Search

The semiconductor manufacturing process is very complex, and it is the most important part of the semiconductor industry. In order to test whether or not wafers are functioning normally, a pass/fail test is conducted; however, time and cost needed for this testing increase as the number of chips increases. To address this, a machine learning technique is adopted and a high-performance classifier is needed to determine whether a pass/fail test is accurate or not. In this paper, a deep belief network (DBN)-based multi-classifier is proposed for fault detection prediction in the semiconductor manufacturing process. The proposed method consists of two phases: The first phase is a data pre-processing phase in which features required for semiconductor data sets are extracted and the imbalance problem is solved. The second phase is to configure the multi-DBN using selected features. A DBN classifier is created for each feature and, finally, fault detection prediction is performed. The proposed method showed excellent performance and can be used in the semiconductor manufacturing process efficiently.

The worldwide trend of aging is highlighting the importance of the silver generation as a major consumer group for various industries. In particular, the necessity of mobile services for the silver generation in the IT industry is increasing. The purpose of this research is to develop a smart phone utilization plan by analyzing the needs of developing mobile services applying the characteristics of the silver generation on smart phones and by studying the smart phone acceptance factors.

A cloud computing is one of a parallel and distributed computing. The cloud computing provides some service for user with virtual resources. However, a user’s service request does not show a time pattern. As a result, each resource also shows a different availability at the same time. This difference affects a quality of service (QoS) and a resource selection for users. Therefore, we propose the resource availability-based multi auction model for cloud service reservation and resource brokering system. The proposed system is to select the proper resource provider based on the users’ request. The proposal adopts the multi phase of the auction to transact resources. The system evaluates the available factor of each resource on the auction phase, and finally reserves the service on the adaptive queue. The proposed model shows the better performance than other existing method.

This paper proposes the Fuzzy Rule-based Adaptive Coronary Heart Disease Prediction Support Model (FbACHD_PSM), which gives content recommendation to coronary heart disease patients. The proposed model uses a mining technique validated by medical experts to provide recommendations. FbACHD_PSM consists of three parts for heart disease risk prediction. First, a fuzzy membership function is constructed using medical guidelines and statistical methods. Then, a decision-tree rule induction technique creates mining-based rules that are subjected to validation by medical experts. As the rules may not be medically suitable, the experts add rules that have been verified and delete inappropriate rules. Thirdly, using fuzzy inference based on Mamdani's method, the model predicts the risk of heart disease. Based on this, final recommendations are provided to patients regarding normal living, nutrition control, exercise, and drugs. To implement our proposed model and evaluate its performance, we use a dataset from a single tertiary hospital.

Cloud computing is one of the distributed computing environments and utilizes several computing resources. Cloud environment uses a virtual machine to process a requested job. To balance a workload and process a job rapidly, cloud environment uses a provisioning technique and assigns a task with a status of virtual machine. However, a scheduling method for cloud computing requires a definition of virtual machine availabilities, which have an obscure meaning. In this paper, we propose Fuzzy logic driven Virtual machine Provisioning scheduling using Resource Evaluation(FVPRE). FVPRE analyzes a state of every virtual machine and actualizes a value of resource availability. Thus FVPRE provides an efficient provisioning scheduling with a precise evaluation of resource availability. FVPRE shows a high throughput and utilization for job processing on cloud environments.

In order to enhance the performance, rare class prediction are to need the feature selection method for target class-related feature. Traditional data mining algorithms fail to predict rare class, as the class imbalanced data models are inherently built in favor of the majority of class-common characteristics among data instances. In the present paper, we propose the Euclidean distance- and standard deviation-based feature selection and over-sampling for the fault detection prediction model. We study applying the semiconductor manufacturing process control in fault detection prediction. First, the features calculate the MAV (Mean Absolute Value) median values. Secondly, the MeanEuSTDEV (the mean of Euclidean distance and standard deviation) are used to select the most appropriate features of the classification model. Third, to address the rare class over-fitting problem, oversampling is used. Finally, learning generates the fault detection prediction data-mining model. Furthermore, the prediction model is applied to measure the performance.

