about humerus (upper arm bone). In a particle filter framework, the arm joint angle configurations are monitored and the occurrences of the ambiguous arm movements are detected. Inverse kinematics is applied to transfer invalid joint angle configurations from unconstrained movement space into constrained space. Experimental results have demonstrated the efficacy of the proposed approach.

In order to understand the development of stem cells into specialized mature cells it is necessary to study the growth of cells in culture. For this purpose it is very useful to have an efficient computerized cell tracking system. In order to get reliable tracking results it is important to have good and robust segmentation of the cells. To achieve this we have implemented three levels of segmentation: based on fuzzy threshold and watershed segmentation of a fuzzy gray weighted distance transformed image; based on a fast geometric active contour model by the level set algorithm and interactively inspected and corrected on the crucial first frame. For the tracking all cells are classified into inactive, active, dividing and clustered cells. A special backtracking step is used to automatically correct for some common errors that appear in the initial forward tracking process.

In this paper, we present a novel approach for pose estimation and tracking of license plates from monocular camera view. Given an initial estimate, we try to track the location, motion vector and pose of the object in 3D in the successive video frames. We utilize Condensation algorithm for estimating the state of the object and filtering the measurements, according to the extracted image features. We utilize directional gradients as the image features. Each sample of the Condensation algorithm is projected to the image plane by perspective camera model. The overlapping of the image gradients and the sample boundaries, gives a likelihood for each sample of the Condensation algorithm. Our contribution is utilizing condensation algorithm for rigid object tracking, where the object is tracked in 3D. We demonstrate the performance of the approach by tracking license plates in outdoor environment with different motion trajectories.

In this paper, we propose a new multi-view approach to object tracking method that adapts itself to suddenly changing appearance. The proposed method is based on color-based particle filtering. A short-term memory and a global appearance memory are introduced to handle sudden appearance changes and occlusions of the object of interest in multi-camera environments. A new target model update method is implemented for multiple camera views. Our method is robust and versatile for a modest computational cost. Desirable tracking results are obtained.

This paper presents a 3D feature-based people tracking algorithm which combines an interacting multiple model (IMM) algorithm with a cascade multiple feature data association algorithm. The IMM algorithm in this paper only uses an adaptive Kalman Filter and two dynamic models consisting of a constant velocity model (CV) and a current statistics model (CS) to predict the 3D location of people maneuvering and update the prediction with corresponding measurement. The cascade multiple feature data association algorithm in this paper utilizes three hypotheses, including the nearest distance hypothesis, the velocity consistency hypothesis, and the intensity consistency hypothesis, in turn to determine which trajectory a measurement should be assigned to. Experimental results demonstrate the robustness and efficiency of the proposed framework. It is real-time and not sensitive to the variable frame to frame interval time. It also can deal with the occlusion of people and do well in those cases that people rotate and wriggle.

The shunt active power filter has been proved to be a useful means to improve power quality. This paper investigates the utilization of a new control strategy for three-phase three-wire voltage source shunt active power filter. The proposed scheme is composed of a self-tuning fuzzy logic controller and a series of generalized integrators. Self-tuning fuzzy controller, which operates through the error between the dc-link voltage and a reference value, regulates the voltage level of the dc-link and its output is regarded as the amplitude of grid reference current. The PI current controllers using stationary-frame generalized integrators are adopted, which can eliminate the steady state errors and compensate current harmonics selectively. In addition, an improved Fourier analytic approach is proposed to obtain the phase of the grid current reference by analysis of the grid voltage. The feasibility of the proposed scheme is validated by experimental results from a prototype.

A production-delivery system in a supply chain is always expected to reduce its overall production and management cost. In this research, a decision-making model is developed for optimal production rate selection in a single-stage supply chain system with volume flexibility, where raw materials and/or components are procured from suppliers and processed into finished products which are delivered to customers periodically at a fixed quantity with a fixed interval of time. In this model, production rate is perceived as a decision variable and unit production cost becomes a function of production rate. A pragmatic computational approach is presented to solve the proposed model for special unit production cost functions. Finally, a numerical study is conducted to illustrate the optimal solution and computational approach.

Wavelet-domain hidden Markov Tree (HMT) was recently pro-posed and often applied to image processing. In this paper, HMT is app-lied to denoise the flame image of boiler and has gotten a good result. Having compared with other denoise methods such as wavelet, Wiener filter and median filter. HMT can get better denoise result and the content of flame image edges can be kept better. With the development of HMT research, it will be extended to the fields of signal processing, detection of edge and classification.

