• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.9
• /
• pp.189-196
• /
• 2020

Recently, GPU cloud computing technology applying GPU(Graphics Processing Unit) devices to virtual machines is widely used in the cloud environment. In a cloud environment, GPU devices assigned to virtual machines can perform operations faster than CPUs through massively parallel processing, which can provide many benefits when operating high-performance computing services in a variety of fields in a cloud environment. In a cloud environment, a GPU device can help improve the performance of a virtual machine, but the virtual machine scheduler, which is based on the CPU usage time of a virtual machine, does not take into account GPU device usage time, affecting the performance of other virtual machines. In this paper, we test and analyze the performance degradation of other virtual machines due to the virtual machine that performs GPGPU(General-Purpose computing on Graphics Processing Units) task in the direct path based GPU virtualization environment, which is often used when assigning GPUs to virtual machines in cloud environments. Then to solve this problem, we propose a GPGPU task management method for a virtual machine.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.13 no.4
• /
• pp.278-286
• /
• 2008

This paper investigates the fault generation type and the cause of output interruptionsin bulky with $30^{\sim}500[KVA]$ double conversion UPS, and proposes the fault detection method to improve the reliability of power supply used in the critical load in industry. Identifies its existing way of detecting a quality of inverter output it to bypass when exceeds its expectation. Under a UPS managing system, when an inner (Power device, Controller, CPU) fault occurs it disrupts the power supply and these occurrences has been verified by the results of experiments and application results. To overcome these problems, the proposed method constructs independently a fault-detection, a bypass-control device and a triple power supply apart from the conventional UPS operation. Also the detection point is changed to the preceding of a circuit breaker, a reference of fault detection is modified to avoid any clash and the breaking equipment is attached to intercept a spread of accident. As a result of applications of these developed systems to 242 UPS which was installed purposefully to the communication power supply, the service errors has not occurred in the UPS for two years since 2006.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.33-40
• /
• 2023

As South Korea enters the realm of an super-aging society, the demand for elderly welfare services has been steadily rising. However, the current shortage of welfare personnel has emerged as a social issue. To address this challenge, there is active research underway on elderly care robots designed to mitigate the social isolation of the elderly and provide emergency contact capabilities in critical situations. Nonetheless, these functionalities require direct user contact, which represents a limitation of conventional elderly care robots. In this paper, we propose a solution to overcome these challenges by introducing a care robot system capable of interacting with users without the need for direct physical contact. This system leverages commercialized elderly care robots and cameras. We have equipped the care robot with an edge device that incorporates facial expression recognition and action recognition models. The models were trained and validated using public available data. Experimental results demonstrate high accuracy rates, with facial expression recognition achieving 96.5% accuracy and action recognition reaching 90.9%. Furthermore, the inference times for these processes are 50ms and 350ms, respectively. These findings affirm that our proposed system offers efficient and accurate facial and action recognition, enabling seamless interaction even in non-contact situations.

• ETRI Journal
• /
• v.27 no.5
• /
• pp.569-578
• /
• 2005

The purpose of this paper is to describe the implementation of monolithically matching circuits, interface circuits, and RF core circuits to the same substrate. We designed and fabricated on-chip 1 to 6 GHz up-conversion and 1 to 8 GHz down-conversion mixers using a 0.8 mm SiGe hetero-junction bipolar transistor (HBT) process technology. To fabricate a SiGe HBT, we used a reduced pressure chemical vapor deposition (RPCVD) system to grow a base epitaxial layer, and we adopted local oxidation of silicon (LOCOS) isolation to separate the device terminals. An up-conversion mixer was implemented on-chip using an intermediate frequency (IF) matching circuit, local oscillator (LO)/radio frequency (RF) wideband matching circuits, LO/IF input balun circuits, and an RF output balun circuit. The measured results of the fabricated up-conversion mixer show a positive power conversion gain from 1 to 6 GHz and a bandwidth of about 4.5 GHz. Also, the down-conversion mixer was implemented on-chip using LO/RF wideband matching circuits, LO/RF input balun circuits, and an IF output balun circuit. The measured results of the fabricated down-conversion mixer show a positive power conversion gain from 1 to 8 GHz and a bandwidth of about 4.5 GHz.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.2 no.4
• /
• pp.145-154
• /
• 1998

The motion-limiting devices can be used for reducing the maximum and residual displacements of the multi-span continuous bridges with inelastic elements such as isolation bearings and plastic hinges formed in piers. For the design of motion-limiting device, the nonlinear time history analysis is required. But the time history analysis is time consuming and very complex. This study suggests the simple design procedure of the motion-limiting devices using the equivalent elastic analysis method and the acceleration-displacement spectrum concept. The suggested design procedure can be used very effectively for determining the location and gap size of the motion-limiting devices.

