• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.5
• /
• pp.409-416
• /
• 2000

In the shared-memory mutiprocessor systems, shared processing techniques such as time-sharing, space¬sharing, and gang-scheduling are used to improve the overall system utilization for the parallel operations. Recently, LLPC(Loop-Level Process Control) allocation technique was proposed. It dynamically adjusts the needed number of processors for the execution of the parallel code portions based on the current system load in the given job. This method allocates as many available processors as possible, and does not save any processors for the parallel sections of other later-arriving applications. To solve this problem, in this paper, we propose a new processor allocation technique called FPA(Fuzzy Processor Allocation) that dynamically adjusts the number of processors by fuzzifYing the amounts ofueeded number of processors, loads, and estimated execution times of job. The proposed method provides the maximum possibility of the parallism of each job without system overload. We compare the performances of our approaches with the conventional results. The experiments show that the proposed method provides a better performance.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.39 no.8
• /
• pp.703-711
• /
• 2015

In large-scale factories, there are usually openings in the building to increase the efficiency of workers. However, if the factory is heated during winter, openings significantly increase the heating load. Therefore, there is a need for air curtains to be installed at the top of openings in factories to reduce the heating load due to the cold air entering from outside. The main design variables of these air curtains are the discharge angle, speed, and temperature, etc. While there have already been many studies focusing on these design variables, the distance from the opening and the width of the discharge have not been studied even though they also affect the sealing performance. As a result, when the distance from the opening decreases and the width of the discharge increases, we realize an optimum air curtain performance. However, if the distance from the opening is about 1.5 m, by adjusting the discharge angle and the distance from the opening, the sealing performance of the air curtain is improved by 13.7％.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.34 no.6
• /
• pp.643-648
• /
• 2010

Portable fuel cells are commonly operated in the dead-end mode because of such as high fuel utilization. However, the performance of such systems deteriorates continuously with an increase in the amount of by-products such as water vapor and nitrogen. In this study, to verify the effect of water vapor on Proton Exchange Membrane Fuel Cells (PEMFCs), constant-load experiments were carried out for a current density of 600 mA/cm2 and a voltage of 0.4 V, respectively. The performance of the cell was more stable under constant voltage conditions than under constant current density conditions. Condensed water accumulated in the anode channel near the cell outlet. The experimental results show how the relative humidity (RH = 0.15, 0.4 and 0.75) of air at the cathode side affect the performance of PEMFCs with dead-end anode. At RH values higher than 0.15, the mean power density increased by up to 51% and the mean purge duration decreased by up to 25% compared to the corresponding initial values.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.5
• /
• pp.217-222
• /
• 2013

Return-Oriented Programming(ROP) is an advanced code-reuse attack like a return-to-libc attack. ROP attacks combine gadgets in program code area and make functions like a Turing-complete language. Some of previous defense methods against ROP attacks show high performance overhead because of dynamic execution flow analysis and can defend against only certain types of ROP attacks. In this paper, we propose Indirect Branch Target Address Verification (IBTAV). IBTAV detects ROP attacks by checking if target addresses of indirect branches are valid. IBTAV can defends against almost all ROP attacks because it verifies a target address of every indirect branch instruction. Since IBTAV does not require dynamic execution flow analysis, the performance overhead of IBTAV is relatively low. Our evaluation of IBTAV on SPEC CPU 2006 shows less than 15% performance overhead.

• Journal of Advanced Navigation Technology
• /
• v.14 no.1
• /
• pp.93-101
• /
• 2010

In this study, the performance characteristics of turbocharged engine is analyzed. The methods of engine performance improvements are suggested not only for full load characteristics of the engine but also for partial load characteristics of the engine, which is more frequently used in actual driving conditions. The compression ratio of the compressor is increased rapidly in a straight line pattern until 1260 engine rpm, and after that it is increased slowly to 2.5 ratio. Also the brake mean effective pressure increased until 1260 engine rpm and decreased rapidly after 1600 engine rpm. The higher the pressure ratio, the better the fuel consumption, air excess ratio and brake mean effective pressure. But those are higher in the rated revolution range than in the mid-low revolution range. The turbocharger is operated in a stable condition from 1260 rpm and its efficiency is low in the low speed range for the reason of its characteristics. The results of this study can be applied in the fundamental control methods of turbocharged engine for stable load and speed.

