• JSTS:Journal of Semiconductor Technology and Science
• /
• v.16 no.6
• /
• pp.808-816
• /
• 2016

As the number of CPU/GPU cores and IPs in SOC increases and applications require explosive memory bandwidth, simultaneously achieving good throughput and fairness in the memory system among interfering applications is very challenging. Recent works proposed priority-based thread scheduling and channel partitioning to improve throughput and fairness. However, combining these different approaches leads to performance and fairness degradation. In this paper, we analyze the problems incurred when combining priority-based scheduling and channel partitioning and propose dynamic priority thread scheduling and adaptive channel partitioning method. In addition, we propose dynamic address mapping to further optimize the proposed scheme. Combining proposed methods could enhance weighted speedup and fairness for memory intensive applications by 4.2% and 10.2% over TCM or by 19.7% and 19.9% over FR-FCFS on average whereas the proposed scheme requires space less than TCM by 8%.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.24 no.3
• /
• pp.327-333
• /
• 2015

As industrial robots come into wider use, their control techniques are being developed along with enhancements in their performance. Specially, the dynamic performance of a 6-axis articulated industrial robot is greatly changed according to the position and orientation of the robot. This means that the PI parameter tuning of the robot and orientation of the robot. This mconsidering the dynamic characteristics of robot mechanism. In this study, $LabView^{(R)}$ programming was applied to automatically conduct parameter scheduling for various robot motions. Using forward and inverse kinematics of RS2, we can divide the working envelope of RS2 into 24 subspaces. We then conduct the gain-tuning according to each subspace. Finally, we program the actual gain scheduling, in which the optimized gain-tuning for each subspace to be passed should be changed for various robot motions using $LabView^{(R)}$.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.1
• /
• pp.165-170
• /
• 2011

The Pfair scheduling algorithm, which is an optimal algorithm in the hard real-time multiprocessor environments, is based on the fixed quantum size. Recently, several methods that can determine the optimal quantum dynamically are developed in the mode change environments. These methods are based on the reach function and in many cases, we have to do the sequential search to find the optimal quantum. In this paper, we propose a new scheduling method, based on the improved reach function, that can determine the optimal quantum more quickly.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.826-828
• /
• 1996

Genetic Algorithm(GA) is a searching or optimizing algorithm based on natural evolution principle. GA has demonstrated considerable success in providing good solutions to many nonlinear, multi-dimensional optimization problems. The preventative maintenance scheduling is a kind of dynamic optimization problem with constraints. This paper applies GA to the preventative maintenance scheduling problem. In the case study, we can get the preventative maintenance scheduling of 3-generators during 8 weeks using GA. It is shown that GA can be available to the preventative maintenance scheduling problem.

• Industrial Engineering and Management Systems
• /
• v.7 no.1
• /
• pp.99-107
• /
• 2008

This paper is concerned with scheduling in Flexible Manufacturing Systems (FMS) using a Fuzzy Logic (FL) approach. Four fuzzy input variables: machine allocated processing time, machine priority, machine available time and transportation priority are defined. The job priority is the output fuzzy variable, showing the priority status of a job to be selected for the next operation on a machine. The model will first select the machines and then assign operations based on a multi-criteria scheduling scheme. System/machine utilization, minimizing mean flow time and balancing machine usage will be covered. Experimental and comparative tests indicate the superiority of this fuzzy based scheduling model over the existing approaches.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.6
• /
• pp.188-198
• /
• 2013

In order to increase its performance, NAND flash memory-based storage is composed of data buses that are shared by a number of flash memories and uses a parallel technique that can carry out multiple flash memory operations simultaneously. Since the storage performance is strongly influenced by the performance of each data bus, it is important to improve the utilization of the bus by ensuring effective scheduling of operations by the storage controller. However, this is difficult because of dynamic changes in buses due to the unique characteristics of each operation with different timing, cost, and usage by each bus. Furthermore, the scheduling technique for increasing bus utilization may cause unanticipated operation delay and wastage of storage resource. In this study, we suggest various dynamic operation scheduling techniques that consider data bus performance and storage resource efficiency. The proposed techniques divide each operation into three different stages and schedule each stage depending on the characteristics of the operation and the dynamic status of the data bus. We applied the suggested techniques to the controller and verified them on the FPGA platform, and found that program operation decreased by 1.9% in comparison to that achieved by a static scheduling technique, and bus utilization and throughput was approximately 4-7% and 4-19% higher, respectively.

