• The Transactions of the Korea Information Processing Society
• /
• v.7 no.11
• /
• pp.3519-3533
• /
• 2000

In a VOD(Video-On-Demand) system, I/O performance and storage requirements are generally more important than computing capability, Due to the development of CPU tecnology, the gap between the performance of processors and desk speed are widening. Therefore, the disk system typically becomes the bohleneck ina VOD system. In order to allevte those effects from physical constramts, disk array system is used. The I/O performance of disk arrays can be improved by the capability of processors, I/O schefuling polcy, the number of member disks in the array, disk block size, block placemet method etc. In this paper, we considered the disk scheduling schemes including EDF( Earliest Deadine First), SCAN, SGAN-EDF, Round-robm. GSS(Grouped Sweeping scheme), and C-SCAN, LOOK, C-LOOK which are variations of SCAN Mamly focusing on those disk scheduling schemes, we performed the simulation in order to compare and eview the perionmance of disk arrays considening seveial related parameters under diverse environments and analyzed the effects of those parameters to the performance.

• The Journal of the Korea Contents Association
• /
• v.10 no.1
• /
• pp.94-102
• /
• 2010

In recent years, embedded systems such as cellular phones, Portable Multimedia Player, intelligent appliance, automobile engine control are reshaping the way people live, work, and play. Thereby, services application to guarantee various requirements of users become increasingly sophisticated and complicated, such embedded computing platforms use real-time operating systems (RTOSs) with time determinism. These RTOSs must not only provide predictable services but must also be efficient and small in size. Kernel services should also be deterministic by specifying how long each service call will take to execute. Having this information allows the application designers to better plan their real-time application software so as not to miss the deadline of each task. In this paper, we present the complete generalized real-time scheduling algorithm using multi-dimensional methodology to determine the highest priority in the ready list with 2r levels of priorities in a constant time without additional memory overhead.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.2
• /
• pp.31-39
• /
• 2013

In this paper, we propose a power-efficient scheduling scheme that stochastically minimizes the power consumption of a real-time parallel task while meeting the deadline on multicore processors. The proposed scheme applies the parallel processing that executes a task on multiple cores concurrently, and activates a part of all available cores with unused cores powered off, in order to save power consumption. It is proved that the proposed scheme minimizes the mean power consumption of a real-time parallel task with probabilistic computation amount on DVFS-enabled multicore processors with a finite set of discrete clock frequencies. Evaluation shows that the proposed scheme saves up to 81% power consumption of the previous method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10C
• /
• pp.1469-1483
• /
• 2004

We propose a packet scheduling algorithms for streaming media. We assume that the receiver periodically reports back the channel throughput. From the original video data, the importance level of a video packet is determined by its relative position within its group of pictures, taking into account the motion-texture discrimination and temporal scalability. Thus, we generate a number of nested substreams. Using feedback information from the receiver and statistical characteristics of the video, we model the streaming system as a queueing system, compute the run-time decoding failure probability of a Same in each substream based on effective bandwidth approach, and determine the optimum substream to be sent at that moment in time. Since the optimum substream is updated periodically, the resulting sending order is different from the original playback order. From experiments with real video data, we show that our proposed scheduling scheme outperforms the conventional sequential sending scheme.

• The KIPS Transactions:PartA
• /
• v.9A no.1
• /
• pp.9-18
• /
• 2002

In this paper, we propose scalable scheduling techniques based on EDF to efficiently integrate hard real-time and multimedia soft real-time tasks in the distributed real-time multimedia database system. Hard tasks are guarangteed based on worst case execution times, whereas multimedia soft tasks are served based on mean execution times. This paper describes a served-based scheme for partitioning the CPU bandwidth among different task classes that coexist in the same system. To handle the problem of class overloads characterized by varying number of tasks and varying task arrival rates, thus scheme shows how to adjust the fraction of the CPU bandwidth assigned to each class. This scheme fixes the maximum time that each hard task can execute in the period of the server, whereas it can dynamically change the bandwidth reserved to each multimedia task. The proposed method is capable of minimizing the mean tardiness of multimedia tasks, without jeopardizing the schedulability of the hard tasks. The performance of this scheduling method is compared with that of similar mechanisms through simulation experiments.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.50 no.5
• /
• pp.218-225
• /
• 2001