M&S (Modeling and Simulation) has been widely used as a decision supporting tool by modeling the structure and dynamics of real-world systems on a computer and simulating the models to answer various what-if questions. As simulation models become complex in their dynamics and structures, more engineers are experiencing difficulties to simulate the models with various real-world scenarios and to discover knowledge from the massive amount of simulation results within a practical time bound. In this paper, we propose a hybrid methodology where the M&S process is combined with a DM (Data Mining) process. Our methodology includes a step to inject simulation outputs to a DM process which generates a prediction model by analyzing pertaining patterns in the simulation outputs. The prediction model can be used to replace simulations, if we need to expedite the M&S-based decision making process. We have applied the proposed methodology to analyze SAM (Surface-to-air missile) and confirmed the applicability.

An interoperable environment requires a structure and framework to build and manage entities of a remotely coupled system. Existing interoperation approach is based on a complicated and rigid procedure with an expertise of that. This feature causes inefficiency in a time and cost of a training and implementation. This paper proposes a consignment technique of a simulation entity to support a plug-in approach for distributed interoperable environment. Proposed consignment technique provides an implied approach for building simulation entities and coupling between those. We expect that our proposal will contribute to improving a building and operation efficiency of a distributed interoperable environment.

Fault detection prediction of FAB (wafer fabrication) process in semiconductor manufacturing process is possible that improve product quality and reliability in accordance with the classification performance. However, FAB process is sometimes due to a fault occurs. And mostly it occurs “pass”. Hence, data imbalance occurs in the pass/fail class. If the data imbalance occurs, prediction models are difficult to predict “fail” class because increases the bias of majority class (pass class). In this paper, we propose the SMOTE (Synthetic Minority Oversampling Technique) based over sampling method for solving problem of data imbalance. The proposed method solve the imbalance of the between pass and fail by oversampling the minority class of fail. In addition, by applying the fault detection prediction model to measure the performance.

A grid computing consists of the physical resources for processing one of the large-scale jobs. However, due to the recent trends of rapid growing data, the grid computing needs a parallel processing method to process the job. In general, each physical resource divides a requested large-scale task. And a processing time of the task varies with an efficiency and a distance of each resource. Even if some resource completes a job, the resource is standing by until every divided job is finished. When every resource finishes a processing, each resource starts a next job. Therefore, this paper proposes a dynamic resource reallocation scheduling model (DDRSM). DDRSM finds a waiting resource and reallocates an unfinished job with an efficiency and a distance of the resource. DDRSM is an efficient method for processing multiple large-scale jobs.

There is a growing tendency toward a vehicle demand and a utilization of traffic information systems. Due to various kinds of traffic information systems and increasing of communication data, the traffic information service requires a very high IT infrastructure. A cloud computing environment is an essential approach for reducing a IT infrastructure cost. And the traffic information service needs a provisioning scheduling method for managing a resource. So we propose a provisioning scheduling with conditional probability inference (PSCPI) for the traffic information service on cloud environment. PSCPI uses a naive bayse inference technique based on a status of a virtual machine. And PSCPI allocates a job to the virtual machines on the basis of an availability of each virtual machine. Naive bayse based PSCPI provides a high throughput and an high availability of virtual machines for real-time traffic information services.

The face recognition has been used in a variety fields, such as identification and security. The procedure of the face recognition is as follows; extracting face features of face images, learning the extracted face features, and selecting some features among all extracted face features. The selected features have discrimination and are used for face recognition. However, there are numerous face features extracted from face images. If a face recognition system uses all extracted features, a high computing time is required for learning face features and the efficiency of computing resources decreases. To solve this problem, many researchers have proposed various Boosting methods, which improve the performance of learning algorithms. Mutual-Boost is the typical Boosting method and efficiently selects face features by using mutual information between two features. In this paper, we propose a GroupMutual-Boost method for improving Mutual-Boost. Our proposed method can shorten the time required for learning and recognizing face features and use computing resources more effectively since the method does not learn individual features but a feature group.

