• The KIPS Transactions:PartA
• /
• v.10A no.4
• /
• pp.389-396
• /
• 2003

InfiniBand technology is being accepted as the future system interconnect to serve as the high-end enterprise fabric for cluster computing. This paper presents the design and implementation of the InfiniBand system interconnect, focusing on an InfiniBand host channel adapter (HCA) based on dual ARM9 processor cores The HCA is an SoC tailed KinCA which connects a host node onto the InfiniBand network both in hardware and in software. Since the ARM9 processor core does not provide necessary features for multiprocessor configuration, novel inter-processor communication and interrupt mechanisms between the two processors were designed and embedded within the KinCA chip. Kinch was fabricated as a 564-pin enhanced BGA (Bail Grid Array) device using 0.18${\mu}{\textrm}{m}$ CMOS technology Mounted on host nodes, it provides 10 Gbps outbound and inbound channels for transmit and receive, respectively, resulting in a high-performance cluster system.

• The KIPS Transactions:PartD
• /
• v.11D no.5
• /
• pp.1149-1158
• /
• 2004

Today the fastest growing communities of web users are mobile visitors who browse web page with wireless PDAs and cellular phones. However, most web pages are optimiaed exclusively for desktop clients on the broadband network and are inconvenient to users with small screen mobile devices. They display only a few lines of text and cannot run client-side programs or scripts due to lack of system resource. Even worse, their connections are usually slow to support most of the data-intensive applications. In this paper, we propose a pageslet scheme that makes it feasible to browse ordinary web pages on small screen mobile devices. It extracts broadcasting sections of user preference from broadcasting web pages and automatically reorganizes the extracted sections for convenient browsing on mobile devices.

• Korean Chemical Engineering Research
• /
• v.58 no.4
• /
• pp.596-603
• /
• 2020

In this study, a mathematical model for optimal operation of the fusion fuel cycle is developed based on scheduling approach. The fusion fuel cycle consists of a system for storing and supplying deuterium and tritium, and receiving and separating process after the fusion reaction. Except that tritium is a radioactive material, most of these processes consist of catalytic reactions and separation process. For these reasons, it is possible to apply scheduling approach which is also widely utilized to chemical plants to derive the optimal operating scenarios. The developed model determined the optimal regeneration cycle to minimize the amount of tritium inside the vacuum pumps. Based on the characteristics of various device in the fusion reactor, the optimal tritium plant operation scenario is evaluated. The formulated model was applied to the actual tokamak scenario and utilized to analyze the fuel flow and balance of ITER fuel cycle.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.9A
• /
• pp.867-875
• /
• 2005

In this paper, we fabricated along distance wireless communication transmitter and receiver over the 100 which used NRD technology. To make the transmitter and receiver of 400Hz band using the non radiative dielectric wave guide, transmitter was composed of gm oscillator, m modulator and antenna, receiver was composed of local gum oscillator, balanced mixer, 3dB direction coupler and antenna. Also we executed a wireless communication image transmission examination to the transmitter and receiver. We receive the image information in real-time data transmission from receiver after we send the image signal in the wireless distance of the 10km. Therefore, the 400Hz band U transmitter and receiver to be developed will be used widely for the transmission system CATV or broadband transmission system. This will be utilized also to the link device of a long distance high speed wireless communication network.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.4A
• /
• pp.205-213
• /
• 2012

In this paper, we propose a new channel hopping scheme based on IEEE 802.11h as a good countermeasure against jamming attacks in IEEE 802.11 wireless networks. 802.11h Dynamic Frequency Selection (DFS) is a mechanism which enables hopping to a best channel with full channel measurement, not a randomly chosen channel, when the current link quality degradation occurs due to interferers such as military radars. However, under jammer attacks, this needs a time for full channel measurement before a new channel hopping and due to link disconnection during the time network performance degradation is inevitable. In contrast, our proposed schemes make an immediate response right after a jammer detection since every device is aware of next hopping channel in advance. To do this, a next hopping channel is announced by Beacon frames and the channel is selected by full channel measurement within Beacon intervals. Simulation results show that proposed scheme minimizes throughput degradation and keeps the advantages of DFS.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.12
• /
• pp.1289-1297
• /
• 2014

