• 한국통신학회논문지
• /
• 제29권12B호
• /
• pp.1037-1041
• /
• 2004

본 논문에서 위성신호를 매개로 원격지의 로컬클럭이 기준신호에 동기되는 새로운 개념의 슬레이브를럭 시스템을 제안하였으며 이 방식에 의한 실용화 가능성을 확인하였다. 새로이 제시하는 방식은 단계적인 물리계층에 의해 동기되던 기존의 방식에 비해 모든 슬레이브 국소들이 동일한 계위의 품질로 유지될 수 있는 등의 여러 가지 구조적인 장점을 지니고 있다. 슬레이브클럭 시스템의 측정결과 10-12 수준의 주파수정확도를 유지하였으며 ITU-T의 권고(G.811)를 만족하는 MTIE 특성을 보여주었다. 현재 전체적으로 자동화기능을 갖는 초기모델이 구현되었으며 가까운 시일 내에 상용화연구를 통해 통신망 동기용 노드클릭으로 사용될 수 있으리라 기대한다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.176-179
• /
• 2002

In recent years, an increasing amount of computer network research has focused on the problem of cluster system in order to achieve higher performance and lower cost. The load unbalance is the major defect that reduces performance of a cluster system that uses parallel program in a form of SPMD (Single Program Multiple Data). Also, the load unbalance is a problem of MPP (Massive Parallel Processors), and distributed system. The cluster system is a loosely-coupled distributed system, therefore, it has higher communication overhead than MPP. Dynamic load balancing can solve the load unbalance problem of cluster system and reduce its communication cost. The cluster systems considered in this paper consist of P heterogeneous nodes connected by a switch-based network. The master node can predict the average execution time of tasks for each slave node based on the information from the corresponding slave node. Then, the master node redistributes remaining tasks to each node considering the predicted execution time and the communication overhead for task migration. The proposed dynamic load balancing uses execution time prediction to optimize the task redistribution. The various performance factors such as node number, task number, and communication cost are considered to improve the performance of cluster system. From the simulation results, we verified the effectiveness of the proposed dynamic load balancing algorithm.

• 한국철도학회논문집
• /
• 제10권1호
• /
• pp.88-95
• /
• 2007

TCN(Train Communication Network) adopts the master/slave protocol to implement real-time communication. In this network, a fault on the master node, cased by either hardware or software failure, makes the entire communication impossible over TCN. To reduce fault detection and recovery time, this paper propose the contention based mastership transfer algorithm. Slave nodes detect the fault of master node and search next master node using the proposed algorithm. This paper also shows the implementation results of a SoC-based Fault-Tolerant MVB Controller(FT-MVBC) which includes the fault-detect-logic as well as the MVB network logic to verify this algorithm.

• 한국산업정보학회논문지
• /
• 제20권1호
• /
• pp.27-35
• /
• 2015

본 논문에서는 저속이지만 다수 노드의 연결이 요구되는 산업용 제어통신망, 홈네트워크 등에서 널리 사용되는 RS485에 적용할 수 있는 CSMA/CD 기법을 제안하고, 기존의 Master-Slave 프로토콜인 Modbus와 성능을 비교하였다. 모의실험을 위해 실제 산업용 제어통신망을 모델링 하였고, 모의실험을 통하여 여러 전송속도에서 평균 전송 지연시간과 처리량이 크게 개선됨을 보였다.

• 정보통신설비학회논문지
• /
• 제3권2호
• /
• pp.11-17
• /
• 2004

In this paper we propose a new conceptual system for a network clock in which all node clocks are simultaneously synchronized to the national standard by intermediation parameter of satellite time. Experiments have shown the possibility of its adoption by real networks. The new proposed method has various structural benefits, in particular all node clocks can be kept at the same hierarchical quality in contrast to the existing method. The measurement results show that the accuracy of the experimental slave clock system can be kept within a few parts In 1012 and the MTIE (Maximum Time Interval Error) sufficiently meets ITU-T G.811 for the primary reference clock. A prototype system with fully automatic operational functions has been realized at present and is expected to be directly used for communication network synchronization in the near future.

• 제어로봇시스템학회논문지
• /
• 제20권11호
• /
• pp.1131-1135
• /
• 2014

This paper proposes the enhanced faulty node detection scheme with hybrid algorithm using Markov-chain model on BCH (Bose-Chaudhuri-Hocquenghem) code in naval distributed control networked systems. The probabilistic model-driven approach, on Markov-chain model, in this paper uses the faulty weighting interval factors, which are based on the BCH code. In this scheme, the master node examines each slave-nodes continuously using three defined states : Good, Warning, Bad-state. These states change using the probabilistic calculation method. This method can improve the performance of detecting the faulty state node more efficiently. Simulation results show that the proposed method can improve the accuracy in faulty node detection scheme for real-time naval distributed control networked systems.