Objectives: The importance of the prediction of coronary heart disease (CHD) has been recognized in Korea; however, few studies have been conducted in this area. Therefore, it is necessary to develop a method for the prediction and classification of CHD in Koreans. Methods: A model for CHD prediction must be designed according to rule-based guidelines. In this study, a fuzzy logic and decision tree (classification and regression tree [CART])-driven CHD prediction model was developed for Koreans. Datasets derived from the Korean National Health and Nutrition Examination Survey VI (KNHANES-VI) were utilized to generate the proposed model. Results: The rules were generated using a decision tree technique, and fuzzy logic was applied to overcome problems associated with uncertainty in CHD prediction. Conclusions: The accuracy and receiver operating characteristic (ROC) curve values of the propose systems were 69.51% and 0.594, proving that the proposed methods were more efficient than other models.

This study focuses on the relation between humans and smart TVs with emotion recognition function. The study hypothesizes that the emotion- based user interface is most effective in human interaction in comparison to other user-centered devices with passive user interface. Forty participants were given three types of user interfaces such as the remote controller, gesture recognition system, voice recognition system, and emotional recognition system to be used for watching the smart TV. They were given interesting contents and sad contents to choose from within a specific time limit. We use the Fraunhofer IIS SHORETM demo software to automatically detect the facial expressions of the participants they exhibited in response to the contents. A study on the preference of contents according to the individual emotional responses of the users was done simultaneously. Additionally, it figured out the relation between four types of emotion as emotion recognition UI on SMART TV screen, the contents, and the study concentrated on the satisfaction and usability of four different user interfaces. As a result of studying the relations of emotions and content preferences from this research, it is recognized that comedy programs are much more preferred based on the degree of happy emotion detected, and that men appear to have higher preference than women. When they were in sad emotion, the valid result value was not achieved though, which shows slight preference for exciting contents. It is noted that women have a little higher preference than men. These results were similar to the standard mood based management of Zilinman’s theory. Another result is that out of the four user interfaces, the emotion recognition-based user interface which selects the contents on the screens of the smart TV showed a higher satisfaction than other passive user interfaces.

A big data is now widely utilized in real-world situations. And the business field requires some cooperation or coupling between different big data systems. However, the big data system does not have a unified format and environments. Thus, this paper presents an agent-based flexible management for big data fusion services. The proposed system utilizes an Internet Relay Chat (IRC) and a concept of IRC-bots. It may provide efficient data fusion services without a concern about the heterogeneous system.

Advances in mobile technology make most people have their own mobile devices which contain various sensors such as a smartphone. People produce their own personal data or collect surrounding environment data with their mobile devices at every moment. Recently, a broad spectrum of studies on Participatory Sensing, the concept of extracting new knowledge from a mass of data sent by participants, are conducted. Data collection method is one of the base technologies for Participatory Sensing, so networking and data filtering techniques for collecting a large number of data are the most interested research area. In this paper, we propose a data collection model in hybrid network for participatory sensing. The proposed model classifies data into two types and decides networking form and data filtering method based on the data type to decrease loads on data center and improve transmission speed.

In wireless network environment, to detect the location of a mobile node (MN) is an important factor for providing high-quality service to a MN. However, a limited communication range, disconnection, location error caused by various environmental influence of moving paths can be occurred in wireless network environment. In this paper, we propose an Intelligent Agent-based Mobility Prediction (IAMP) method that predicts the mobility of a MN by assigning an intelligent agent (IA) to each base station (BS) in a hybrid wireless network environment and provides suitable services according to change of location. To predict the mobility of a MN we build an ontology and carry out an inference based on some defined rules. To demonstrate superiority of the IAMP, we measured the average location error as time. Experimental results comparing the proposed method with established methods verify effectiveness and efficiency of the IAMP.

