• KIPS Transactions on Computer and Communication Systems
• /
• v.3 no.6
• /
• pp.173-178
• /
• 2014

By exploiting parallel processing, the proposed scheduling scheme enhances energy saving capability of multicore processors for real-time tasks while satisfying deadline and fault tolerance constraints. The scheme searches for a near minimum-energy schedule within a polynomial time, because finding the minimum-energy schedule on multicore processors is a NP-hard problem. The scheme consumes manifestly less energy than the state-of-the-arts method even with low parallel processing speedup as well as with high parallel processing speedup, and saves the energy consumption up to 86%.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.7
• /
• pp.505-509
• /
• 2009

It is the most important object of real-time computing to make real-time tasks keep their given time conditions. In this paper, we implemented BCC(Basic Concurrency Constraint) scheduler which is provided as an essential element of TMO(Time-triggered Message-triggered Object) model, and TMO Supporting Library that supports object-oriented design for TMO. BCC scheduler is a means to design timeliness-guaranteed computing, and it predicts the start of SpMs first, and then it makes the execution of SvMs deferred when it is predicted that any SpM begins to run currently. In this way, BCC is able to prevent collisions between SpM and SvM, and it gives higher priority to SpMs than SvMs.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.5_6
• /
• pp.279-287
• /
• 2004

This paper describes a processor architecture, named Dynamically Instruction Scheduled VLIW (DISVLIW). The DISVLIW Processor architecture is designed for dynamic scheduling VLIW instructions using dependency information. The DISVLIW instruction format is augmented to allow dependency bit vectors to be placed in the same VLIW word. The DISVLIW processor dynamically schedules each instruction in long instructions using functional unit and dynamic scheduler pairs. Features such as explicit parallelism, balanced scheduling effort, and dynamic scheduling of VLIW instructions can be used to provide a sound frustructure for supercomputing. We simulate the DISVLIW processor architecture and show that the DISVLIW processor performs significantly better than the VLIW processor for a wide range of cache sites and across numerical benchmark applications.

• Journal of Internet Computing and Services
• /
• v.5 no.1
• /
• pp.67-84
• /
• 2004

Real-time database systems (RTDBS) are database systems whose transactions are associated with timing constraints such as deadlines. Therefore transaction needs to be completed by a certain deadline. Besides meeting timing constraints, a RTDBS needs to observe data consistency constraints as well. That is to say, unlike a conventional database system, whose main objective is to provide fast average response time, RTDBS may be evaluated based on how often transactions miss their deadline, the average lateness or tardiness of late transactions, the cost incurred in transactions missing their deadlines. Therefore, in RTDBS, transactions should be scheduled according to their criticalness and tightness of their deadlines, even If this means sacrificing fairness and system throughput, And It always must guarantee preceding process of the transaction with the higher priority. In this paper, we propose an efficient real-time scheduling algorithm (Multi-level EFDF) that alleviates problems of the existing real-time scheduling algorithms, a real-time concurrency control algorithm(2PL-FT) for firm and soft real-time transactions. And we compare the proposed 2PL F[ with AVCC in terms of the restarting ratio and the deadline missing ratio of transactions. We show through experiments that our algorithms achieve good performance over the other existing methods proposed earlier.

• The KIPS Transactions:PartA
• /
• v.18A no.3
• /
• pp.93-98
• /
• 2011

While the use of battery-powered embedded systems has been rapidly increasing, the current level of battery technology has not kept up with the drastic increase in power consumption by the system. In order to prolong system usage time, the battery size needs to be increased. The amounts of power consumption by embedded systems are determined by their hardware configuration and software for manipulating hardware resources. In spite of that, the hardware provides features for lowering power consumption, if those cannot be utilized efficiently by software including operating system, reduction in power consumption is not optimized. In this paper, we propose a BET(Break Even Time)-based scheduling method using task partition to reduce power consumption of multimedia applications in a mobile embedded system environment.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.39 no.12
• /
• pp.539-546
• /
• 2002

In this paper, we consider a wireless multimedia environment to service both real-time video traffic and non-real-time WWW-application traffic In our suggested new packet scheduling algorithm, we consider both the accumulation counter and SIR to reduce delay in real-time traffic. In addition, our packet scheduling algorithm gives priority first to real-time video traffic service and then to non-real-time internet Packet service when real-time traffic service is absent. From the simulation results, we find that the AC (Accumulation Counter) scheme has much smaller delay than the conversional M-LWDF scheme for real-time video data users, which has a special quality sensitive to its own delay. We also consider the transmission structure of using both the frame period in the time-axis and the OVSF codes in the code-axis at the same time, which is similar to the structure of HSDPA system.

