• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.22 no.1
• /
• pp.15-28
• /
• 2018

Recently Machine Learning algorithms are widely used to process Big Data in various applications and a lot of these applications are executed in run time. Therefore the speed of Machine Learning algorithms is a critical issue in these applications. However the most of modern iteration Machine Learning algorithms use a successive iteration technique well-known in Numerical Linear Algebra. But this technique has a very low convergence, needs a lot of iterations to get solution of considering problems and therefore a lot of time for processing even on modern multi-core computers and clusters. Tchebychev iteration technique is well-known in Numerical Linear Algebra as an attractive candidate to decrease the number of iterations in Machine Learning iteration algorithms and also to decrease the running time of these algorithms those is very important especially in run time applications. In this paper we consider the usage of Tchebychev iterations for acceleration of well-known K-Means and SVM (Support Vector Machine) clustering algorithms in Machine Leaning. Some examples of usage of our approach on modern multi-core computers under Apache Spark framework will be considered and discussed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.10
• /
• pp.3520-3537
• /
• 2014

Video copy detection methods have emerged in recent years for a variety of applications. However, the lack of efficiency in the usual retrieval systems restricts their use. In this paper, we propose a parallel implementation strategy for content based video copy detection (CBCD) by using a multi-core processor. This strategy can support video copy detection effectively, and the processing time tends to decrease linearly as the number of processors increases. Experiments have shown that our approach is successful in speeding up computation and as well as in keeping the performance.

• Transactions of Materials Processing
• /
• v.28 no.1
• /
• pp.15-20
• /
• 2019

The superplastic forming/diffusion bonding process has been developed to fabricate a core frame structure with joint nodes out of tubes, for the development of a titanium high performance bicycle. The hydroforming process has been applied for bulging of a tube in the superplastic condition before, and during the diffusion bonding process. In this experiment, a commercial Ti-3Al-2.5V tube was selected as raw material for the study. The forming experiment has been performed using a servo-hydraulic press with a capacity of 200 ton. Next, nitrogen gas was used to acquire necessary pressure for the bulging and bonding of the tubes to fabricate the joint nodes. The pertinent processing temperature was $870^{\circ}C$ for the superplastic hydroforming/diffusion bonding (SHF/DB) process, using the Ti-3Al-2.5V tube. The bonding quality and the progress of bulging and diffusion bonding have been observed by the investigation of the joining interfaces at the cross section of the joint structure. The control of the nitrogen pressure throughout the SHF/DB process, was an important factor to avoid any significant defects in the joint structure. The whole progress stage of the diffusion bonding could be observed at a joint interface. A core structure with 5 joint nodes to manufacture a titanium bicycle could be obtained in a SHF/DB process.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.05a
• /
• pp.111-113
• /
• 2013

멀티 코어 시스템의 보급으로 일반 시스템에서도 프로그램의 병렬 실행이 가능해지고 있다. 본 연구에서는 멀티 코어를 사용하는 단일 시스템에서 분자 동역학 코드인 LAMMPS를 대상으로 병렬 수행 성능을 확인하고 분석하여 효과적인 실행 조건을 살펴보았다. LAMMPS의 구조적인 특징과 공간 분할 방식의 사용으로 인하여 단일 시스템에서도 메시지 전달 방식에 의한 병렬 수행이 보다 효율적임을 확인할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.04a
• /
• pp.146-148
• /
• 2012

최근 GPGPU를 활용한 병렬처리가 각광을 받고 있는 가운데 GPU의 구조적 특성인 매니코어(many core)기반에서 쓰레드(thread)의 구성이 성능에 얼마나 영향을 미치는지에 관해 수치적 해답을 얻고자 하였다. 이는 멀티코어 (multi core)기반으로 작성된 프로그램을 GPGPU로 변환하는 과정에서 쓰레드의 최대활용도를 빠르게 추측 할 수 있도록 도움을 얻고자 하는데 일차적인 목적이 있다. 현재 GPGPU의 쓰레드 구성은 입력되는 데이터의 양을 고려하여 충분한 테스트를 거쳐 경험적인 최적화 수를 지정해 주워야 한다. 이번 연구를 통해 GPGPU로 변환하는 과정에서 최적의 쓰레드 수구성 방법을 추측 할 수 있으며 더 나아가 동적으로 최적의 수를 구할 수 있도록 하는데 목적이 있다.