This paper analyses the performance of network system using the CAN(Controller Area Network) protocol. Given messages are assumed to be scheduled by the DMS(Deadline Monotonic Scheduling) algorithm. The mathematical models for time-delay that can be occurred between CAN nodes are defined. The effectiveness of modeling is shown by comparing the difference of time-delay between simulations and practical experiments. We analyse the results according to the variation of factors, such as the number of nodes, the transmission speed, the message size and the number of aperiodic messages through simulation and confirm the real-time performance of lower priority messages. We also investigate the real-time performance of periodic messages when aperiodic message generates.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.313-317
• /
• 2008

센서 네트워크는 여러 분야에서 활용할 수 있는 기술이다. 센서 노드가 외부에서 채취한 데이터를 실시간으로 사용자에게 전달하는 것은 매우 중요하다. UC 버클리에서 개발된 TinyOS는 센서 노드에서 동작하는 운영체제 중 가장 많이 사용되고 있다. TinyOS는 Event-driven 방식이며 Component 기반의 센서 네트워크 운영체제이다. 기본적으로 비선점 방식의 스케줄러를 사용함으로써 TinyOS의 실시간성을 보장하기 어렵다. 최근 연구에서 TinyOS의 빠른 반응성을 위해 Priority Level Scheduler라는 선점 기능이 제안되었다. 여기서 본 논문은 TinyOS의 실시간성의 보장을 위해 Priority Level Scheduler에 EDF(Earliest Deadline First)를 적용한 선점형 EDF 스케줄링 방식을 제안하고자 한다.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.591-592
• /
• 2008

본 논문은 배터리를 이용하는 시스템의 사용시간을 극대화하기 위하여 두 가지 해결책을 제시한다. 첫 번째, 우리는 멀티 프로세서 시스템에서 Dynamic Voltage Scaling(DVS)을 이용하여 에너지 소모를 최소화시킨다. 다른 어프로치와의 큰 차이점은 테스크의 실행 시간을 deadline까지 확장시켜 에너지 소모를 최소화할 뿐만 아니라 테스크의 실행 사이클 수가 감소할것을 고려하여 테스크를 나누어 다른 동작 주파수를 적용 시키고 이를 수학적 방법으로 도출한다. 두 번째, 배터리의 discharge 특성인 capacity rate effect와 recovery effect를 고려하여 프로세서들의 에너지 소모 프로파일을 재구성함으로서 배터리 라이프타임을 최적화시킨다.

• Journal of the military operations research society of Korea
• /
• v.10 no.1
• /
• pp.41-55
• /
• 1984

Clarke-Wright algorithm is a well-known heuristic algorithm for the vehicle routing problem, that is, the problem to find routes through which vehicles deliver goods to retailers. The purpose of this paper is to extend the Clarke-Wright algorithm to the vehicle routing problem with the following restrictions: ${\cdot}$ capacities of vehicles may not be the same ${\cdot}$ vehicles should pick up empty boxes on the way back ${\cdot}$ there is a deadline before delivery should be completed for every retailer ${\cdot}$ only certain kinds of vehicles may enter the unloading sites of retailers. This paper modifies the Clarke-Wright algorithm for this problem and shows how to solve a real problem by the newly developed algorithm.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10c
• /
• pp.161-163
• /
• 1999

본 논문은 고정 우선순위를 가지는 주기 태스크와 동적으로 발생하는 비주기 태스크를 스케쥴링하는 방법으로 슬랙 스틸링(slack stealing) 알고리즘을 소개하고 발생되는 문제점으로부터 새로운 알고리즘을 제시한다. 기존의 실시간 스케쥴링 알고리즘을 RM 스케쥴링에 의해 할당된 주기 태스크의 우선순위가 동적으로 발생하는 비주기 태스크의 수행으로 인한 주기 태스크들간의 긴급함에 적절하게 대응하지 못하고 주기 태스크들이 자신의 제한시간을 넘길 수 있다. 이는 슬랙 스틸링(slack stealing) 알고리즘에서 슬랙 타임을 구하는 함수의 한 요소로서 주기 태스크의 필요 계산시간 값이 긴급하지 않은 상위 우선 순위 태스크의 계산시간 까지 포함하는데 있다. 따라서 제안한 확장된 실시간 스케쥴링 알고리즘은 RM 스케쥴링에 의한 스케쥴링 가능성을 위배하지 않으면서 동적으로 발생하는 비주기적 태스크로부터 긴급도의 변화에 적응할 수 있는 우선 순위 체제의 알고리즘을 제시하고, 제안한 알고리즘이 다른 실시간 알고리즘보다 성능 개선이 있음을 모의 실험을 통하여 증명한다.