• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.22 no.6
• /
• pp.648-653
• /
• 2011

In this paper, coordinated space division multiple access(SDMA) technology is proposed to mitigate inter-cell interference by using partial channel state information in cooperative wireless communications system with limited feedback. Each AT selects an optimal cluster transmission mode and sends it back to a cluster scheduler, and at the cluster scheduler, ATs are scheduled within a AT group with the identical cluster transmission mode, and the optimal transmission mode and the corresponding scheduled ATs are determined to maximize scheduling priority. Also, in order to enhance multiuser diversity gain, an extended transmission feedback method is proposed to feed back multiple preferred cluster transmission modes at each AT. It is shown that the proposed coordinated SDMA scheme outperforms existing non-coordinated SDMA schemes in terms of the average system throughput.

• Journal of Korean Society for Geospatial Information Science
• /
• v.25 no.1
• /
• pp.29-36
• /
• 2017

In this paper, we implement a spatial big data analysis prototype based on Spark which is an in-memory system and compares the performance by the spatial split algorithm on this basis. In cluster computing environments, big data is divided into blocks of a certain size order to balance the computing load of big data. Existing research showed that in the case of the Hadoop based spatial big data system, the split method by spatial is more effective than the general sequential split method. Hadoop based spatial data system stores raw data as it is in spatial-divided blocks. However, in the proposed Spark-based spatial analysis system, there is a difference that spatial data is converted into a memory data structure and stored in a spatial block for search efficiency. Therefore, in this paper, we propose an in-memory spatial big data prototype and a spatial split block storage method. Also, we compare the performance of existing spatial split algorithms in the proposed prototype. We presented an appropriate spatial split strategy with the Spark based big data system. In the experiment, we compared the query execution time of the spatial split algorithm, and confirmed that the BSP algorithm shows the best performance.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.103-105
• /
• 2002

이 논문에서는, PC 클러스터 환경에서 질의 확장을 사용하는 정보 검색 시스템 (IR)을 설계하고 구현한 내용을 기술한다. 이 정도 검색 시스템은 문서 집합을 저장하고, 문서 집합은 역색인 파인 (IIF)로 색인되고, 랭킹 방법으로 벡터 모델을 사실하며, 질의 확장 방법으로 코사인 유사도를 사용한다. 질의 확장이란 사용자가 준 원래의 질의에 연관된 단어를 추가하여 검색 효율을 향상시키는 것이다. 여기서 제안하는 병렬 정보 검색 시스템에서는 역색인 과일은 여러 개로 분활되는데 lexical 분할 방법과 greedy 분할 방법을 사용한다. 사용자의 질의가 들어오면 질의확장을 하여 여러 개의 단어로 이루어진 확장된 질의가 만들어 지는데 이 확장된 질의를 구성하는 단어들은 각 단어와 연관된 IIF를 가지고 있는 노드에 보내어져서 병렬로 처리된다. 실험을 통하여 병렬 IR 시스템의 성능이 질의 확장과 IIF의 두 가지 분한 방법에 의해 어떻게 영향을 받는지 보인다. 실험에는 표준 한국어 테스트 말뭉치인 EKSET과 KTSET을 사용하였다. 실험에 따르면 greedy 분활 방법이 lexical 분할 방법에 비해 20%정도의 성능 향상을 보였다.

• Annual Conference of KIPS
• /
• 2015.10a
• /
• pp.204-207
• /
• 2015

이 연구는 공학 및 실험과학에서 활용되는 몬테 칼로 시뮬레이션 기법 중 하나인 무작위 행보 알고리즘의 성능 개선을 목표로 하였다. 이를 위해 무작위 행보 과정에서 난수 발생부와 행보 진행부를 분리하여 처리 시간을 단축하는 방안과, 문제 영역의 계산 규모를 2단계로 분할하여 시뮬레이션의 수렴 속도를 향상 시키는 방안을 제안한다. 또한 대규모 문제를 병렬처리 가능하도록 구현하고, 서로 다른 작업 분할 방식을 혼합하여 최적화를 수행 하였다. 순차 알고리즘만으로 실험한 결과 단순 구현방법과 비교해 실행시간과 에너지 소모량이 각각 18%의 성능향상을 얻었으며, 병렬 알고리즘을 8개의 노드(16코어)의 클러스터에서 실행했을 때 행 분할 방식의 성능이 블록 분할 방식보다 8% 빨라지는 것을 확인하였다.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.25 no.3
• /
• pp.357-364
• /
• 2014

In this work, we construct a CPU cluster to implement a parallel finite-difference time domain(FDTD) algorithm for fast electromagnetic analyses. This parallel FDTD algorithm can reduce the computational time significantly and also analyze electrically larger structures, compared to a single FDTD counterpart. The parallel FDTD algorithm needs communication between neighboring processors, which is performed by the MPI(Message Passing Interface) library and a 3-D domain decomposition is employed to decrease the communication time between neighboring processors. Compared to a single-processor FDTD, the speed up factor of a-CPU-cluster-based parallel FDTD algorithm is investigated for the normal mode and the hypermode and finally analyze an electrically large concrete structure by the developed parallel algorithm.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.16 no.5
• /
• pp.608-615
• /
• 2006