Grid computing has been noticed as an issue to solve complex problems of large-scale bioinformatics applications and helps to improve data accuracy and processing speed on multiple computation platforms. Outlier detection helps classification success rate high and makes processing time reduce. This paper focuses on a data clustering and classification method with outlier detection which is an important bioinformatics application in grid environment. This paper proposes a grid-based and outlier detection-based clustering and classification(GODDCC) using grid computational resources with geographically distributed bioinformatics data sets. This GODDCC is able to operate large-scale bioinformatics applications in guaranteeing high bio-data accuracy with reasonable grid resources. This paper evaluates performance of GODDCC in comparing to the data clustering and classification(DCC) without outlier detection. The average of processing time of the GODDCC model records the lowest processing time and provides the highest resources utilization than the other DCC models. The outlier detection method reduces processing time for DCC models with maintaining high classification success rate and grid computing gives a great promise of high performance processing with geographically distributed and large-scale bio-data sets in bioinformatics applications.

Algorithms and methods for analyzing large amounts of data are studied and developed. This paper presents a Data Mining (DM) method operated in grid computing environment. Because DM technology uses large amounts of data and requires costs to compute, utilizing and sharing computing data and resources are key issues in DM. Therefore, a Dynamic Load Balancing (DLB) algorithm and a decision range readjustment algorithm are proposed and applied to the Grid-based Data Mining (GDM) method. And we analyzed the average waiting time for learning and computing time. For a performance evaluation, the system execution time, computing time, and average waiting time for learning are measured. Experimental results show that GDM with the DLB method provides many advantages in terms of processing time and cost.

The intelligent agent works powerful jobs for handling system complexity and making systems more modular. Especially a reasoning agent is effective on organizing for decision-making process of systems. This paper introduces an Ontology-based Intelligent Agent for a Grid Resource Management System (OIAGRMS), which uses ontology reasoning to select a suitable resource supplier, is proposed. This paper focuses on effective grid resource management and the improvement of resource utilization through transaction management for the OIAGRMS. For performance evaluation with accuracy and reliability, the OIAGRMS is compared with the Prediction-based Agent for Grid Resource Management System(PAGRMS) and the Random-based Agent for Grid Resource Management System(RAGRMS). The OIAGRMS recorded over 90 percents trade success, but the PAGRMS and RAGRMS recorded less than a 90 percents trade success. In comparing of resource utilization rate, maximum deviation, standard deviation, the OIAGRMS were about 9.4 and 9.8 percents but the PAGRMS are about 22.9 and 16.3 percents, the RAGRMS were about 61.6 and 21.7 percents. The empirical results demonstrate the usefulness and improvement utilization with stable performances of the intelligent agent base on ontology reasoning in grid environment.

Grid technology enables to share and use of various heterogeneous resources. However, it is very difficult to satisfy user requirements and benefits of resource providers. This paper proposes an ontology-based resource management model (ORMM). The model builds an ontology of ORMM for computational grid. As using the ontology and the inference based on some rules, we can reflect requirements of resource users, conclude more transactions, and reduce resting resources. To evaluate performances, the number of transactions, transaction period, and cumulative profits are measured. In experiments, ORMM increases about 32% of greater transactions, reduces about 40% of transaction period. Experimental results demonstrate that the proposed model can improve the efficiency of resource management in the computational grid environment.

Recently, a lot of boosting-based feature selection methods have been introduced. However, the effective methods that can select more discriminative features at lower computational cost are needed. In this paper, we propose a novel boosting algorithm, called the Segment-Boost, and describe an improved feature selection method based on Segment-Boost. Segment-Boost learns the weak classifiers for the feature sets of various sizes through given training examples. Hence, our proposed feature selection method can consider discriminations and complementarities of features for selecting features. In addition, the proposed feature selection method requires lower computational cost than other boosting-based feature selection methods. For experiments, we extracted Gabor feature vectors from 400 face images of ORL facial database through Gabor filters. And randomly generated intra/extra-personal feature vectors from the Gabor feature vectors were used for training and testing examples. From the experimental results, we verified that Segment-Boost can select more discriminative features at lower computational cost than other boosting-based feature selection methods.