• 스마트미디어저널
• /
• 제6권3호
• /
• pp.68-74
• /
• 2017

빅 데이터 처리는 파일이나 이미지, 동영상 등 다양한 형태의 데이터를 처리하여 문제를 해결하고 통찰력 있는 유용한 정보를 제공한다. 현재 빅 데이터 처리를 위해 다양한 플랫폼이 사용되지만, 하둡이 가지는 단순성, 생산성, 확장성, 그리고 내고장성 때문에 많은 기관, 기업에서 빅 데이터 처리에 하둡을 사용하고 있다. 또한, 하둡은 다양한 하드웨어 플랫폼으로 클러스터를 구축할 수 있으며, 네임 노드(Master)와 데이터 노드(Slave)로 구분하여 빅 데이터를 처리한다. 본 논문에서는 실제 기관과 기업에서 사용하는 완전분산모드를 사용하였으며 원활한 테스트를 위해 저전력이고 저가인 싱글 보드를 사용하여 하둡 클러스터를 구축하였다. 네임 노드의 성능 영향 분석은 싱글 보드와 랩톱을 네임 노드로 사용하여 같은 데이터 처리를 통하여 비교하였으며 데이터 노드의 개수에 따른 영향 분석은 싱글 보드를 기존 클러스터의 개수에서 2배까지 늘려가며 데이터 노드가 미치는 영향을 분석하였다.

• 정보처리학회논문지A
• /
• 제8A권2호
• /
• pp.117-124
• /
• 2001

At NOW (Networks Of Workstations), the load sharing is very important role for improving the performance. The known load sharing strategy is fixed-granularity, variable-granularity and adaptive-granularity. The variable-granularity algorithm is sensitive to the various parameters. But Send algorithm, which implements the fixed-granularity strategy, is robust to task granularity. And the performance difference between Send and variable-granularity algorithm is not substantial. But, in Send algorithm, the computing time and the communication time are not overlapped. Therefore, long latency time at the network has influence on the execution time of the parallel program. In this paper, we propose the preSend algorithm. In the preSend algorithm, the master node can send the data to the slave nodes in advance without the waiting for partial results from the slaves. As the master node sent the next data to the slaves in advance, the slave nodes can process the data without the idle time. As stated above, the preSend algorithm can overlap the computing time and the communication time. Therefore we reduce the influence of the long latency time at the network and the execution time of the parallel program on the NOW. To compare the execution time of two algorithms, we use the $320{\times}320$ matrix multiplication. The comparison results of execution times show that the preSend algorithm has the shorter execution time than the Send algorithm.

• 한국정밀공학회지
• /
• 제20권12호
• /
• pp.148-158
• /
• 2003

In this paper, efficient and accurate contact search algorithm is proposed for the contact problems by the finite element method. A slave node and a maser contact segment is defined using the side of a finite element on the contact surface. The specific goal is to develop techniques of reducing the nonsmoothness of the contact interactions arising from the finite element discretization of the contact surface. Contact detection is accomplished by monitoring the territory of the slave nodes throughout the calculation for possible penetration of a master surface. To establish the validity of the proposed algorithm, some different process and geometries examples were simulated. Efforts are focused on the error rate that is based on the penetrated area through the simulations fur large deformation with contact surface between deformable bodies. A proposed algorithm offers improvements in contact detection from the simulation results.

• Journal of Communications and Networks
• /
• 제7권3호
• /
• pp.284-293
• /
• 2005

A critical issue in applications involving networks of wireless sensors is their ability to synchronize, and mitigate the fading propagation channel effects. Especially when distributed 'slave' sensors (nodes) reach-back to communicate with the 'master' sensor (gateway), low power cooperative schemes are well motivated. Viewing each node as an antenna element in a multi-input multi-output (MIMO) multi-antenna system, we design pilot patterns to estimate the multiple carrier frequency offsets (CFO), and the multiple channels corresponding to each node-gateway link. Our novel pilot scheme consists of non-zero pilot symbols along with zeros, which separate nodes in a time division multiple access (TDMA) fashion, and lead to low complexity schemes because CFO and channel estimators per node are decoupled. The resulting training algorithm is not only suitable for wireless sensor networks, but also for synchronization and channel estimation of single- and multi-carrier MIMO systems. We investigate the performance of our estimators analytically, and with simulations.