• Journal of Software Engineering Society
• /
• v.13 no.3
• /
• pp.18-30
• /
• 2000

기존 워크플로우 시스템은 비즈니스 프로세스 내의 정보와 제어의 흐름 수행에 있어서 중앙 집중적인 워크플로우 엔진이 전체 실행의 모든 부분을 담당하고 있다. 이러한 워크플로우 시스템이 갖는 성능, 확장성 등 구조적 제약의 단점을 극복하기 위하여 워크플로우 엔진의 분산, 이동 에이전트의 적용, 웹 기반 등 다양한 연구가 진행되고 있다. 본 논문에서는 단위 업무 스케쥴링 기능과 관련된 정보를 분산된 형태로 참가자에게 일임하여 워크플로우 엔진 부하를 줄이고, 프로세스 인스턴스 단위의 이동 에이전트를 적용하여 비동기적 특성을 이용한 성능 및 확장성의 향상을 제공하는 워크플로우 시스템의 설계 및 구현을 소개한다. 또한 제안한 시스템을 위하여 각 참가자별 스케쥴링 정보를 WfMC의 프로세스 정의 표준 언어인 WPDL로부터 추출하여 프로세스 테이블로 구성하는 방법을 제안한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.552-554
• /
• 2000

기존 워크플로우 시스템은 비즈니스 프로세스 내의 정보와 제어의 흐름 수행에 있어서 중앙 집중적인 워크플로우 엔진이 전체 실행의 모든 부분을 담당하고 있다. 이러한 워크플로우 시스템이 갖는 성능, 확장성 등 구조적 제약의 단점을 극복하기 위하여 워크플로우 엔진의 분산, 이동 에이전트의 적용, 웹 기반 등 다양한 연구가 진행되고 있다. 본 논문에서는 단위 업무 스케쥴링 기능과 관련된 정보를 분산된 형태로 참가자에게 일임하여 워크플로우 엔진 부하를 줄이고, 프로세스 인스턴스 단위의 이동 에이전트를 적용하여 비동기적 특성을 이용한 성능 및 확장성의 향상을 제공하는 워크플로우 시스템의 설계를 제안한다. 또한 제안한 시스템을 위하여 각 참가자별 스케쥴링 정보를 WfMC의 프로세스 정의 표준 언어인 WPDL로부터 추출하는 방법을 제안한다.

• The Journal of the Korea Contents Association
• /
• v.7 no.1
• /
• pp.32-39
• /
• 2007

An important application domain for online services is an interactive, multi-player game. In recent, many increase of users that use on-line services through networks have caused a heavy load to the server. In this paper, we propose a MMORPG(Massively Multi-player Online Role Playing Game) distributed game server using flayer-Cell. Our method provides efficient solution of a MMORPG distributed game server for large numbers of users. It is shown through the experiments that our method outperforms existing methods in terms of memory utilization rate and processing speed.

• The Transactions of the Korea Information Processing Society
• /
• v.2 no.2
• /
• pp.156-167
• /
• 1995

VDFS(V-kernel Distributed File System) constructed in this paper proposes an asymmetric Distributed File System so that various problems of existing symmetric structures like NFS ＆ RFS may be solved in small-sized system. Particulary, in this paper, server workload can be reduced by using a caching mechanism and remote access transparency can be improved by using a sharing file duplicately.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04a
• /
• pp.127-129
• /
• 2000

1990년대 초반에 등장한 인터넷을 기반으로 하는 서비스 중에 하나인 웹은 대중적인 인기를 확보하여 사용량이 기하급수적으로 증가하고 있다. 인터넷 사용자와 서비스 제공자의 증가에 따라 같은 객체에 대한 중복요청이 네트워크 대역폭의 상당 부분을 차지하여 불필요하게 낭비되며 일부 인기있는 서버로 부하가 집중되어 응답시간이 길어지게 된다. 이와 같은 문제를 해결하기 위해 인터넷 캐슁기법이 도입되어 연구되어 왔다. 사용자의 지연시간 및 네트워크 전송용량이 사용효율은 어떻게 캐슁을 설계하고 운영하는가에 따라 많은 영향을 받게 된다. 본 논문에서는 인터넷 캐슁을 위해 설계된 기존의 전략들을 살펴보고 우리의 새로운 알고리즘을 제안한다. 그리고, 각 알고리즘의 성능을 trace-driven 모의실험을 통해 검증한다.