Grid computing has been noticed as a solution for digital content management. But, existing grid computing-based digital content management systems cannot meet complex and detailed requirements of a grid user due to the simple search mechanism base on few keywords. Therefore, this paper constructs a digital content ontology that represents digital content information. And, this paper proposes a semantic Web-based digital content management system. In the system, a semantic Web agent analyzes user requirements and infers a proper digital content from the digital content ontology. And, the system selects a grid resource with the cheapest and fastest data service among grid resources which can provide the inferred digital content. Digital content data is transferred from the selected grid resource to the grid user through a local grid resource agent.

The current RFID system is based on a distributed computing system to process a great volume of tag data. Therefore, RFID M/W should provide load balancing for consistent and high system performance. Load balancing is a technique to solve workload imbalance among edge M/Ws and improve scalability of RFID systems. In this paper, we propose a fuzzy logic control-based load balancing agent for distributed RFID systems. The agent predicts data volume of an edge M/W and evaluates workload of the edge M/W by fuzzy logic control. And, the agent acquires workload information of an edge M/W group to which the edge M/W belongs. Also, the agent adjusts over and under-loaded workload according to a dynamic RFID load balancing algorithm. Experimental results demonstrates that the fuzzy logic control-based load balancing agent provides 95% and 6.5% reductions in overload time and processor usage and 13.7% and 12.2% improvements in throughput and turn around time as compared the mobile agent-based load balancing system.

Currently network uses, especially the number of internet users, increase rapidly. Also, high quality of service is required and this requirement results a sudden network traffic increment. As a result, an efficient management system for huge network traffic becomes an important issue. Ontology/data engineering based context awareness using the System Entity Structure (SES) concepts enables network administrators to access traffic data easily and efficiently. The network traffic analysis system, which is studied in this paper, is an effective management system with a high throughput and a low response time designed and implemented based on a model and simulation based data engineering methodology. Extensible Markup Language (XML) is used for metadata language in this system. The information which is extracted from the network traffic analysis system could be modeled and simulated in Discrete Event Simulation (DEVS) methodology for further works such as post simulation evaluation, web services, and etc.

As demand and supply of Grid resources in Grid computing environments increases rapidly, many Grid researchers have taken particular interest in efficient management of Grid resources. In this paper, a user demand prediction-based resource management model (UDPRMM), which uses the user demand trend of Grid resources as a critical management factor, is proposed. This paper introduces a user demand prediction approach with historical user demands and aims for effective Grid resource management through transaction management between resource users and resource suppliers. For performance evaluation with accuracy and reliability, the UDPRMM is compared with the user demand fixed-based resources management model (UDFRMM) and the user demand random-based resources management model (UDRRMM). The UDPRMM demonstrates improvement greater than 70% with regard Grid resource transactions and greater than 16% with regard to Grid resource transaction quantities in comparison with the UDFRMM and the UDRRMM. In evaluation, the UDPRMM reduces delay time by 21%, over that of the UDFRMM, and 19% over that of the UDRRMM. The empirical results demonstrate the usefulness of the UDPRMM with effective economics-based resource management.

This paper proposes the QLP-LBS (Quantization and Location Prediction-based LBS). This QLP-LBS system is based on quantization theory and uses statistical location prediction mechanism. This LBS applies the quantum range of quantization theory to each mobile user and reduces location update costs by comparing results between moving distance of mobile user and quantum range. But, this LBS system generates location errors from the quantization. In order to solve this problem, we apply statistical location prediction mechanism to LBS system. This prediction mechanism predicts location of mobile user using its historical path and decrease location errors by quantization and makes more reliable LBS system. For performance evaluation, this paper measures location accuracy and reduction rate of location update costs with various quantum ranges. This experiments show that QLP-LBS effectively reduces location update costs of LBS system. Also, QLP-LBS solves problem of location errors using location prediction mechanism which is problem of general quantized system. Therefore, QLP-LBS is solution for reduction of location update costs and has reliable location accuracy.