This paper proposes a fuzzy model tree consisting of local linear models using fuzzy cluster for data modeling. First, cluster centers are calculated by fuzzy clustering method using all input and output attributes. And then, linear models are constructed at internal nodes with fuzzy membership values between centers and input attributes. The expansion of internal node is determined by comparing errors calculated in parent node with ones in child node, respectively. As a final step, data prediction is performed with a linear model having the highest fuzzy membership value between input attributes and cluster centers in leaf nodes. To show the effectiveness of the proposed method, we have applied our method to various dataset. Under various experiments, our proposed method shows better performance than conventional model tree and artificial neural networks.

• The KIPS Transactions:PartD
• /
• v.10D no.4
• /
• pp.591-602
• /
• 2003

Delay and discontinuance phenomenon of service are cause by sudden increase of the network communication amount and the quantity consumed of resources when Internet users are driven excessively to a conventional single large database sewer. To solve these problems, spatial database cluster consisted of several single nodes on high-speed network to offer high-performance is risen. But, research about spatial join operation that can reduce the performance of whole system in case process at single node is not achieved. So, in this paper, we propose efficient parallel spatial join processing method in a spatial database cluster system that uses data partitions and replications method that considers the characteristics of space data. Since proposed method does not need the creation step and the assignment step of tasks, and does not occur additional message transmission between cluster nodes that appear in existent parallel spatial join method, it shows performance improvement of 23% than the conventional parallel R-tree spatial join for a shared-nothing architecture about expensive spatial join queries. Also, It can minimize the response time to user because it removes redundant refinement operation at each cluster node.

• Journal of the Korean Geographical Society
• /
• v.47 no.2
• /
• pp.226-244
• /
• 2012

This paper aims to analyze the process of the evolution of Gumi electronics industrial cluster and to understand the role of governments for local industrial dynamics. Gumi was a typical satellite platform type new industrial district up to mid-1990s. At that time, Gumi industrial park was the agglomeration of branch plants headquartered in Capital Region with weak local linkages. During the last two decades, however, Gumi has evolved to an electronics industrial cluster with considerable local interfirm linkages and innovation activities of SMEs. Recognizing government industrial policies is critical in understanding the process of the evolution of Gumi electronics cluster. At the early stage, the state was the developer and locator of business activities within the confines of the Gumi industrial park. In recent years, central government's innovative cluster policy contributed to strengthening networks among firms, universities, and research centers to form local innovation networks as well as networks between large branch plants and SMEs. Gumi city and Gyungsangbuk-do promoted innovative activities of SMEs through the supports of cooperative networks between universities and SMEs. The increasing roles of SMEs and local governments in addition to the large branch plants and the central government have become the basis of the evolution of industrial cluster in Gumi.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.38 no.1
• /
• pp.79-85
• /
• 2001

In this paper, we propose a new technology mapping algorithm for CPLD consider area under time constraint(TMFCPLD). This technology mapping algorithm detect feedbacks from boolean networks, then variables that have feedback are replaced to temporary variables. Creating the temporary variables transform sequential circuit to combinational circuit. The transformed circuits are represented to DAG. After traversing all nodes in DAG, the nodes that have output edges more than two are replicated and reconstructed to fanout free tree. This method is for reason to reduce area and improve total run time of circuits by TEMPLA proposed previously. Using time constraints and delay time of device, the number of graph partitionable multi-level is decided. Initial cost of each node are the number of OR-terms that it have. Among mappable clusters, clusters of which the number of multi-level is least is selected, and the graph is partitioned. Several nodes in partitioned clusters are merged by collapsing, and are fitted to the number of OR-terms in a given CLB by bin packing. Proposed algorithm have been applied to MCNC logic synthesis benchmark circuits, and have reduced the number of CLBs by 62.2% than those of DDMAP. And reduced the number of CLBs by 17.6% than those of TEMPLA, and reduced the number of CLBs by 4.7% than those of TMCPLD. This results will give much efficiency to technology mapping for CPLDs.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.679-688
• /
• 2021

Gait is an evaluation index used in various clinical area including brain nervous system diseases. Signal source localizing and time-frequency analysis are mainly used after extracting independent components for Electroencephalogram data as a method of measuring and analyzing brain activation related to gait. Existing treadmill-based walking EEG analysis performs signal preprocessing, independent component analysis(ICA), and source localizing by merging data after the multiple EEG measurements, and extracts representative component clusters through inter-subject clustering. In this study we propose an analysis method, without merging to single dataset, that performs signal preprocessing, ICA, and source localization on each measurements, and inter-subject clustering is conducted for ICs extracted from all subjects. The effect of data merging on the IC clustering and time-frequency analysis was investigated for the proposed method and two conventional methods. As a result, it was confirmed that a more subdivided gait-related brain signal component was derived from the proposed "non-merging" method (4 clusters) despite the small number of subjects, than conventional method (2 clusters).