• The KIPS Transactions:PartA
• /
• v.12A no.5 s.95
• /
• pp.355-364
• /
• 2005

With fast advances in computing power and network technologies, online streaming services of continuous media (CM) have been popularly implemented on the Web. To implement such services, a variety of CM streams need to be processed efficiently, so that the Sweep scheme was proposed. This scheme has several advantages such as hiccup-free playbacks and seek-time optimization. In this scheme, however, the entire CM streams are scheduled with a single scheduling period, called a cycle. Since only one scheduling period is allowed in this scheme, a significant amount of disk time is usually wasted because of its inflexible disk schedules. To solve this, we propose a new dynamic Sweep scheme. For this, we devise an algorithm that is able to expand scheduling periods of serviced CM streams and propose a new admission control mechanism guaranteeing hiccup-free playbacks. To show performance gains, we execute various simulation experiments. From the experimental results, we can see that the proposed scheme outperforms the Sweep scheme in terms of disk utilization and scheduling flexibility.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.9
• /
• pp.520-529
• /
• 2002

This paper presents an approach for scheduling network messages with real-time dynamic algorithms. We present the method that calculates an intermediate relative deadline of the message based on the EDF(Earliest Deadline First) scheduling policy. We adjust the slack of message by using this intermediate relative deadline to allocate a priority of message. The priority of the message can be determined accurately by using the slack that calculates in our approach, which increases the schedulability efficiency of the message. As a result, we reduce the worst-case response time and improve the guarantee ratio of real-time messages. Also, we describe the analysis method to check the schedulability on message sets, and show the efficiency of our approach by comparing the results of the DM(Deadline Monotonic) approach and the existing EDF approach with that of the improved EDF in our approach through the simulation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.8
• /
• pp.1755-1764
• /
• 1997

In this paper, we proposed the mechanism for the buffer allocation and a cell scheduling method with logical separation a single buffer in the ATm switch, and analyzed the cell loss probability and the delay of each trafic (CBR/VBR/ABR) based on the weighted value and the dynamic cell service scheduling algorithm. The proposed switch buffering system classifies composite trafics incoming to the switch, according to the characteristic of traffic, then stores them in the logically separated buffers, and adopts the round-robin service with weighted value in order to transmit cells in buffers though one output port. We analyzed 4 cell service scheduling algorithms with dynamic round-robinfor each logically separated service line of a single buffer, in which buffers have the respective weighted values and 3 classes on mixed traffic which characteristized by traffic descriptor. In simulation, using SIMCRIPT II.5., we model the VBR and the ABR traffics as ON/OFF processes, and the CBR traffic as a Poisson processes. As the results of analysis according to the proposed buffer management mechanism and cell service algorithm, we have found that the required QoS of each VC can be quaranteed depends on a scale of weighted values allocated to buffers that changed the weighted values, and cell scheduling algorithm.

• The KIPS Transactions:PartC
• /
• v.8C no.6
• /
• pp.805-814
• /
• 2001

In this paper, we propose a new scheduling algorithm that guarantees the delay property of real-time traffic, not considered in previous DWRR(Dynamic Weighted Round Robin) algorithm and also transmits non-real-time traffic efficiently. The proposed scheduling algorithm is a variation of DWRR algorithm to guarantee the delay property of real-time traffic by adding cell transmission method based on delay priority. It also uses the threshold to prevent the cell loss of non-real-time traffic due to cell transmission method based on delay priority. Proposed scheduling algorithm may increase some complexity over conventional DWRR scheme because of cell transmission method based on delay priority. However, the consideration of delay priority can minimize cell delay and require less size of temporary buffer. Also, the results of our performance study shows that the proposed scheduling algorithm has better performance than conventional DWRR scheme due to reliable ABR service and congestion avoidance capacity.