• Proceedings of the KIEE Conference
• /
• 2001.11c
• /
• pp.149-151
• /
• 2001

In real-time control system, it is essential to confirm the timing behavior of all tasks because these tasks of real-time controller have to finish their processes within the specified time intervals called a deadline. In order to satisfy this objective, the timing analysis of a real-time system such as a schedulability test must be performed during the system design phase. This paper presents a Matlab toolbox for simulation of real-time control system based on OSEK-OS, which is one of the most widely adopted real-time operating systems in automotive industry. The toolbox allows the user to explore the timely behavior of control algorithms, and to study the interaction between the object of the OSEK-OS, such as task, scheduler and resource etc.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1994.04a
• /
• pp.357-363
• /
• 1994

In many industrial situations the human acts as the primary scheduler since there often exist various constraints and considerations that may not be mathematically or quantitatively defined. For proper design of interactive scheduling systems, how human strategy and performance are affected by the fashion of human-computer interaction at various levels of task complexity should be investigated. In this study, two scheduling experiments were conducted. The first one showed that human schedulers could perform better than simple heuristic rules with each of typical performance measures such as average machine utilization, average tardiness, and maximum tardiness. In experiment 2, the effect of providing computer-generated initial solution was investigated. The results was that in complex problems the subjects performed significantly better when the initial solutions were generated by themselves, evidencing the importance of the continuity of strategic search through the problem.

• International Journal of Automotive Technology
• /
• v.8 no.1
• /
• pp.39-48
• /
• 2007

This paper addresses the problem of period and priority assignment in networked control systems (NCSs) using a fixed priority scheduler. The problem of assigning periods and priorities to tasks and messages is formulated as an optimization problem to allow for a systematic approach. The temporal characteristics of an NCS should be considered by defining an appropriate performance index (PI) which represents the temporal behavior of the NCS. In this study, the sum of the end-to-end response times required to process all I/Os with precedence relationships is defined as a PI. Constraints are derived from the task and message deadline requirements to guarantee schedulability. Genetic algorithms are used to solve this constrained optimization problem because the optimization formulation is discrete and nonlinear. By considering the effects of communication, an optimum set of periods and priorities can be holistically derived.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.04a
• /
• pp.58-60
• /
• 2014

임베디드 소프트웨어는 국방, 항공 우주, 자동차와 같이 다양한 응용분야에서 그 중요성이 부각되고 있다. 이와 함께 시스템 자원의 효율성을 높이고 응용 소프트웨어 간 안전한 실행환경을 제공하기 위해서 자원 파티셔닝의 필요성이 강조되고 있다. 최근 임베디드 시스템의 응용 분야가 다양해지면서 주기적인 파티션과 함께 비주기적인 파티션들에 대한 요구가 증가하고 있다. 하지만 기존 시스템들은 비주기적인 파티션은 고려하고 있지 않거나, 주기적인 파티션에 비해서 우선순위가 낮게 취급하고 있다. 이러한 문제를 해결하기 위해서 본 논문은 높은 우선순위의 비주기적인 태스크 그룹을 지원하기 위한 구조를 제안하고, 리눅스의 cgroup 프레임워크를 확장하여 구현한다.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.8
• /
• pp.1-6
• /
• 2015

The MapReduce Program of Hadoop Distributed File System operates on any unspecified nodes due to distributed-parallel process and block replicate for data stability. Since it is difficult to guarantee the cache locality when a Solid State Drive is used as a cache in hadoop, cache hit-rate is decreased. In this paper, we suggest a method to improve cache hit rate by pre-loading the input data of the MapReduce onto the SSD cache. To perform this method, we estimated the blocks that are used on each node by using capacity scheduler and block metadata. Eventually we could increase the performance of SSD cache by loading the blocks onto SSD cache before the Map Task run.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.4
• /
• pp.87-92
• /
• 2020

Although GPGPU (General-Purpose GPU) can maximize performance by parallelizing a task with tens of thousands of threads, those threads are internally grouped into a thread block, which is a base unit for processing and resource allocation. A thread block scheduler is a specialized hardware gadget whose role is to allocate thread blocks to GPGPU processing hardware in a round-robin manner. However, round-robin is a sequential allocation policy and is not optimized for GPGPU resource utilization. In this paper, we propose a thread block scheduler model which can analyze and quantify performances for various thread block scheduling policies. Experiment results from the implemented simulator of our model show that the legacy hardware thread block scheduling does not behave well when workload becomes heavy.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.1
• /
• pp.109-119
• /
• 2013

