• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권5호
• /
• pp.2003-2021
• /
• 2020

Today's computer systems are widely integrated with CPU and GPU to achieve considerable performance, but energy consumption of such system directly affects operational cost, maintainability and environmental problem, which has been aroused wide concern by researchers, computer architects, and developers. To cope with energy problem, we propose a task-scheduling framework to reduce energy under performance constraint by rationally allocating the tasks across the CPU and GPU. The framework first collects the estimated energy consumption of programs and performance information. Next, we use above information to formalize the scheduling problem as the 0-1 knapsack problem. Then, we elaborate our experiment on typical platform to verify proposed scheduling framework. The experimental results show that our proposed algorithm saves 14.97% energy compared with that of the time-oriented policy and yields 37.23% performance improvement than that of energy-oriented scheme on average.

• ETRI Journal
• /
• 제28권5호
• /
• pp.561-573
• /
• 2006

In this paper, we present a practical implementation of the MPEG-21 multimedia framework for broadcasting and communication convergent services. MPEG-21 standard technology was exploited to build a convergent service framework. Using this framework, a service model and several scenarios have been successfully designed and implemented. In addition, interoperability, which is the main objective of a multimedia framework, especially in a convergent environment consisting of heterogeneous networks and various types of devices, has been addressed in detail. The experimental results show that the implemented test bed provides a next-generation multimedia service; that is, universal multimedia access (UMA), meeting the requirements of a broadcasting and communication convergent environment.

• 전기학회논문지
• /
• 제67권2호
• /
• pp.270-276
• /
• 2018

Apache Spark is one of the high performance in-memory computing frameworks for big-data processing. Recently, to improve the performance, general-purpose computing on graphics processing unit(GPGPU) is adapted to Apache Spark framework. Previous Spark-GPGPU frameworks focus on overcoming the difficulty of an implementation resulting from the difference between the computation environment of GPGPU and Spark framework. In this paper, we propose a Spark framework based on a heterogenous pipeline computing with OpenCL to further improve the performance. The proposed framework overlaps the Java-to-Native memory copies of CPU with CPU-GPU communications(DMA) and GPU kernel computations to hide the CPU idle time. Also, CPU-GPU communication buffers are implemented with switching dual buffers, which reduce the mapped memory region resulting in decreasing memory mapping overhead. Experimental results showed that the proposed Spark framework based on a heterogenous pipeline computing with OpenCL had up to 2.13 times faster than the previous Spark framework using OpenCL.

• Bulletin of the Korean Chemical Society
• /
• 제19권4호
• /
• pp.422-428
• /
• 1998

In the present paper, we report a molecular dynamics (MD) simulation of non-rigid zeolite-A framework only as the base case for a consistent study of the role of intraframework interaction on several zeolite-A systems using the same technique in our previous studies of rigid zeolite-A frameworks. Usual bond stretching, bond angle bending, torsional rotational, and non-bonded Lennard-Jones and electrostatic interactions are considered as intraframework interaction potentials. The comparison of experimental and calculated structural parameters confirms the validity of our MD simulation for zeolite-A framework. The radial distribution functions of non-rigid zeolite-A framework atoms characterize the vibrational motion of the framework atoms. Mean square displacements are all periodic with a short period of 0.08 ps and a slow change in the amplitude of the vibration with a long period of 0.53 ps. The displacement auto-correlation (DAC) and neighbor-correlation (DNC) functions describe the up-and-down motion of the framework atoms from the center of α-cage and the back-and-forth motion on each ring window from the center of each window. The DAC and DNC functions of the framework atoms from the center of α-cage at the 8-ring windows have the same period of the up-and-down motion, but those functions from the center of 8-ring window at the 8-ring windows are of different periods of the back-and-forth motion.

• 한국해양공학회지
• /
• 제36권4호
• /
• pp.243-250
• /
• 2022

Subsea power cables are subjected to various external loads induced by environmental and mechanical factors during manufacturing, shipping, and installation. Therefore, the prediction of the structural strength is essential. In this study, experimental and theoretical analyses were performed to investigate the axial stiffness of subsea power cables. A uniaxial tensile test of a 6.5 m three-core AC inter-array subsea power cable was carried out using a 10 MN hydraulic actuator. In addition, the resultant force was measured as a function of displacement. The theoretical model proposed by Witz and Tan (1992) was used to numerically predict the axial stiffness of the specimen. The Newton-Raphson method was employed to solve the governing equation in the theoretical analysis. A comparison of the experimental and theoretical results for axial stiffness revealed satisfactory agreement. In addition, the predicted axial stiffness was linear notwithstanding the nonlinear geometry of the subsea power cable or the nonlinearity of the governing equation. The feasibility of both experimental and theoretical framework for predicting the axial stiffness of subsea power cables was validated. Nevertheless, the need for further numerical study using the finite element method to validate the framework is acknowledged.

