• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제10권5호
• /
• pp.385-393
• /
• 2004

현재 클러스터 시스템은 여러 분야의 문제들을 위하여 폭 넓게 이용되어지고 있으며 유용한 고성능 컴퓨팅 자원으로 인식되고 있다. 클러스터 시스템의 사용자가 늘어남에 따라 클러스터 시스템의 성능 개선 못지 않게 안정적인 운영을 유지하는 것도 중요한 상황이다. 하드웨어 예방 정비가 정상 운영을 위해서 중요한 것임에도 불구하고, 예방 점검 시간에 일반적인 클러스터 시스템을 위하여 사용될 수 있는 검사 도구는 주요 관심사가 되지 못했다. 본 논문에서는 하드웨어 예방 정비를 고려하여 클러스터 시스템을 위한 하드웨어 검사 도구를 제안한다. CTS(Cluster system Test Suite)로 명명된 클러스터 시스템 검사 도구는 메모리와 NIC를 점검하기 위한 두개의 검사 루틴을 가지고 있다. CTS를 설계시, CTS가 일반 적인 클러스터 시스템이 가지는 공통된 특징을 지원하도록 노력하였으며 검사 조건 설정에서 결과 조회까지 모든 작업은 통합 GUI 환경에서 진행될 수 있도록 하였다. 두 종류의 클러스터 시스템을 점검할 때, CTS를 사용하였고 클러스터 시스템을 관리하는데 유용한 정보가 제공됨을 확인하였다.

• 한국정보과학회논문지:정보통신
• /
• 제31권1호
• /
• pp.62-69
• /
• 2004

클러스터 구조는 Ad hoc 네트워크내의 전체 호스트로 데이타를 방송하는 경우에, 재전송 되는 메시지의 수를 감소시킨다. 이러한 클러스터 구조의 이점을 보존하기 위해 클러스터 유지방법이 이용된다. 그러나 기존의 클러스터 유지방법들은 이웃정보 파악을 위한 제어메시지 외에도 클러스터 재구성을 위한 추가적인 메시지 교환을 필요로 한다. 이로 인해 클러스터 구조의 유지에 따른 이점은 크게 약화된다. 본 논문에서는 클러스터 구조의 중첩성을 이용하여 Hello 시간에 클러스터 헤드들만 제어메시지를 broadcast 전송하고 일부 멤버 호스트들은 제어메시지의 unicast 전송을 통해 분리된 게이트웨이를 파악하는 클러스터 유지방법을 제안한다. 제안하는 방법은 클러스터 재구성이 필요할 때에도, 각 호스트간에 전송되는 제어메시지를 최소한으로 줄이기 위한 전략을 사용한다. 제안된 방법은 이 과정에서 2흡 클러스터의 정의를 파괴하지 않으며, 클러스터를 완전히 분산된 방법으로 생성한다. 본 논문에서 제안한 방법은 실험결과에 의해 LCC(1)보다 우수한 것으로 평가된다.

• 산업공학
• /
• 제25권1호
• /
• pp.127-133
• /
• 2012

This paper considers the preventive maintenance scheduling problem of the cluster tool which is one of the most important manufacturing equipments in the next-generation semiconductor production environment. We define a random process that expresses the successive amount of chemicals accumulating inside the tool. Based on the renewal theory, we find the expected value and probability distribution of the time that the amount of accumulated chemicals exceeds a predetermined level. For a given probability that the accumulated chemicals exceeds the predetermined level we present a method to obtain the number of chamber operations to perform the preventive maintenance of that chamber. In addition, a method to get the preventive maintenance schedule for the whole cluster tool is presented. A numerical example is provided to illustrate our method.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2005년도 춘계 학술대회논문집
• /
• pp.41-46
• /
• 2005

A new PC cluster was designed and constructed based on Windows XP Operating system. Primary target of the present design was the high node density per rack by using the general PC parts those are cost-effective and readily available in the market. Other major design points were system cooling and the convenient maintenance using standard PC parts. Presently 24 nodes per rack seems to be optimum considering the specification of the network switching device, system cooling and power supply, but 40 nodes can be accommodated within a single rack at maximum. Windows XP was selected as a high-performance computing environment considering the cost and the convenience in acquisition, maintenance and education. Both fast-Ethernet and Gigabit Ethernet network connection were tested and compared with previous data, especially for Linux doter using Myrinet. The result shows that there is no significant difference between the operating systems and the Fast-Ethernet and/or Gigabit Ethernet are good solution for the high-performance PC cluster considering the cost and performance.

• Molecules and Cells
• /
• 제43권2호
• /
• pp.198-202
• /
• 2020

Comprehensive inhibition of RUNX1, RUNX2, and RUNX3 led to marked cell suppression compared with inhibition of RUNX1 alone, clarifying that the RUNX family members are important for proliferation and maintenance of diverse cancers, and "cluster regulation of RUNX (CROX)" is a very effective strategy to suppress cancer cells. Recent studies reported by us and other groups suggested that wild-type RUNX1 is needed for survival and proliferation of certain types of leukemia, lung cancer, gastric cancer, etc. and for their one of metastatic target sites such as born marrow endothelial niche, suggesting that RUNX1 often functions oncogenic manners in cancer cells. In this review, we describe the significance and paradoxical requirement of RUNX1 tumor suppressor in leukemia and even solid cancers based on recent our findings such as "genetic compensation of RUNX family transcription factors (the compensation mechanism for the total level of RUNX family protein expression)", "RUNX1 inhibition-induced inhibitory effects on leukemia cells and on solid cancers through p53 activation", and "autonomous feedback loop of RUNX1-p53-CBFB in acute myeloid leukemia cells". Taken together, these findings identify a crucial role for the RUNX cluster in the maintenance and progression of cancers and suggest that modulation of the RUNX cluster using the pyrrole-imidazole polyamide gene-switch technology is a potential novel therapeutic approach to control cancers.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권10호
• /
• pp.3439-3457
• /
• 2014