• Journal of Ceramic Processing Research
• /
• v.13 no.spc1
• /
• pp.42-46
• /
• 2012

The effect of the parameters for few-layered graphene growth by thermal CVD on nickel substrate was investigated. Graphene can be synthesized by using different strategies. Chemical vapor deposition (CVD) has known as one of the most attractive methods to produce graphene due to its good film uniformity, compatibility and large scale production. The control of parameters such as temperature, growth time and pressure in CVD process has been widely recognized as the most important process in graphene growth. Different carbon precursors, methane and acetylene, were introduced in the quartz tube with a variety of growth conditions. Raman spectroscopy was used to confirm the presence of a few- or multi-layered graphene.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.1
• /
• pp.1-14
• /
• 2024

Time-series forecasting is extensively used in the actual world. Recent research has shown that Transformers with a self-attention mechanism at their core exhibit better performance when dealing with such problems. However, most of the existing Transformer models used for time series prediction use the traditional encoder-decoder architecture, which is complex and leads to low model processing efficiency, thus limiting the ability to mine deep time dependencies by increasing model depth. Secondly, the secondary computational complexity of the self-attention mechanism also increases computational overhead and reduces processing efficiency. To address these issues, the paper designs an efficient multi-layer attention-based time-series forecasting model. This model has the following characteristics: (i) It abandons the traditional encoder-decoder based Transformer architecture and constructs a time series prediction model based on multi-layer attention mechanism, improving the model's ability to mine deep time dependencies. (ii) A cross attention module based on cross attention mechanism was designed to enhance information exchange between historical and predictive sequences. (iii) Applying a recently proposed sparse attention mechanism to our model reduces computational overhead and improves processing efficiency. Experiments on multiple datasets have shown that our model can significantly increase the performance of current advanced Transformer methods in time series forecasting, including LogTrans, Reformer, and Informer.

• Journal of Communications and Networks
• /
• v.7 no.3
• /
• pp.377-384
• /
• 2005

The purpose of this research is to develop and evaluate a traffic engineering architecture that uses local state information. This architecture is applied to an Internet protocol radio access network (RAN) that uses multi-protocol label switching (MPLS) and differentiated services to support mobile hosts. We assume mobility support is provided by a protocol such as the hierarchical mobile Internet protocol. The traffic engineering architecture is router based-meaning that routers on the edges of the network make the decisions onto which paths to place admitted traffic. We propose an algorithm that supports the architecture and uses local network state in order to function. The goal of the architecture is to provide an inexpensive and fast method to reduce network congestion while increasing the quality of service (QoS) level when compared to traditional routing and traffic engineering techniques. We use a number of different mobility scenarios and a mix of different types of traffic to evaluate our architecture and algorithm. We use the network simulator ns-2 as the core of our simulation environment. Around this core we built a system of pre-simulation, during simulation, and post-processing software that enabled us to simulate our traffic engineering architecture with only very minimal changes to the core ns-2 software. Our simulation environment supports a number of different mobility scenarios and a mix of different types of traffic to evaluate our architecture and algorithm.

• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.2
• /
• pp.25-30
• /
• 2020

The thermal image system was widely used in the defence-related industry because of detect infrared light from the object without light. but, as the demand in the security system and automobile market increases, the civilian industry are expanding to the private sector. There are difficult to apply various requirement because of previous systems are based by FPGA, so it need new system that apply to various requirement. The proposed paper is thermal image processing system using common image processor. It has various requirement and scalable to support image input/output interface and device driver. If it is used to proposed system, it reduce development cost and period than previous system based FPGA. Because there has very high accessibility. In addition, it expect to have satisfaction of customer requirements, development cost, development period, release date of product.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.10
• /
• pp.1432-1440
• /
• 2022

Recently, as Virtual Desktop Infrastructure(VDI) is widely used not only in office work environments but also in workloads that use high-spec multi-core-based applications, the requirements for real-time and stability of VDI are increasing. Accordingly, the display protocol used for remote access in VDI and performance optimization of virtual machines have also become more important. In this paper, we propose two ways to configure desktop virtualization for multi-core-based application operation. First, we propose a codec configuration of a display protocol with optimal performance in a high load situation due to multi-processing. Second, we propose a virtual CPU scheduling optimization method to reduce scheduling delay in case of CPU contention between virtual machines. As a result of the test, it was confirmed that the H.264 codec of Blast Extreme showed the best and stable frame, and the scheduling performance of the virtual CPU was improved through scheduling optimization.