• Journal of the Korean Vacuum Society
• /
• v.22 no.6
• /
• pp.335-340
• /
• 2013

This paper is for the simulation design to enhance the breakdown voltage of MHEMTs with an InP-etchstop layer. Gate-recess and channel structures has been simulated and analyzed for the breakdown of the MHEMT devices. The fully removed recess structure at the drain side of MHEMT shows that the breakdown voltage enhances from 2 V to almost 4 V as the saturation current at gate voltage of 0 V is reduced from 90 mA to 60 mA at drain voltage of 2 V. This is because the electron-captured negatively fixed charges at the drain-side interface between the InAlAs barrier and the $Si_3N_4$ passivation layers deplete the InGaAs channel layer more and thus decreases the electron current passing the channel layer and thus the impact ionization in the channel become smaller. In addition, the replaced InGaAs/InP composite channel with the same thickness in the same asymmetrically recessed structure increases the breakdown voltage to 5 V due to the smaller impact ionization and mobility of the InP layer at high drain voltage.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.9
• /
• pp.1847-1853
• /
• 2011

Analysis of an indoor propagation channel has conventionally used the ray-tracing method. But, in this paper, we had modelling the channel for three dimensional indoor structure by the finite difference time domain method for three dimensional full wave analysis. An excitation signal of the FDTD method used plane wave. The plane wave was excited using the total field/scattered field method. And absorbing boundary condition used the perfectly matched layer method with 7 layers. An living room for the simulation of indoor channel modeling is surrounded the wall that be composed of the wood, the conductor, the glass and concrete. When there are furniture in the living room or not, it were simulated, respectively. As simulation results, we could identify the fading effect of multipath at indoor propagation environment, calculated mean excess delay and rms delay spread for the receiver design.

• Journal of Communications and Networks
• /
• v.10 no.4
• /
• pp.371-383
• /
• 2008

Application of network coding in wireless two-way relay channels (TWRC) has received much attention recently because its ability to improve throughput significantly. In traditional designs, network coding operates at upper layers above (including) the link layer and it requires the input packets to be correctly decoded. However, this requirement may limit the performance and application of network coding due to the unavoidable fading and noise in wireless networks. In this paper, we propose a new wireless network coding scheme for TWRC, which is referred to as soft network coding (SoftNC), where the relay nodes applies symbol-by-symbol soft decisions on the received signals from the two end nodes to come up with the network coded information to be forwarded. We do not assume further channel coding on top of SoftNC at the relay node (channel coding is assumed at the end nodes). According to measures of the soft information adopted, two kinds of SoftNC are proposed: amplify-and-forward SoftNC (AF-SoftNC) and soft-bit-forward SoftNC (SBF-SoftNC). We analyze the both the ergodic capacity and the outage capacity of the two SoftNC schemes. Specifically, analytical form approximations of the ergodic capacity and the outage capacity of the two schemes are given and validated. Numerical simulation shows that our SoftNC schemes can outperform the traditional network coding based two-way relay protocol, where channel decoding and re-encoding are used at the relay node. Notable is the fact that performance improvement is achieved using only simple symbol-level operations at the relay node.

• Journal of Broadcast Engineering
• /
• v.26 no.4
• /
• pp.429-440
• /
• 2021

With the recent development of deep convolutional neural network learning, deep learning techniques applied to single image super-resolution are showing good results. One of the existing deep learning-based super-resolution techniques is RDN(Residual Dense Network), in which the initial feature information is transmitted to the last layer using residual dense blocks, and subsequent layers are restored using input information of previous layers. However, if all hierarchical features are connected and learned and a large number of residual dense blocks are stacked, despite good performance, a large number of parameters and huge computational load are needed, so it takes a lot of time to learn a network and a slow processing speed, and it is not applicable to a mobile system. In this paper, we use the residual dense structure, which is a continuous memory structure that reuses previous information, and the residual dense channel attention block using the channel attention method that determines the importance according to the feature map of the image. We propose a method that can increase the depth to obtain a large receptive field and maintain a concise model at the same time. As a result of the experiment, the proposed network obtained PSNR as low as 0.205dB on average at 4× magnification compared to RDN, but about 1.8 times faster processing speed, about 10 times less number of parameters and about 1.74 times less computation.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.154-154
• /
• 2012

Flexible complementary inverters based on thin-film transistors (TFTs) are important because they have low power consumption and high voltage gain compared to single type circuits. We have manufactured flexible complementary inverters using pentacene and amorphous indium gallium zinc oxide (IGZO) for the p-channel and n-channel, respectively. The circuits were fabricated on polyimide (PI) substrate. Firstly, a thin poly-4-vinyl phenol (PVP) layer was spin coated on PI substrate to make a smooth surface with rms surface roughness of 0.3 nm, which was required to grow high quality IGZO layers. Then, Ni gate electrode was deposited on the PVP layer by e-beam evaporator. 400-nm-thick PVP and 20-nm-thick ALD Al2O3 dielectric was deposited in sequence as a double gate dielectric layer for high flexibility and low leakage current. Then, IGZO and pentacene semiconductor layers were deposited by rf sputter and thermal evaporator, respectively, using shadow masks. Finally, Al and Au source/drain electrodes of 70 nm were respectively deposited on each semiconductor layer using shadow masks by thermal evaporator. The characteristics of TFTs and inverters were evaluated at different bending radii. The applied strain led to change in voltage transfer characteristics of complementary inverters as well as source-drain saturation current, field effect mobility and threshold voltage of TFTs. The switching threshold voltage of fabricated inverters was decreased with increasing bending radius, which is related to change in parameters of TFTs. Throughout the bending experiments, relationship between circuit performance and TFT characteristics under mechanical deformation could be elucidated.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.02a
• /
• pp.479-479
• /
• 2011