Clustering is an effective method for improving the performance of large scale mobile ad hoc networks. However, when the moving speed is very fast, the topology changes quickly, which leads to frequent cluster topology updates. The drastically increasing control overheads severely threaten the throughput of the network. SCBCS (Signal Characteristic Based Cluster Scheme) is proposed as a method to potentially reduce the control overheads caused by cluster formation and maintenance in aeronautical ad hoc networks. Each node periodically broadcasts Hello packets. The Hello packets can be replaced by data packets, which preserve bandwidth. The characteristics of the received packets, such as the Doppler shift and the power of two successive Hello packets, help to calculate the relative speed and direction of motion. Then, the link connection lifetime is estimated by the relative speed and direction of motion. In the clustering formation procedure, the node with the longest estimated link connection time to its one-hop neighbors is chosen as the cluster head. In the cluster maintenance procedure, re-affiliation and re-clustering schemes are designed to keep the clusters more stable. The re-clustering phenomenon is reduced by limiting the ripple effect. Simulations have shown that SCBCS prolongs the link connection lifetime and the cluster lifetime, which can reduce the topology update overheads in highly dynamic aeronautical ad hoc networks.

• 한국도로학회논문집
• /
• 제15권5호
• /
• pp.91-99
• /
• 2013

PURPOSES : This study aims at proposing the methodology for benefit evaluations in pavement maintenance methods and timings using KoPMS(Korean Pavement Management System) software which was developed for efficient pavement management. METHODS : This study classified pavement sections into 4 clusters considering AADT(Annual Average Daily Traffic) and ESAL(Equivalent Single-Axle Load) using cluster analysis and used the deterioration models in each cluster. Increased user costs due to pavement deterioration as time goes by and agent costs for maintenance were estimated. Based on deterioration model and KoPMS software, Methodology for benefit evaluation was proposed in pavement maintenance methods and with/without implementation using real pavement section data. RESULTS : This study verified that considering agent costs only would be constrained to decide pavement maintenance methods and timings, and ascertained that decision making with agent and user costs would be effective. In addition, this study revealed that pavement maintenance methods and timings can be affected by AADT and ESAL and frequent pavement maintenances can be more efficient for benefits in pavement sections with more AADT and ESAL. Also this study found that user costs would be more affected to decision making than agent costs. Moreover, Delay of conducting pavement maintenance caused increased vehicle operating costs and environmental costs because of poor conditions of pavements. CONCLUSIONS : This study proposed LCCA and benefit estimation methodology of pavement with considering agent and user costs. The results of this study can be used for baseline data of efficient pavement asset management.

• Structural Monitoring and Maintenance
• /
• 제5권2호
• /
• pp.273-295
• /
• 2018

In this study, an investigation of a damage detection methodology for global condition assessment is presented. A particular emphasis is put on the utilization of wireless sensors for more practical, less time consuming, less expensive and safer monitoring and eventually maintenance purposes. Wireless sensors are deployed with a sensor roving technique to maintain a dense sensor field yet requiring fewer sensors. The time series analysis method called ARX models (Auto-Regressive models with eXogeneous input) for different sensor clusters is implemented for the exploration of artificially induced damage and their locations. The performance of the technique is verified by making use of the data sets acquired from a 4-span bridge-type steel structure in a controlled laboratory environment. In that, the free response vibration data of the structure for a specific sensor cluster is measured by both wired and wireless sensors and the acceleration output of each sensor is used as an input to ARX model to estimate the response of the reference channel of that cluster. Using both data types, the ARX based time series analysis method is shown to be effective for damage detection and localization along with the interpretations and conclusions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권4호
• /
• pp.1661-1678
• /
• 2016

Clustering is one of the key technologies in vehicular networks. Constructing and maintaining stable clusters is a challenging task in high mobility environments. DWCM (Distributed and Weighted Clustering based on Mobility Metrics) is proposed in this paper based on the d-hop dominating set of the network. Each vehicle is assigned a priority that describes the cluster relationship. The cluster structure is determined according to the d-hop dominating set, where the vehicles in the d-hop dominating set act as the cluster head nodes. In addition, cluster maintenance handles the cluster structure changes caused by node mobility. The rationality of the proposed algorithm is proven. Simulation results in the NS-2 and VanetMobiSim integrated environment demonstrate the performance advantages.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권2호
• /
• pp.664-677
• /
• 2014

Wind energy has proven its viability by the emergence of countless wind turbines around the world which greatly contribute to the increased electrical generating capacity of wind farm operators. These infrastructures are usually deployed in not easily accessible areas; therefore, maintenance routines should be based on a well-guided decision so as to minimize cost. To aid operators prior to the maintenance process, a condition monitoring system should be able to accurately reflect the actual state of the wind turbine and its major components in order to execute specific preventive measures using as little resources as possible. In this paper, we propose a fault detection approach which combines cluster analysis and frequent pattern mining to accurately reflect the deteriorating condition of a wind turbine and to indicate the components that need attention. Using SCADA data, we extracted operational status patterns and developed a rule repository for monitoring wind turbine systems. Results show that the proposed scheme is able to detect the deteriorating condition of a wind turbine as well as to explicitly identify faulty components.