• Proceedings of the Korean Society of Computer Information Conference
• /
• 2018.07a
• /
• pp.211-212
• /
• 2018

머신 러닝 환경에서 많은 양의 데이터를 한꺼번에 학습하게 되면 데이터 트래픽이 증가함에 따라 흐름 정체가 발생하고 학습 품질이 저하되며 학습속도 지연 등의 문제가 발생한다. 본 연구는 머신러닝 환경에서 빅 데이터 학습 데이터 분할을 위한 핵심 목표인 Greedy 알고리즘에 대해 설명하고 간단한 Greedy 알고리즘을 사용하여 각각의 데이터 파티션을 생성하여 학습 속도의 효율성을 향상시키는 방법을 제안한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.24-26
• /
• 2000

본 논문에서는 FDS를 응용한 하드웨어 분할 방법을 강 제약 조건을 만족하면서 FDS를 응용하는 방법보다 낮은 복잡도의 분할 알고리즘을 제안한다. 기존의 FDS 응용 방법은 힘값 계산에서 종속성에 의해 후위 연산이 받는 영향값을 계산하여야 하므로 이로 인한 시간 복잡도가 가중되었다. 본 논문에서는 이러한 복잡도를 저하시키기 위해 노드의 분포 그래프와 구현에 소요되는 비용, 그리고 해당 파티션에서의 실행시간 등에 의해 상대적 긴박도를 정의하여 분할을 수행하지만, 종속성 검사는 종속성 제약조건에 의한 분포그래프의 변화와 스케쥴에 대해서만 고려되며 힘값 계산에는 고려하지 않는다. 또한, 분할 단계에서 스케쥴링을 함께 고려함으로써 합성 이후에 재 스케쥴링의 부하를 경감할 수 있도록 하였다. 제안 알고리즘 결과는 ILP 결과와 비교 분석하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.10b
• /
• pp.965-968
• /
• 2000

통합설계에서 제약사항을 만족하는 최적의 시스템을 구현하기 위해 시스템을 기술하는 각 부분을 하드웨어부와 소프트웨어부로 나누어 매핑의 권역을 찾는 분할은 중요한 문제이다. 기존의 분할 알고리즘들[1]은 파티션과 스케줄링을 2단계로 분리하여 분할 단계의 결과를 스케줄링하는 과정에 의해 진행되었다. 이러한 작업과정은 스케줄링 결과 스케줄이 불가능한 경우 시스템을 재설계 해야하는 문제점을 가진다. 본 논문에서는 분할 단계에서 스케줄링을 함께 고려하는 낮은 복잡도의 알고리즘을 제안한다. FDS를 응용한 기존 논문[4]이 고려하지 못한 자원제약에 의한 힘값 변이를 고려할 수 있도록 하였고 알고리즘 복잡도를 개선하기 위하여 종속성 제약 조건에 의해 받는 다른 노드의 힘값 계산 방법을 수정하였다. 수정된 계산 방법에서는 특정 노드와 경쟁 노드들의 제어구간별 상대적 스케줄 요구값의 크기에 의해 분할 대상 노드를 선택하게 된다. 제안된 논문의 실험결과는 시스템 제약시간을 만족하면서 구현비용을 저하시키고 알고리즘 실행시간 측면에서 효과적임을 보인다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.8
• /
• pp.1776-1784
• /
• 2003

In this paper, we propose the architecture of stream producer for MPEG­4 Video encoding. This module receives the quantized coefficient from DCT and Quantization module in macroblock unit and performs the VLC coding according to the encoding mode, and supports the error concealment mode of MPEG­4 and data partitioning mode. Using the VHDL, we designed the module using this architecture and performed the evaluations of this module by performing the post­-simulation.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.24 no.11
• /
• pp.883-889
• /
• 2014

Mufflers have been widely used in the exhaust system to reduce the noise. However, installing muffler may deteriorate the efficiency due to the increase of back pressure. Mufflers usually consist of partition plates and perforated holes in a expansion chamber. In this paper, the influences of the location of the partition and hole on the acoustic TL and back pressure were examined. The acoustic TL was predicted using virtual lab commercial software, while the back pressure were predicted using CFX commercial software. The results were used to set up a database for finding their trends. The optimal muffler model in user-interested frequency range could be selected by analyzing this database.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.119-122
• /
• 2010