• ETRI Journal
• /
• v.37 no.2
• /
• pp.233-240
• /
• 2015

The novelty of this study resides in a 6"-wafer-level microfabrication protocol for a microdevice with a fluidic control system for the separation of circulating tumor cells (CTCs) from human whole blood cells. The microdevice utilizes a lateral magnetophoresis method based on immunomagnetic nanobeads with anti-epithelial cell adhesive molecule antibodies that selectively bind to epithelial cancer cells. The device consists of a top polydimethylsiloxane substrate for microfluidic control and a bottom substrate for lateral magnetophoretic force generation with embedded v-shaped soft magnetic microwires. The microdevice can isolate about 93% of the spiked cancer cells (MCF-7, a breast cancer cell line) at a flow rate of 40/100 mL/min with respect to a whole human blood/buffer solution. For all isolation, it takes only 10 min to process 400 mL of whole human blood. The fabrication method is sufficiently simple and easy, allowing the microdevice to be a mass-producible clinical tool for cancer diagnosis, prognosis, and personalized medicine.

• Journal of Sensor Science and Technology
• /
• v.11 no.5
• /
• pp.294-300
• /
• 2002

In this study, we fabricated thermopile infrared sensor with floating membrane structure. Floating membrane was formed by SOI(Silicon On Insulator) structure. In SOI structure, silicon dioxide layer between top silicon layer and bottom silicon substrate was etched by HF solution, then membrane was floated over substrate. After membrane was floated, thermopile pattern was formed on membrane. By insertion of SOI technology, we could obtain thermal isolation structure easily and passivation process for sensor pattern protection was not required during fabrication process. Then, the amplifier circuit for thermopile sensor was fabricated by using $1.5{\mu}m$ CMOS process. The voltage gain of fabricated amplifier was about two hundred.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2009.06a
• /
• pp.401-401
• /
• 2009

Recently, demands for the development of compact, lightweight power supplies with higher power density and higher efficiency have been increased. Since Piezoelectric Transformer (PT) was emerged in device and material industry, it has been suggested as a viable alternative to the magnetic transformer in some applications. PT has some advantages such as low profile and mechanical energy transfer with little electromagnetic interface (EMI). Also, PT can provide high voltage stepping ratio with good isolation and requires no copper windings saving copper usage especially for large voltage conversion differences. Conventional control of PT converter has mainly two-way. One is the pulse frequency modulation (PFM) control method and the other is the pulse width modulation (PWM) control with frequency fixed method. It is known that the maximum PT efficiency can be obtained when it operates near the resonant frequency of the PT. And, also PT's resonant frequency moves according to the load condition. Therefore, selection of PT converter control method is very difficult. This paper analyzes general piezo-electric converter modeling and proposes a guide-line to selection of control method and stabilization control.

• Journal of electromagnetic engineering and science
• /
• v.8 no.3
• /
• pp.129-133
• /
• 2008

An integrated passive device(IPD) technology by semi-insulating(SI)-GaAs-based fabrication has been developed to meet the ever increasing needs of size and cost reduction in wireless applications. This technology includes reliable NiCr thin film resistor, thick plated Cu/Au metal process to reduce resistive loss, high breakdown voltage metal-insulator-metal(MIM) capacitor due to a thinner dielectric thickness, lowest parasitic effect by multi air-bridged metal layers, air-bridges for inductor underpass and capacitor pick-up, and low chip cost by only 6 process layers. This paper presents the Wilkinson power divider with excellent performance for digital cellular system(DCS). The insertion loss of this power divider is - 0.43 dB and the port isolation greater than - 22 dB over the entire band. Return loss in input and output ports are - 23.4 dB and - 25.4 dB, respectively. The Wilkinson power divider based on SI-GaAs substrates is designed within die size of $1.42\;mm^2$.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2014.04a
• /
• pp.352-357
• /
• 2014

Recently the construction of stations under railway lines and railway sections passing through central area of cities are increasing, calling for an urgent establishment of countermeasures against railway vibration and its subsequent second-phase noise. Of technology developed up to now, the most efficient countermeasure is the floating slab track, a track system isolated from the sub-structure by springs. Developed in this study, anti-vibration device for floating slab track (HLRM-High Load Rubber Mount) Haman station affiliated to the primary measurement was conducted in October 2012, one year after the second measurement after 2013 to be carried out in November and we want to change that. It is expected to raise awareness for the need of technology self-support and to make a meaningful contribution to mitigating vibration and noise produced by the next-generation high-speed railway.