• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.1
• /
• pp.19-24
• /
• 2021

In this paper, we study performance factor of machine learning in the distributed environment using Apache Spark and presents an efficient distributed processing method through experiments. This work firstly presents performance factor when performing machine learning in a distributed cluster by classifying cluster performance, data size, and configuration of spark engine. In addition, performance study of regression analysis using Spark MLlib running on the Hadoop cluster is performed while changing the configuration of the node and the Spark Executor. As a result of the experiment, it was confirmed that the effective number of executors was affected by the number of data blocks, but depending on the cluster size, the maximum and minimum values were limited by the number of cores and the number of worker nodes, respectively.

• Electronics and Telecommunications Trends
• /
• v.28 no.6
• /
• pp.1-12
• /
• 2013

미래 데이터 트래픽은 유선에서 무선으로 트래픽의 비중이 증가하고 있으며 데이터 트래픽의 규모도 급격히 증가될 것으로 예측되고 있다. 데이터 트래픽을 효과적으로 수용하기 위하여 단일한 전송 기술을 이용하는 방식 보다는 다양한 유무선 전송 기술을 기반으로 효과적으로 다양한 유무선 자원을 제어하고 상황에 따라 최적의 전송 기술을 적용함으로써 제한된 유무선 자원을 효율적으로 이용하는 기술이 필요하다. 미래 통신 네트워크는 다양한 액세스 망을 통합 관리함으로써 망구축 비용 및 운영비용을 절감하고 효과적으로 망 자원을 이용하며 새로운 요구사항을 가지는 신규 서비스에 적절히 대응하기 위하여 유연하고 지능화된 유무선 융합 네트워크 구조로 발전되어야 한다. 본 문서에서는 유무선 융합 네트워크 기술을 중심으로 지금까지 진행된 융합 기술에 대해 간단히 살펴보고 유럽을 중심으로 현재 연구되고 있는 유무선 융합 기술에 대하여 설명한다. 그리고, 2020년의 1,000배 트래픽을 수용하기 위한 5G 모바일 코어망의 요구사항에 대하여 살펴 본다.

• Journal of the Korean Magnetics Society
• /
• v.6 no.6
• /
• pp.375-381
• /
• 1996

This study is conducted to develop the high frequency inductor with ferro-yoke for the surface shape control. A 3- Dimensional simulation program based on the finite element method is developed for calculating electromagnetic field of the high frequency inductor. The validity of the program and the capability to control the liquid metal shape are confirmed by comparison between the results from calculation and those from measurement.

• Electronics and Telecommunications Trends
• /
• v.30 no.4
• /
• pp.12-25
• /
• 2015

최근 서버 관련 에너지 소비가 꾸준히 증가함에 따라 IT부문의 에너지 효율성을 높이기 위한 IT기술의 요구가 급증하고 있다. 이러한 요구에 따라 IT시스템 운용 시 사용자로 하여금 에너지 사용 상태를 정확히 인지 및 식별 가능하도록 함으로써 이를 토대로 에너지 제어를 통한 결과를 예상할 수 있도록 하며 에너지절감 유도까지 가능하게 하는 에너지 효율성 컴퓨팅 기술이 활발히 연구되고 있다. 향후에 엑사스케일 컴퓨터나 대규모 매니코어 컴퓨터 개발에 있어서 이러한 핵심기술은 선택이 아니라 필수 요소가 될 것이다.

• Journal of Wetlands Research
• /
• v.14 no.3
• /
• pp.375-384
• /
• 2012

The sedimentary processes of the Ganghwa tidal flat has been changed over 20 years because of the large-scale construction projects. The sedimentary environment of the Donggum tidal flat, located in the eastern part of Ganghwa tidal flat and in the lower reaches of Yeomha channel, was affected by changes the tidal current regime and estuarine circulation. These resulted an occurrence of rapid deposition in the tidal flat. The silt-clay laminated silt facies in the upper parts of two core sediments suggested that deposition had been relatively high in the tidal flat. The sedimentation rates from the cores using $^{210}Pb$ analysis were 3.25cm/year(st. 3) and 3.47cm/year(st. 5). However the short-term sediment accumulation rates from 2010 to 2012 were mostly less than 1cm/year, indicated that the sediments deposited relatively low rates. As a result, the sediment in the Donggum tidal flat rapidly accumulated during 2000s due to constructions of man-made structures. Recently, the increase of elevation in the tidal flat resulted to show relatively low sedimentation rate with seasonal variations.