본 논문에서는 Xen 가상 머신에 연성 실시간 기능을 제공하기 위한 방안을 제시한다. 현재 실시간 하이퍼바이저(hypervisor)는 대부분 경성 실시간(hard real-time) 보장을 위하여 자원을 파티션(partition)하고 고정된 태스크에 대한 수행을 지원한다. 본 연구에서는 오픈 환경을 위한 조합적 실시간 프레임워크에 기반한 계층형 실시간 스케줄링 시스템을 제안하고 있다. 제안하는 시스템은 오픈 환경에서 동적으로 태스크 및 컴포넌트의 유입이 가능하게 설계되었다. 제안하는 시스템에 대한 사전 연구 결과로써 Xen 플랫폼의 오버헤드 분석 결과를 제시하고 있다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.4B
• /
• pp.646-680
• /
• 2000

In this paper, we discuss the design and implementation of a concurrency control manager for a main memory DBMS(MMDBMS). Since an MMDBMS, unlike a disk-based DBMS, performs all of data update or retrieval operations by accessing main memory only, the portion of the cost for concurrency control in the total cost for a data update or retrieval is fairly high. Thus, the development of an efficient concurrency control manager highly accelerates the performance of the entire system. Our concurrency control manager employs the 2-phase locking protocol, and has the following characteristics. First, it adapts the partition, an allocation unit of main memory, as a locking granule, and thus, effectively adjusts the trade-off between the system concurrency and locking cost through the analysis of applications. Second, it enjoys low locking costs by maintaining the lock information directly in the partition itself. Third, it provides the latch as a mechanism for physical consistency of system data. Our latch supports both of the shared and exclusive modes, and maximizes the CPU utilization by combining the Bakery algorithm and Unix semaphore facility. Fourth, for solving the deadlock problem, it periodically examines whether a system is in a deadlock state using lock waiting information. In addition, we discuss various issues arising in development such as mutual exclusion of a transaction table, mutual exclusion of indexes and system catalogs, and realtime application supports.

• Journal of KIISE
• /
• v.45 no.1
• /
• pp.53-60
• /
• 2018

This study describes a spatial web object retrieval system using Spark, an in - memory based distributed processing system. Development of social networks has created massive amounts of spatial web objects, and retrieval and analysis of data is difficult by using exist spatial web object retrieval systems. Recently, development of distributed processing systems supports the ability to analyze and retrieve large amounts of data quickly. Therefore, a method is promoted to search a large-capacity spatial web object by using the distributed processing system. Data is processed in block units, and one of these blocks is converted to RDD and processed in Spark. Regarding the discussed method, we propose a system in which each RDD consists of spatial web object index for the included data, dividing the entire spatial region into non-overlapping spatial regions, and allocating one divided region to one RDD. We propose a system that can efficiently use the distributed processing system by dividing space and increasing efficiency of searching the divided space. Additionally by comparing QP-tree with R-tree, we confirm that the proposed system is better for searching the spatial web objects; QP-tree builds index with both spatial and words information while R-tree build index only with spatial information.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.3
• /
• pp.551-560
• /
• 2011

Recently, There has been increasing concern about WLAN-based indoor positioning system. Most of the existing WLAN-based positioning systems use a fingerprinting method as a main approach. In the fingerprinting approach, the accuracy of the location of a mobile objects is proportional to the number of reference points. However, depending on the increasing number of reference points in the training phase, it requires more time and effort to create fingerprint database. To solve these problems, we propose the new indoor positioning algorithm that calculate the distance between a mobile objects and an AP using the information of surrounding environment WLAN based APs and applied the particle filter to the proposed algorithm in order to improve the accuracy of the estimated location in this paper. To implement this algorithm, at first environmental information database such as wall, iron door, glass door, partition etc. existing in the periphery of the AP should be established. The positioning use attenuation model and path loss model. Our experimental results with proposed algorithm are verified that the positioning accuracy was low but solved the problems with fingerprinting, compared with other positioning algorithms.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.4
• /
• pp.147-152
• /
• 2016

One of major issues in machine learning that extracts and acquires knowledge implicit in data is to find an appropriate way of representing it. Knowledge can be represented by a number of structures such as networks, trees, lists, and rules. The differences among these exist not only in their structures but also in effectiveness of the models for their problem solving capability. In this paper, we propose partition utility as a criterion function for clustering that can lead to simplification of the model and thus avoid overfitting problem. In addition, a heuristic is proposed as a way to construct balanced hierarchical models.