The mobile grid introduces various research challenges distinguished from existing grid computing systems. They are low bandwidth, low processing power, low battery capacity, frequent disconnectivity, and mobility. Mobility of the grid node increases the system load of the mobile grid in a constrained operating environment by increasing the number of communication messages required to confirm the location between the grid broker and mobile grid node. Therefore, this paper proposes an adaptive distance filter that can effectively reduce communication traffic between the mobile grid node and grid broker. This filter constructs clusters based on the mobility and velocity of the grid node and filters the location updates. However, the reduction of location updates generates location errors, which occur when the grid broker cannot acquire the exact location of mobile nodes. To solve this problem, if the location updates are filtered, the grid broker can estimate the location of the mobile node using a statistical estimation method. For the performance evaluation of the adaptive distance filter, we modeled the mobility of the grid nodes. We then measured the reduction in location updates and location errors. In these experiments, we prove that the adaptive distance filter is an effective scheme for reducing location updates and the grid broker can reduce location errors through location estimation.

This paper deals with congestion control in which GRID network community has been interested. In order to solve problems of existing congestion control models like FOBS, this paper proposes the node availability-based congestion control model. The model monitors not only storage availability but also network availability of grid nodes to detect and control congestion. As using the fuzzy adaptive packet allocation mechanism, the model estimates congestion level of grid nodes and adjusts their packet transmission rate for stable and reliable data transfer in computational grid. Experiment results demonstrate that our model outperforms FOBS which is a typical congestion control model with the improved QoS.

The key purpose of computational grid is how to allocate grid resources to grid applications effectively. Existing resource management models with centralized architecture like GRAM and nimrod-G broker are liable to generate communication overheads and bottleneck. Therefore, we propose the service agent-based resource management model using virtuali-zation for computational grid. Virtualization and agent-based system applied to our model is able to improve load imbalance of grid broker and reduce communication overheads. Experiment results demonstrate that the service agent-based resource management model decreases 25.8% more job latencies and 151.6% more communication messages than the centralized resource management model.

Grid computing provides high performance like a super computer through sharing and using distributed heterogeneous computing resources. Grid computing processing time and cost are widely varied since a grid user or a grid middleware can select a variety of distributed heterogeneous resources. Therefore, grid computing absolutely needs a grid resource management method and model. In this paper, we propose two types of resource management model with resource reliability. The first model is the idle resource supplement model. The point of the idle resource supplement model is that adds to idle resources when existing resources can not process jobs. The second model is the validity time designation model which considers grid users. This model processes jobs during validity time that is decided by grid users. This paper evaluates system performance such as utilization, job-loss rate and average turn-around time. And, we estimate experiment results of our models in comparison with those of existing models such as a random model and a round-robin model. The experiment results demonstrate that the two models based on resource reliability measurement improve resource utilization and provide reliable job processing. And, we expect that our proposed models improve grid computing QoS.

Protein sequence data have been revealed in current genome research and have been noticed in demand of classifier for new protein classification. This paper proposes the optimal clustering-based ART1 classifier for the GPCR data classification and processes the GPCR data classification. We focuses on a demand of optimal classifier system for protein sequence data classification. The optimal clustering-based ART1 classifier reduces processing cost for classification effectively. We compare classification success rate to those of Backpropagation Neural Network and SVM. In experimental result of the optimal clustering-based ART1 classifier, classification success rate of ClassA group is 99.7% and that of the others group is 96.6%. This result demonstrates that the optimal clustering-based ART1 classifier is useful to the GPCR data classification. The classification processing time of the optimal clustering-based ART1 classifier is the 27% less than that of the Backpropagation Neural Network and is the 39% less than that of the SVM in an optimal clustering rate which is 15%. And the classification processing time of the optimal clustering-based ART1 classifier is the 39% less than that of the optimal clustering-based ART1 classifier in a prediction success rate which is 96%. This result demonstrates that the optimal clustering-based ART1 classifier provides the high performance classification and the low processing cost in the GPCR data classification.