• Journal of Internet Computing and Services
• /
• v.1 no.2
• /
• pp.61-67
• /
• 2000

In case of the remote controllable web-camera that was took notice especially in internet real-time broadcasting systems, a great many clients connect the server of web-camera to request the service. So, the scheduling methods are important. Web-camera systems have used to the traditional FIFO(First In First Out) or SDF(Shortest Distance First) scheduling method. But they does not satisfy both the minimization of moving distance on the web-camera and the fairness on the users. In this paper, We propose the 2D scheduling method, As a result, the moving distance of the web-camera decreases compared with FIFO scheduling method. And the starvation state on the user's request does not happen compared with SDF scheduling method. Thus, if the remote controllable web-camera systems use the 2D scheduling method, they are satisfied with the minimization of moving distance on the remote controllable web-camera and the fairness on the users simultaneously. Therefore the user's satisfaction and the performance of the systems are improved.

• The Journal of Society for e-Business Studies
• /
• v.23 no.3
• /
• pp.1-12
• /
• 2018

Mixed-criticality systems consist of components with different criticality. Recently, components are categorized depending on criticality by ISO 26262 standard and DO-178B standard in automotive and avionic domain. Existing mixed-criticality system research achieved efficient and safe scheduling through system-level criticality mode. The drawback of these approaches is performance degradation of low-criticality tasks on high-criticality mode. Task-level criticality mode is one method to address the problem and improve the performance of low-critical tasks. In this paper, we propose probabilistic performance metric for the approach. In simulation results with probabilistic performance metric, we showed that our approach has better performance than the existing approaches.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04a
• /
• pp.148-150
• /
• 2003

본 논문에서는 인터넷 종단 호스트에서 공유 자원의 대역폭 제약조건과 종료시한 제악조건 모두를 만족시킬 수 있는 자원 할당 구조를 제안한다. 제안된 구조는 두 단계로 구성된다. 상위 단계에서는 비례 분배 스케줄러(proportional share scheduler)인 EFT-C/D (Earliest Finish Time Credit/Debit) 스케줄러가 수행된다. 이 스케줄러는 CPU와 같은 시분할 공유 자원을 하위 단계 스케줄러들에게 지정된 비율로 분배하는 역할을 한다. 그리고 하위 단계에서는 서로 다른 시간 제약조건이 부여된 태스크들을 스케줄링 하기위해 다양한 실시간 스케줄러가 수행된다. 본 연구의 주요 성과는 두 가지로 요약된다. 첫째, 이상적인 GPS (Generalized Processor Sharing) 서버와 거의 동등한 수준으로 자원을 공평하게 분배하는 EFT-C/D 알고리즘을 개발하였다. 둘째, 하위 단계에서 수행되는 EDF 스케줄러에 대해 이용율(utilization)에 기반한 스케룰링 가능성 분석 방법을 개발하였다. 이 방법은 주어진 태스크 집합에 대해 단순히 이용율만을 계산하여 스케줄링 가능성을 판별할 수 있다. 따라서 새로운 태스크가 생성될 때 수락 여부를 시스템 수행 중에 제어할 수 있는 장점이 있다.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.37 no.4
• /
• pp.9-15
• /
• 2000

The purpose of this paper is to minimize the a slack computation time of the scheduling of a soft aperiodic real-time tasks in a fixed priority real-time system. The proposed algorithm reduces the computation overhead at on-line time and supports the maximum slack time assigned for aperiodic real-time tasks. The proposed algorithm has 10~20% more response time for aperiodic real-time tasks than that of Slack Stealing Algorithm that offers optimal response time in fixed priority real-time system. However, the performance of the proposed algorithm is seven times better in a scheduling overhead.