This paper presents the design of the Safety Programmable Logic Controller (SPLC) used in the Nuclear Power Plants, an analysis of a reliability for the SPLC using a markov model. The architecture of the SPLC is designed to have the multiple modular redundancy composed of the Dual Modular Redundancy(DMR) and the Triple Modular Redundancy(TMR). The operating system of the SPLC is designed to have the non-preemptive state based scheduler and the supervisory task managing the sequential scheduling, timing of tasks, diagnostic and security. The data communication of the SPLC is designed to have the deterministic state based protocol, and is designed to satisfy the effective transmission capacity of 20Mbps. Using Markov model, the reliability of SPLC is analyzed, and assessed. To have the reasonable reliability such as the mean time to failure (MTTF) more than 10,000 hours, the failure rate of each SPLC module should be less than $2{\times}10^{-5}$/hour. When the fault coverage factor (FCF) is increased by 0.1, the MTTF is improved by about 4 months, thus to enhance the MTTF effectively, it is needed that the diagnostic ability of each SPLC module should be strengthened. Also as the result of comparison the SPLC and the existing safety grade PLCs, the reliability and MTTF of SPLC is up to 1.6-times and up to 22,000 hours better than the existing PLCs.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.6
• /
• pp.121-126
• /
• 2018

IEEE 802.1 TSN TG is developing standards for end-to-end delay bounds and zero packet loss based on Ethernet technology. We focus on packet forwarding techniques. TSN packet forwarding techniques can be classified into Synchronous and Asynchronous framework. Synchronous approach allocates fixed time period for a class, yet is complex for large networks. Asynchronous approach provides delay guarantee by regulator-scheduler pair, yet is unnecessarily complex, too. We propose network components for TSN Asynchronous architecture, which remove the complexity of maintaining flow state for regulation decisions. Despite such a simplicity, the proposed architecture satisfies the TSN's delay requirements provided the limited high priority traffic's maximum packet length.

• The KIPS Transactions:PartC
• /
• v.13C no.2 s.105
• /
• pp.203-210
• /
• 2006

Fair packet scheduling algorithms supporting quality-of-services of real-time multimedia applications can be classified into the following two design schemes in terms of the reference time used in calculating the timestamp of arriving packet: Finish-time Design (FD) and Start-time Design (SD) schemes. Since the former can adjust the latency of a flow with raising the flow's reserved rate, it has been applied to a router for the guaranteed service of the IETF (Internet Engineering Task Force) IntServ model. However, the FD scheme may incur severe bandwidth loss for traffic flows requiring low-rate but strong delay bound such as internet phone. In order to verify the usefulness of the SD scheme based router for the IETF guaranteed service, this paper analyzes and compares two design schemes in terms of bandwidth and payload utilizations. It is analytically proved that the SD scheme is better bandwidth utilization than the FD one, and the simulation result shows that the SD scheme gives better payload utilization by up to 20%.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2013.01a
• /
• pp.9-12
• /
• 2013

리눅스 기반 안드로이드 단말에서는 CFS(Completely Fair scheduler)가 사용되고 있다. 그리고 CFS는 태스크의 nice값 조절을 통해서 응용프로그램의 CPU 점유율을 제어할 수 있다. 하지만 안드로이드를 업그레이드할 때마다 수많은 태스크의 nice값을 적절하게 맞추는 일은 매우 어려운 일이다. 이러한 문제를 해결하기 위하여 안드로이드 단말은 리눅스의 cgroup(control group)을 사용하여 태스크들을 그룹으로 나눈다. 고성능과 빠른 응답 특성이 필요한 태스크들을 apps 그룹에 할당하여 높은 CPU 점유율을 보장하고, 그렇지 않은 태스크들을 background 그룹에 할당한다. 하지만 안드로이드의 버전이 업그레이드 되면서 각 그룹에 속한 태스크들에도 변화가 생긴다. 그 결과 동일하게 제작된 태스크들의 CPU 점유율이 달라지게 되고 예기치 못한 성능 저하가 발생할 수 있다. 본 연구에서는 안드로이드 버전 향상에 따라 동종 태스크들이 이전 버전에서보다 성능이 하락하는 현상의 원인을 파악하였다. 아울러 분석과 실험을 통하여 태스크의 nice 값보다 그룹 스케줄링 메커니즘이 어떻게 태스크의 CPU 점유율을 결정 짓는지 규명하였다.