Myungil Kim

The resource running time manager for integrated environment

Latest scientific application and simulation are requiring powerful computing resources. To get the sufficient resources, more and more computer experts are constructing high-end supercomputer, cluster computing or cloud computing. Although most corporations are starting to use high performance computing by organizing cluster, they rarely add some extra resources dynamically when computing power is not sufficient. In addition, they are also expensive to use insufficient resources as cloud computing resources. To overcome resource limitations when using cloud computing to conduct large computations, we exploit underutilized personal desktops by organizing a heterogeneous cluster environment. An open-source based scheduler, Son of Grid Engine (SGE) is provided in the integrated environment to manage or control the extra computing resources. Two kinds of groups are classified according to the individuals degree of utilization: a full-time group that provides resources at any time, and a custom-time group that provides resources only at designated times. We developed a resource manager that can turn on/off and schedule the SGE execution daemon; this manager performs as the core engine. The resource running time manager efficiently allocates and controls the members of affiliated computing resources.

A Study on PDR Heading Angle Error Correction Algorithm Using WLAN Based Localization Information

Indoor localization is a core technology for providing ‘Seamless LBS(Location Based Service)’. Localization is divided as two parts, which are indoor localization and outdoor localization. At the outdoor localization, dominant technology is existed like a GPS. However, this technology cannot be used in the indoor due to restriction of arrival of signal. Therefore, study of indoor localization technology is needed for providing precise position information as a GPS of the outdoor. For this, recently research institution, laboratory are developed various indoor system which is using various communication technologies. Geomagnetic sensor typically is used to estimate the direction of the user by PDR. To calculate the azimuth from geomagnetic sensors, the error of estimated azimuth can occur by axis change of geomagnetic sensors. However, regardless of the position of user, azimuth information can be obtained in WLAN-based localization system. To overcome this PDR problem, this thesis introduces research for improve estimation the direction of user using WLAN-based localization information.

Industrial HPC activities in Korea

We have described current crisis of industries, especially small and medium-sized businesses (SMB). To overcome the crisis, industries must take a revolutionary step and look for modern technologies. Computer modeling and simulation has brought about dramatic innovation in engineering field. In the process of product design and performance verification, computer aided engineering with utilizing supercomputers and software proper to industrial products surely makes product cost and time be definitely reduced, thereby industries can have an opportunity to overcome the current crisis. In order to promote SMB through supercomputing, we are going to establish an industrial supercomputing service model that are composed of R&D IT infrastructure, joint development, education, and communities relationship. And we present some success stories obtained from SMB supercomputing.

Development of large-scale structural analysis system on a supercomputer

In this paper, a linear structural analysis system for the model with large scale degrees of freedom is presented. The proposed system has 3-tier architecture and three main components; geometry viewer, pre-/post-processor and supercomputing manager. To analyze large-scale problems, the ADVENTURE solid solver was adopted as a general-purpose finite element solver and the supercomputer named ‘tachyon’ was adopted as a parallel computational platform. The problem solving performance and scalability of this structural analysis system is demonstrated by illustrative examples with different sizes of degrees of freedom.

Efficient data synchronization method on integrated computing environment

To execute scientific applications and simulations of enormous scale, the computing paradigm is evolving into one of cluster computing and cloud computing that can exploit the large number of available computing resources. To maximize the utilization of them, company or research center needs a scheduler engine and its data space to construct a cluster computing environment. However, if certain data space is shared, problems related to the security of node, the network traffic imbalance between nodes, and the data protection could arise. To solve these issues, a manager synchronizing the shared data space for the nodes that constitute a cluster computing environment is designed. The synchronization manager shares data in two ways: First, under the cluster environment, the full synchronization group can mount a specific directory space of the master node via NFS. It is used for the data which can be globally referenced. Second, the partial synchronization group delivers data to assigned workers through rsync. It can be used to locally share data for the isolation. The partial synchronization group is superior to full synchronization group in security and efficiency because data are shared in separate manner. By applying adequate data-sharing method, the designed manager efficiently mediate sharing data as purposed.

Job submission and monitoring management in integrated computing environment for finite element analysis

Modeling and simulation (M&S) is a promising solution for manufacturers to verify and resolve errors in advance. Recently, M&S using high-performance computing has attracted attention as a technology trend for its efficiency in terms of the time and the cost. However, a huge amount of computing resources is still required to perform an M&S analysis workflow within a short time. For the resolution of this issue, we organize a heterogeneous cluster environment with individual nodes of the diverse performance and OS. An open source-based task scheduler named Son of Grid Engine is used for managing the heterogeneous resources for the experiment. To verify the efficiency and validity of the organization scheme, we have developed an environment for analyzing products and monitoring work through CalculiX, an open source-based finite element analysis solver.

Distributed Network Operations and Management on Future Internet Based on GMOC and DvNOC Federation

Traditional Internet has confronted quite a few problems in terms of network security, scalability, performance, etc., mainly due to the rapid increase of the number of end-users and various new service demands. Therefore, Future Internet researches have been resolving the fundamental weakness of traditional Internet. As a part of Future Internet developments, the federation and management of distributed Future Internet testbeds is becoming the central challenge to support reliable end-to-end researches. Under the fundamental requirements that Future Internet researches and experiments need to be performed over inter-domain networks, countries, and even continents, this paper presents a VFN (Virtual Federated Network) framework that federates and manages the resource information of autonomous distributed networks (ADNs). VFN environment is a common federation interface between ADNs so that the core network resource information can be federated and provided for end users. Also, we will elaborate the experiment results accomplished by the international collaborative works between GMOC (GENI Meta Operations Center) in the US and DvNOC (Distributed virtual Network Operations Center) in Korea based on the proposed federation framework.

Multimedia Network Service Environment Based on Distributed Virtual Network Management

Multimedia applications are generally based on continuous streams such as audios and videos, which are very dependent on time and space constraints. This dependency leads to the requirement of functional services that support processing and communication of the multimedia data in distributed environments. The functional services, therefore, need to acquire media-specific meta-data from the distributed environment in order to provide end-to-end services with proper Quality of Service (QoS) specifications. In this context, it is very important that distributed multimedia applications become network-aware by gathering and utilizing meta datasets from underlying network resources. DvNOC (Distributed virtual Network Operations Center) is originally devised for user-oriented virtual network management based on distributed network resources. In this paper, we have come up with the DvNOC framework adopted for multimedia applications. The proposed multimedia network service environment based on DvNOC incorporates the functionalities of meta-data gathering, sharing, abstracting, federating, etc., which is ultimately capable of having applications and/or end-users aware of the underlying network resource information associated with media-specific attributes.

Fan Simulator Using Supercomputer

A fan is the most common airflow machinery and is being used in various different industries such as for heavy machinery, home appliances and automobile. An axial is used in a wide variety of applications, ranging from small cooling fans for electronics to the giant fans used in wind tunnels. An axial fan can generate large air volume used to cool equipment, but is less efficient. A sirocco fan is an efficient device for moving air by centrifugal force and can generate high pressure. Fans that affect the performance and noise of a product are important components. It is also a time and cost consuming equipment to develop a fan through physical experiments. In order to overcome this problem, we have designed and developed the fan simulator for the fluid analysis of axial fans and sirocco fans using a supercomputer. The fan simulator supports a numerical analysis of an axial fan and a sirocco fan through the web portal. The fan simulator consists of four automated processes: fan shape design, pre-processing, solving and post-processing. The prediction of flow rate and noise based on data mining without numerical analyses is also developed for the conceptual design of a fan.