• Proceedings of the Korean Magnestics Society Conference
• /
• 2004.06a
• /
• pp.66-67
• /
• 2004

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.12 no.5
• /
• pp.519-526
• /
• 1987

An adaptive estimation system which operates propoerly in the environments corrupted by additive impulse noise in addition to the white Gaussian noise has been proposed. A feed forward loop is inserted into the adaptive estimator proposed by R. L. Moose for a system with an unknown measurement bias by which the improved adaptive estimator is processed successfully without the sum of the time varying weights being zero even when the measurement system is added impulue noise. Successfully processed adaptive estimator has been obtained under the large impulse noise in addition to randomly varying unknown biases condition by giving sufficient large value to the elements of discrete vector on the computer simulation.

• Journal of the Korean Vacuum Society
• /
• v.9 no.4
• /
• pp.382-388
• /
• 2000

Hydrogenated amorphous carbon (a-C:H) films were fabricated by electron cyclotron resonance plasma-enhanced chemical vapor deposition. The bonding structure of carbon and hydrogen in the a-C:H films has been investigated by varying the deposition conditions such as ECR power, gas composition of methane and hydrogen, deposition time, and negative DC self bias voltage. The bonding characteristics of the a-C:H thin film were analyzed using FTIR spectroscopy. The IR absorption peaks of the film were observed in the range of $2800\sim3000 \textrm{cm}^{-1}$. The atomic bonding structure of a-C:H film consisted of $sp^3$ and $sp^2$ bonding, most of which is composed of $sp^3$ bonding. The structure of the a-C:H films changed from $CH_3$ bonding to $CH_2$ or CH bonding as deposition time increased. We also found that the amount of dehydrogenation in a-C:H films was increased as the bias voltage increased.

• Journal of the Korean Magnetics Society
• /
• v.21 no.2
• /
• pp.52-55
• /
• 2011

Spin-transfer torque is a useful tool to control the magnetic state in nanostructures. In magnetic tunnel junctions, the spin-transfer torque has two components, the in-plane spin torque and the perpendicular spin torque. While properties of the in-plane spin-transfer torque are relatively well understood, properties of the perpendicular spin-transfer torque still remain controversial. A recent experiment demonstrated that in asymmetric magnetic tunnel junctions, the bias voltage dependence of the perpendicular spin-transfer torque contains both linear and quadratic terms in the bias. However it still remains unexplored how the bias voltage dependence changes as a function of material parameters. In this paper, we systematically investigate the perpendicular spin-transfer torque in asymmetric magnetic tunnel junction by varying spin splitting energy, work function difference, and Fermi energy of the ferromagnetic metal leads.

• Journal of the Korean Magnetics Society
• /
• v.8 no.5
• /
• pp.288-294
• /
• 1998

Effect of degree of intermixing at the Ta/NiFe interface induced by varying applied substrate bias voltage during NiFe free layer deposition on change of magnetoresistance in Substrate/Ta/NiFe/Cu/NiFe/FeMn/Ta spin valve multilayers was investigated. It was found that the optimum NiFe free layer thickness showing a maximum MR increase with increasing the bias voltage. The increase of the optimum thickness was due to the increase of the intermixed layer thickness with a bias voltage. The weak ferromagnetic or non ferromagnetic intermixed layer plays as a spin-independent scattering region and does not contribute on spin-dependent scattering. The existence of the intermixed layer was proved by the means of electrical resistivity and magnetization changes. In the present study, the optimum "effective" free layer thickness which gives the highest MR ratio was a constant independent of the magnitude of the bias voltage we have used.have used.

• Journal of Korea Port Economic Association
• /
• v.30 no.3
• /
• pp.1-14
• /
• 2014