Communication Data Management (CDM) is an important issue in complex and large-scale modern geo-computation applications. In this paper, we review existing CDM schemes and propose an efficient adaptive quantization-based CDM scheme for geocomputation in grid environment. This model is based on theory of quantization and uses distributed clusters which are concerned about the features and geographical adjacency of communication objects. And, this model uses various thresholds which are suitable for individual clusters and effectively reduces communication messages by clustering and diverse thresholds. For realization of this scheme, we design and develop a HLA (High Level Architecture)-based system for geo-computation in grid environment. For performance evaluation, this paper measures reduction rate of message traffic and location errors. This scheme has more effective message traffic reduction and less location errors with precise location accuracy rate than general quantization-based scheme. The empirical result apparently presents that adaptive quantization- based CDM scheme is a suitable scheme for high-performance geo-computation.

Computational grid technology has been noticed as an issue to solve large-scale bioinformatics-related problems and improves data accuracy and processing speed on multiple computation platforms with distributed bioDATA sets. This paper proposes a grid-based intelligence ART1 classifier which operates an ART1 clustering classification using grid computational resources with distributed GPCR data sets. This paper evaluates performance of the grid-based ART1 classifier in comparing to the grid-based ART1 classifier and the grid-based 3-tier ART1 classifier. The number of communication message of the grid-based intelligence ART1 classifier is less than the others classifier. And the classification processing time of the grid-based intelligence ART1 classifier is 13% classification processing time of the grid-based ART1 classifier and is 25% classification processing time of the grid-based 3-tier ART1 classifier. Computational grid in bioinformatics applications gives a great promise of high performance processing with large-scale and geographically distributed bioDATA sets and provides a good progress in guaranteeing high bioData accuracy with reasonable processing resources

Computational Grid technology has been noticed as an issue to solve large-scale bioinformatics-related problems and improves data accuracy and processing speed on multiple computation platforms with distributed bioDATA sets. This paper focuses on a GPCR data mining processing which is an important bioinformatics application. This paper proposes a Grid-based 3-tier ART1 classifier which operates an ART1 clustering data mining using grid computational resources with distributed GPCR data sets. This Grid-based 3-tier ART1 classifier is able to process a large-scale bioinformatics application in guaranteeing high bioDATA accuracy with reasonable processing resources. This paper evaluates performance of the Grid-based ART1 classifier in comparing to the ART1-based classifier and the ART1 optimum classifier. The data mining processing time of the Grid-based ART1 classifier is 18% data mining processing time of the ART1 optimum classifier and is the 12% data mining processing time of the ART1-based classifier. And we evaluate performance of the Grid-based 3-tier ART1 classifier in comparing to the Grid-based ART1 classifier. As data sets become larger, data mining processing time of the Grid-based 3-tier ART1 classifier more decrease than that of the Grid-based ART1 classifier. Computational Grid in bioinformatics applications gives a great promise of high performance processing with large-scale and geographically distributed bioDATA sets.

Importance and need of grid process scheduling have been increased in accordance with development of grid computing. In order to distribute and utilize grid processors efficiently, grid computing system needs scheduling policies that manage and schedule grid process. This paper reviews current scheduling policies and proposes an efficient scheduling model which is called the predictive process scheduling model. For efficient scheduling, this paper presents the processing time prediction algorithm to resolve problems of grid scheduling. The predictive process scheduling model predicts processing times of processors, allocates a job to a processor with minimum processing time, and minimizes overall system execution times. For performance evaluation, this paper measures turn-around time, job loss, throughput, and utilization. Empirical results show that the predictive process scheduling model operates with the less 69.5% of turn-around time and the less 76.4% of job loss and improves the more 119.6 % of throughput and the more 117.8% of utilization than those of existing scheduling models such as random scheduling and round-robin scheduling.