Image segmentation is kernel part of image analysis and processing. Fast and robust portrait segmentation and fusion is still a challenging problem by far. In this paper, we present two fast and robust methods to segment portrait in blue background. One method is threshold obtaining based on histogram peak distribution, the other is Quantum Evolution Algorithm (QEA) threshold searching based on group variance. Detailed experiments and comparison analyses are presented to demonstrate the performance of our methods.

In this paper we have studied the suppression of chaotic vibration in the system with impacts. Period pulse method of chaos suppression in the system with impacts has been presented. We consider the stable fixed point as the control target. Once the state signals derivate from the control target, the system dynamically produces period impulse signals to suppress chaos and bifurcation. The method has been employed to a two-degree-of–freedom reciprocating impact vibration model. Using the stable fixed point of Poincaré map equation as the control target, suppressing bifurcation and chaos under different parameters by numerical simulation. The results show that the method can suppress chaotic motion effectively.

A spiking-neuron based cognitive model with classical conditioning behaviors is proposed. With a reflex arc structure and a reinforcement learning method based on the Hebb rule, the cognitive model possesses the property of ‘stimulate-response-reinforcement’ and can simulate the learning process of classical conditioning. An experiment on the inverted pendulum validated that this model can learn the balance control strategy by classical conditioning.

A new machine learning method called probabilistic tangent subspace is introduced to improve the performance of the equalization for the M-QAM modulation signals in wireless communication systems. Due to the mobility of communicator, wireless communication channels are time variant. The uncertainties in the time-varying channel’s coefficients cause the amplitude distortion as well as the phase distortion of the M-QAM modulation signals. On the other hand, the Probabilistic Tangent Subspace method is designed to encode the pattern variations. Therefore, we are motivated to adopt this method to develop a classifier as an equalizer for time-varying channels. Simulation results show that this equalizer performs better than those based on nearest neighbor method and support vector machine method for Rayleigh fading channels.

We have proposed a new feature extraction method and a new feature fusion strategy based on generalized canonical correlation analysis (GCCA). The proposed method and strategy have been applied to facial feature extraction and recognition. Compared with the face feature extracted by canonical correlation analysis (CCA), as in a process of GCCA, it contains the class information of the training samples, thus, aiming for pattern classification it would improve the classification capability. Experimental results on ORL and Yale face image database have shown that the classification results based on GCCA method are superior to those based on CCA method. Moreover, those two methods are both better than the classical Eigenfaces or Fishierfaces method. In addition, the newly proposed feature fusion strategy is not only helpful for improving the recognition rate, but also useful for enriching the existing combination feature extraction methods.

In this paper we propose a novel clustering algorithm in mobile adhoc network. By selecting the node optimally in both time connectivity and space connectivity as the cluster head with Genetic Algorithm (GA), the resulting clustering algorithm can provide a generic, stable and lower communication overhead cluster structure for the upper-layer protocols. For this clustering scheme, we give analytical model and evaluate the performance by simulation.

As a novel computational approach from swarm intelligence, an Ant Colony Optimization algorithm attracts more researches, and has been applied in many fields. The paper proposes a pair-ant colony algorithm for multiuser detecting problem based on the Max-Min ant colony algorithm. The optimum multiuser detector has the best performance, but its computation complexity is very high, it is an NP-complete problem. Experiment results show that the proposed method improves the search quality, lower the iteration times, its performances are better than those of the conventional detector and decorrelating detector, and its complexity is lower than that of optimum detector.

In this paper, we present an enhanced massively multi-agent system based on the previous MMAS for discovering the unique dynamics of HIV infection [1]. The enhanced MMAS keeps the spacial characteristics of cellular automata (CA), and employs mathematical equations within sites. Furthermore, new features are incorporated into the model, such as the sequence representation of HIV genome, immune memory and agent remote diffusion among sites. The enhanced model is closer to the reality and the simulation captures two extreme time scales in the typical three stages dynamics of HIV infection, which make the model more convincing. The simulation also reveals two phase-transitions in the dynamics of the size of immune memory, and indicates that the high mutation rate of HIV is the fatal factor with which HIV destroys the immune system eventually. The enhanced MMAS provides a good tool to study HIV drug therapy for its characterizing the process of HIV infection.

In this paper, we introduce an efficient feature extraction method for character recognition. The EA strategy is used to maximize the Fisher linear discriminant function (FLD) over a high order Pseudo-Zernike moment. The argument, which maximizes the FLD criteria, is selected as the proposed weight function. To evaluate the performance of the proposed feature, experimental studies are carried out on the historic Middle-Age Persian characters. The numerical results show 96.8% recognition rate on the selected database with the weighted Pseudo-Zernike feature (with order 10) and 65, 111,16 neurons for the input, hidden, and output layers while this amount for the original Pseudo-Zernike is 93%.