• KIPS Transactions on Computer and Communication Systems
• /
• v.7 no.3
• /
• pp.67-72
• /
• 2018

Recently, noise reduction in an axial fan producing the small pressure rise and large flow rate, which is one type of turbomachine, is recognized as essential. This study describes the design and optimization techniques of MPI parallel program to simulate the flow-induced noise in the axial fan. In order to simulate the code using 100 million number of grids for flow and 70,000 points for noise sources, we parallelize it using the 2D domain decomposition. However, when it is involved many computing cores, it is getting slower because of MPI communication overhead among nodes, especially for the noise simulation. Thus, it is adopted the one-sided communication to reduce the overhead of MPI communication. Moreover, the allocated memory and communication between cores are optimized, thereby improving 2.97x compared to the original one. Finally, it is achieved 12x and 6x faster using 6,144 and 128 computing cores of KISTI Tachyon2 than using 256 and 16 computing cores for the flow and noise simulations, respectively.

• Journal of KIISE
• /
• v.43 no.2
• /
• pp.163-176
• /
• 2016

The contention overhead on the same channel in an NoC architecture can significantly increase a communication delay due to the simultaneous communication requests that occur. To reduce the overall overhead, we propose task-to-tile binding techniques for an NoC-based manycore platform, whereby it is assumed that the task mapping decision has already made. Since the NoC architecture may have multiple memory tiles as its size grows, memory clustering is used to balance the load of memory by making applications access different memory tiles. We assume that the information on the communication overhead of each application is known since it is specified in a dataflow task graph. Using this information, this paper proposes two heurisitics that perform binding of multiple tasks at once based on a proper memory clustering method. Experiments with an NoC simulator prove that the proposed heurisitic shows performance gains that are 25% greater than that of the previous binding heuristic.

• Journal of Information Technology and Architecture
• /
• v.10 no.1
• /
• pp.23-31
• /
• 2013

A rapid growth of IT convergence into different application areas draws a lot of interest in high performance platform and embedded system. Industry needs well educated computer professionals with the practical understanding on the emerging technologies and core issues of contemporary popular services. In this paper, we present an education framework for computer system architecture based on rigorous analyses of the characteristics of IT convergence services and information technology trends. The proposed framework puts emphasis on real-world and hands-on subjects related to multicore architecture, embedded system and parallel processing. We believe effective use in the development and management of computer system architecture courses encouraging both industries and students.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.32 no.6
• /
• pp.391-399
• /
• 2019

In this study, an analytical method was proposed to correct the analysis results and minimize the errors between column shortening predictions and real values in high-rise buildings. In this regard, the construction sequence analysis of 41-story reinforced concrete buildings was performed and the results were compared to four assumed field measurements that were divided into the column and the core. The analysis correction was applied at a stage over the error limit in the column and at all stages in the core. Since the error occurred continuously after the analysis was corrected, additional corrections of the analysis resulted in a smaller error. By applying the proposed analytical correction method, it was confirmed that the long-term shortening value can be accurately predicted.

• The KIPS Transactions:PartC
• /
• v.10C no.6
• /
• pp.809-814
• /
• 2003

The multicast transmitting the real-time data to groups may easily have many attacks from abnormal attacks because it has many links as compared to the unicast. The existing group key management architectures for preventing these problems are designed for protocols suitable for a large scale. Thus these architectures applied to a small scale routing protocols may have many overheads with key distribution and a constant core tree. Therefore this paper proposes a groups key management protocol for a secure multicast in PIM-SM multicast group communication. The proposed method divide multicast groups with RO(Rendezvous-Point), and subgroup key managers are established in each RP and can be transmitted groups keys between senders and receivers, so the security cannel is set up for secure data transfer, And this does not have needs of the data translation for group keys and the new key distribution for path change. As a result of this, the data transmission time can be reduced.