In this paper, we use the ECMA-392 standard, the first cognitive radio networks to meet the demand for IT services in the ship. We propose a ECMA-392 cooperative MAC protocol to ensure QoS to support multimedia services in the ship area in a multi-hop wireless bridge using ECMA-392 networks. Because the proposed protocol transmits directly or relay transmission by selecting efficient routes based on the information of the ECMA-392 devices, it provides communication to ensure QoS of the device. The simulation results show that the proposed cooperative MAC protocol improves a throughput of devices compared to the existing method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.12
• /
• pp.1267-1279
• /
• 2014

In this paper, we suggest the development of object detection systems for the safety of the ship through the study of the properties of short-range radar. Many of the short-range radars developed for special purpose like cars has cheaper price advantages but it is not proper to every application. In order to overcome such obstacles we need to analysis data from experiments in various environments and feature analysis of the device is essential. Also, the data clustering algorithms to display correct classified moving objects is necessary. In this paper we propose the advanced fast moving object detection system using short range radars with better detection accuracy. And we proposed a clustering algorithm using the value of the RCS and the speed and trajectory information of the radar data that are reflected.

• The Korean Journal of Applied Statistics
• /
• v.31 no.5
• /
• pp.539-554
• /
• 2018

Effective interaction between user and device is considered an important ability of IoT devices. For some applications, it is necessary to recognize human facial expressions in real time and make accurate judgments in order to respond to situations correctly. Therefore, many researches on facial image analysis have been preceded in order to construct a more accurate and faster recognition system. In this study, we constructed an automatic recognition system for facial expressions through two steps - a facial recognition step and a classification step. We compared various models with different sets of data with pixel information, landmark coordinates, Euclidean distances among landmark points, and arctangent angles. We found a fast and efficient prediction model with only 30 principal components of face landmark information. We applied several prediction models, that included linear discriminant analysis (LDA), random forests, support vector machine (SVM), and bagging; consequently, an SVM model gives the best result. The LDA model gives the second best prediction accuracy but it can fit and predict data faster than SVM and other methods. Finally, we compared our method to Microsoft Azure Emotion API and Convolution Neural Network (CNN). Our method gives a very competitive result.

• Journal of Korean Society of Water and Wastewater
• /
• v.31 no.6
• /
• pp.567-575
• /
• 2017

It is not cost effective to raise the density of catch basins in preparation for heavy rainfall in terms of construction and maintenance. Our researchers have developed the new catch basin for increasing interception capacity of runoff with internal filtration structure. To compare interception capacity of an existing catch basin with the invented catch basin, a hydraulic experiment device with 4% of road gradients and 0.2% of road gradients was constructed. For runoff conditions of 4.4 l/s, 6.7 l/s and 10.4 l/s, capability of runoff and separation capability of debris (sand and leaves) were evaluated. As the main experimental results, the effectiveness of the developed catch basin has been verified with an increase in interception rate of approximately 22% for the runoff of 6.7 l/s as heavy rainfall. However, the results of invented catch basin showed only 4.5% of settlement rate of debris regarding sand. Therefore, the authors proposed an improved tilted screen structure additionally. After reviewing the performance of improved catch basin, application of the invented catch basin is expected to drain runoff effectively when it is applied to the faulty road drainage section.

• Journal of the Korean Wood Science and Technology
• /
• v.45 no.6
• /
• pp.797-808
• /
• 2017

Automatic wood species identification systems have enabled fast and accurate identification of wood species outside of specialized laboratories with well-trained experts on wood species identification. Conventional automatic wood species identification systems consist of two major parts: a feature extractor and a classifier. Feature extractors require hand-engineering to obtain optimal features to quantify the content of an image. A Convolutional Neural Network (CNN), which is one of the Deep Learning methods, trained for wood species can extract intrinsic feature representations and classify them correctly. It usually outperforms classifiers built on top of extracted features with a hand-tuning process. We developed an automatic wood species identification system utilizing CNN models such as LeNet, MiniVGGNet, and their variants. A smartphone camera was used for obtaining macroscopic images of rough sawn surfaces from cross sections of woods. Five Korean softwood species (cedar, cypress, Korean pine, Korean red pine, and larch) were under classification by the CNN models. The highest and most stable CNN model was LeNet3 that is two additional layers added to the original LeNet architecture. The accuracy of species identification by LeNet3 architecture for the five Korean softwood species was 99.3%. The result showed the automatic wood species identification system is sufficiently fast and accurate as well as small to be deployed to a mobile device such as a smartphone.