• IEIE Transactions on Smart Processing and Computing
• /
• 제3권5호
• /
• pp.251-258
• /
• 2014

This paper proposes an efficient token flow design methodology for a decoder in the MPEG Reconfigurable Media Coding (RMC) framework. The MPEG RMC framework facilitates a decoder to be configured with a set of modules called functional units (FUs) that are connected by tokens. Such a modular design philosophy of the MPEG RMC framework enables the reusability and reconfigurability of FUs. One drawback of the MPEG RMC framework is that the decoder performance can be affected by increasing the token transmissions between FUs. The proposed method improves the design of the FU network in the RMC framework toward real-time decoder implementation. In the proposed method, the merging of FU, the separation of token flow, and the merging of token transactions are applied to minimize the token traffic between FUs. The experimental results of the MPEG-4 SP decoder show that the proposed method reduces the total decoding time by up to 77 percent compared to the design of the RMC simulation model.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권3호
• /
• pp.1182-1194
• /
• 2016

The single image dehazing algorithm in existence can satisfy the demand only for improving either the effectiveness or efficiency. In order to solve the problem, a novel variational framework for single image dehazing based on restoration is proposed. Firstly, the initial atmospheric scattering model is transformed to meet the kimmel's Retinex variational model. Then, the green light component of image is considered as an input of the variational framework, which is generated by the sensitivity of green wavelength. Finally, the atmospheric transmission map is achieved by multi-resolution pyramid reduction to improve the visual effect of the results. Experimental results demonstrate that the proposed method can remove haze effectively with less memory consumption.

• International Journal of Contents
• /
• 제8권1호
• /
• pp.47-54
• /
• 2012

The dynamic adaptation between various service environments using the application profiles for the service continuity is a key issue of the profile-based service continuity framework (PSCF) for N-screen service using next generation networks. PSCF offers an optimized service framework for providing continuous user services, which are multimedia video streaming, educational broadcasting, game, etc., using the various devices that are not restricted by the service environment of the user. This paper specifies the functional model of PSCF, service scenario and explains the experimental results of the service continuity for N-screen service using PSCF.

• Journal of information and communication convergence engineering
• /
• 제13권1호
• /
• pp.15-20
• /
• 2015

The limited battery power in the mobile environment, lack of sufficient wireless bandwidth, limited resources of mobile terminals, and frequent breakdowns of the wireless network have become major hurdles in the development of mobile cloud computing (MCC). In order to solve the abovementioned problems, This paper propose a proxy-based MCC framework by adding a proxy server between mobile devices and cloud services to optimize the access to cloud services by mobile devices on the network transmission, application support, and service mode levels. Finally, we verify the effectiveness of the developed framework through an experimental analysis. This framework can ensure that mobile users have efficient access to cloud services.

• Bulletin of the Korean Chemical Society
• /
• 제20권4호
• /
• pp.445-450
• /
• 1999

Molecular dynamics (MD) simulation of non-rigid H-Y zeolite framework are performed at 298.15 and 5.0 K. Usual bond stretching, bond angle bending, torsional rotational, and non-bonded Lennard-Jones and electrostatic interactions are considered as intraframework interaction potentials. Calculated atomic parameters are in good agreement with the experiment, which indicates the successful reproduction of the framework structure and its motion. Both calculated bond lengths and bond angles are also in good agreement with the experiment except generally for a little longer bond lengths and a little smaller T-O-H bond angles. The calculated overall site occupation of HI keeps the order O(2) > O(3) > O(4) > O(t) at 298.15 K, which is very different from the experimental prediction, O(l) > O(3) > O(2) at 5 K. Calculated IR spectra of the H-Y zeolite framework show that most of the main peaks of the O-H bonds are in the broad region 3700-5000 cm-1 and that the O-T stretching bands appeared in 0-2000 cm-1 and at 2700 cm-1