This paper discusses a concept of DEVS middleware that would provide a set of services for constructing discrete event models and executing them in simulation or in real-time. Although generic middleware such as CORBA and MPI supports simulation and simulation-oriented middleware such as High Level Architecture Runtime Infrastructure exists, there are many issues that either not addressed at all or have been inadequately addressed in existing systems. DEVS Middleware would improve modeling and simulations due to the beneficial formal properties of DEVS, enhance interoperability among components adhering to the DEVS protocol, and reduce programming complexity by hiding lower level Simulation and execution details. The DEVS Middleware concept includes components for model construction and for mapping models onto simulators or real-time execution engines to support the novel concept of "distributed programming by modeling." This paper discusses several sub-layers that have been identified within the overall DEVS middleware concept, existing work that can be brought to bear on developing these layers and the integration needed to create a widely accepted standard.

Complex and large-scale distributed systems are characterized by numerous interactive data communication among distributed components over network. This paper proposes a communication data multiplexing approach as an efficient message traffic reduction scheme. This approach encodes joint outputs of sender components into a single message and applies to a distributed system involving components moving and interacting in multi-dimensional space. For performance evaluation, this paper applies uses a projectile/missile case study with realistic multi-dimensional dynamics. This paper investigates variation of system accuracy and network bandwidth requirement, while a ratio of active components and a time granule are varied. Analytical and empirical data clearly indicate the advantages of multiplexing in saving communication resources in a large-scale distributed simulation. In addition, this paper discusses effectiveness and limitation of the multiplexing approach while considering the tradeoff between system accuracy and performance.

Grid computing is the next generation platform that maximizes computing performance of system as using numerous computing resources on computer networks. Grid computing not only shares computing resources but also efficiently utilizes available computing resources. Therefore, grid computing needs resource management policies that handle and manage grid computing resources. This paper reviews existing resource management models based on economic principles such as auction model and commodity market model. This paper proposes an efficient resource management model, which is called the grid resource trade network model, and presents its bidding algorithms for grid users and grid providers. The proposed model reduces idle resources, improves resource utilization, and provides effective resource distribution. Empirical results show usefulness and efficiency of the grid resource trade network model.

This paper reviews existing data management schemes and presents a design and development of a data management scheme with load balancing in a distributed computing. This scheme defines a variety of degree of load balancing, maps each degree to each data management configuration, and reduces data traffic among distributed components. The scheme allows to share geographically dispersed data assets collaboratively and execute a complex large-scale distributed computing system including cooperating and distributed components with reasonable computation and communication resources. In addition, this paper introduces a HLA (High Level Architecture) bridge middleware environment for data communication among multiple federations. We analyze system performance and scalability with a variety of degree of load balancing configurations. The empirical result on a heterogeneous OS distributed system apparently presents advantages of the data management scheme with load balancing in terms of system performance and scalability.

There is an increasing demand to execute complex large-scale grid computing systems and to share dispersed data assets and computing resources collaboratively. To perform grid computing within reasonable communication and computation resources, this paper introduces a scalable data management approach and its framework. This paper presents a design and development of the framework which performs the scalable data management approach. For development of this framework, this paper uses a hierarchical and modular object-oriented technology supported by a system modeling theory and specification. The framework provides a superior solution to reduce data traffic loads and local computation demands with a practical application.

This paper describes a design and development of an intelligence balancing for communication data management in a grid computing. The intelligence balancing allows to execute a complex large-scale grid computing system and share dispersed data assets collaboratively. After a brief review of communication data management, we discuss the design issue of the intelligence balancing. We analyze system performance and scalability, especially with non-intelligence and intelligence-based configuration. The empirical result on the heterogeneous OS distributed system indicates the intelligence balancing is effective and scalable due to the use of communication data reduction.