This paper aims at measuring how new information is incorporated into volatility estimates. Various GARCH models are compared and estimated with daily BDI(Baltic Dry Index) data. While most researchers agree that volatility is predictable, they differ on how this volatility predictability should be modelled. This study, hence, introduces the asymmetric or leverage volatility models, in which good news and bad news have different predictability for future. We provide the systematic comparison of volatility models focusing on the asymmetric effect of news on volatility. Specifically, three diagnostic tests are provided: the sign bias test, the negative size bias test, and the positive size bias test. From the Ljung-Box test statistic for twelfth-order serial correlation for the level we do not find any significant serial correlation in the unpredictable BDI. The coefficients of skewness and kurtosis both indicate that the unpredictable BDI has a distribution which is skewed to the left and significantly flat tailed. Furthermore, the Ljung-Box test statistic for twelfth-order serial correlations in the squares strongly suggests the presence of time-varying volatility. The sign bias test, the negative size bias test, and the positive size bias test strongly indicate that large positive(negative) BDI shocks cause more volatility than small ones. This paper, also, shows that three leverage models have problems in capturing the correct impact of news on volatility and that negative shocks do not cause higher volatility than positive shocks. Specifically, the GARCH model successfully reveals the shape of the news impact curve and is a useful approach to modeling conditional heteroscedasticity of daily BDI.

• Applied Chemistry for Engineering
• /
• v.18 no.1
• /
• pp.24-28
• /
• 2007

The etching of zinc oxide (ZnO) thin films has been studied using a high density plasma in a $Cl_2/Ar$ gas. The etch characteristics of ZnO thin films were systematically investigated on varying $Cl_2$ concentration, coil rf power, dc-bias voltage, and gas pressure. With increasing $Cl_2$ concentration, the etch rate of ZnO thin film increased, the redeposition around the etched patterns decreased but the sidewall slope of the etched patterns slanted. As the coil rf power and dc-bias voltage increased, the etch rates of ZnO thin films increased and etch profiles of ZnO thin films were improved. With increasing gas pressure, the etch rate of ZnO thin films slightly increased but little change in etch profile was observed. Based on these results, the optimal etching conditions of ZnO thin film were selected. Finally, the etching of ZnO thin films with a high degree of anisotropy of approximately $75^{\circ}{\sim}80^{\circ}$ without the redepositions and residues was successfully achieved at the etching conditions of 20% $Cl_2$ concentration, coil rf power of 1000 W, dc-bias voltage of 400 V, and gas pressure of 5 mTorr.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.21 no.12
• /
• pp.1051-1056
• /
• 2008

In this study, the investigations of the TiN etching characteristics were carried out with addition of $Cl_2$ gas in an inductively coupled $BCl_3$-base plasma system. Dry etching of the TiN was studied by varying the etching parameters including $Cl_2$ gas addition ratio, RF power, DC-bias voltage and pressure. The etch rate of TiN thin film was maximum when the $Cl_2$ gas addition flow was 2 sccm with fixed other conditions. As the RF power DC-bias voltage were increased, the etch rate of TiN thin film showed increasing tendency. $BCl_3/Cl_2$/Ar plasmas were characterized by optical emission spectroscopy (OES) analysis. The chemical reaction on the surface of the etched TiN films was investigated with X-ray photoelectron spectroscopy (XPS).

• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.20-27
• /
• 2020

Sound event detection in real-world environments suffers from the interference of non-stationary and time-varying noise. This paper presents an adaptive noise reduction method for sound event detection based on non-negative matrix factorization (NMF). In this paper, we proposed a deep learning model that integrates Convolution Neural Network (CNN) with Non-Negative Matrix Factorization (NMF). To improve the separation quality of the NMF, it includes noise update technique that learns and adapts the characteristics of the current noise in real time. The noise update technique analyzes the sparsity and activity of the noise bias at the present time and decides the update training based on the noise candidate group obtained every frame in the previous noise reduction stage. Noise bias ranks selected as candidates for update training are updated in real time with discrimination NMF training. This NMF was applied to CNN and Hidden Markov Model(HMM) to achieve improvement for performance of sound event detection. Since CNN has a more obvious performance improvement effect, it can be widely used in sound source based CNN algorithm.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.16 no.4 s.95
• /
• pp.345-350
• /
• 2005

An X-band push-push VCO for low cost applications is proposed. The designed push-push oscillator achieves a wide tuning range in the X-band by the collector bias tuning instead of extra varactor diodes. The measurement shows a wide tuning bandwidth of $900\;\cal{MHz}\;from\;10.9\;\cal{GHz}\;to\;11.8\;\cal{GHz}$ with a drain bias voltage varying from 4 to 9 V, excellent fudamental suppression of $-30\;\cal{dBc}$ and good phase noise of $-115\;\cal{dBc/Hz}\;@\;1\;\cal{MHz}$ offset.