Flexible inverters based on complementary thin-film transistor (CTFTs) are important because they have low power consumption and other advantages over single type TFT inverters. In addition, integrated CTFTs in flexible electronic circuits on low-cost, large area and mechanically flexible substrates have potentials in various applications such as radio-frequency identification tags (RFIDs), sensors, and backplanes for flexible displays. In this work, we introduce flexible complementary inverters using pentacene and amorphous indium gallium zinc oxide (IGZO) for the p-channel and n-channel, respectively. The CTFTs were fabricated on polyimide (PI) substrate. Firstly, a thin poly-4-vinyl phenol (PVP) layer was spin coated on PI substrate to make a smooth surface with rms surface roughness of 0.3 nm, which was required to grow high quality IGZO layers. Then, Ni gate electrode was deposited on the PVP layer by e-beam evaporator. 400-nm-thick PVP and 20-nm-thick ALD Al2O3 dielectric was deposited in sequence as a double gate dielectric layer for high flexibility and low leakage current. Then, IGZO and pentacene semiconductor layers were deposited by rf sputter and thermal evaporator, respectively, using shadow masks. Finally, Al and Au source/drain electrodes of 70 nm were respectively deposited on each semiconductor layer using shadow masks by thermal evaporator. Basic electrical characteristics of individual transistors and the whole CTFTs were measured by a semiconductor parameter analyzer (HP4145B, Agilent Technologies) at room temperature in the dark. Performance of those devices then was measured under static and dynamic mechanical deformation. Effects of cyclic bending were also examined. The voltage transfer characteristics (Vout- Vin) and voltage gain (-dVout/dVin) of flexible inverter circuit were analyzed and the effects of mechanical bending will be discussed in detail.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.08a
• /
• pp.166-166
• /
• 2010

Recently, TFTs based on amorphous oxide semiconductors (AOSs) such as ZnO, InZnO, ZnSnO, GaZnO, TiOx, InGaZnO(IGZO), SnGaZnO, etc. have been attracting a grate deal of attention as potential alternatives to existing TFT technology to meet emerging technological demands where Si-based or organic electronics cannot provide a solution. Since, in 2003, Masuda et al. and Nomura et al. have reported on transparent TFTs using ZnO and IGZO as active layers, respectively, much efforts have been devoted to develop oxide TFTs using aforementioned amorphous oxide semiconductors as their active layers. In this thesis, I report on the performance of thin-film transistors using amorphous indium gallium zinc oxides for an active channel layer at room temperature. $SiO_2$ was employed as the gate dielectric oxide. The amorphous indium gallium zinc oxides were deposited by RF magnetron sputtering. The carrier concentration of amorphous indium gallium zinc oxide was controlled by oxygen pressure in the sputtering ambient. Devices are realized that display a threshold voltage of 1.5V and an on/off ration of > $10^9$ operated as an n-type enhancement mode with saturation mobility with $9.06\;cm^2/V{\cdot}s$. The devices show optical transmittance above 80% in the visible range. In conclusion, the fabrication and characterization of thin-film transistors using amorphous indium gallium zinc oxides for an active channel layer were reported. The operation of the devices was an n-type enhancement mode with good saturation characteristics.

• Ocean and Polar Research
• /
• v.39 no.2
• /
• pp.85-106
• /
• 2017

The Lago Sofia conglomerate in southern Chile is a deep-marine gravelly deposit, which is hundreds of meters thick and kilometers wide and extends laterally for more than 100 km, filling the foredeep trough of the Cretaceous Magallanes Basin. For understanding the depositional processes and environments of this gigantic deep-sea conglomerate, detailed analyses on sedimentary facies, architecture and paleoflow patterns were carried out, highlighting the differences between the northern (Lago Pehoe and Lago Goic areas) and southern (Lago Sofia area) parts of the study area. The conglomerate bodies in the northern part occur as relatively thin (< 100 m thick), multiple units intervened by thick mudstone-dominated sequences. They show paleoflows toward ENE and S to SW, displaying a converging drainage pattern. In the southern part, the conglomerate bodies are vertically interconnected and form a thick (> 400 m thick) conglomerate sequence with rare intervening fine-grained deposits. Paleoflows are toward SW. The north-to-south variations are also distinct in sedimentary facies. The conglomerate bodies in the southern part are mainly composed of clast-supported conglomerate with sandy matrix, which is interpreted to be deposited from highly concentrated bedload layers under turbidity currents. Those in the northern part are dominated by matrix- to clast-supported conglomerate with muddy matrix, which is interpreted as the products of composite mass flows comprising a turbidity current, a gravelly hyperconcentrated flow and a mud-rich debris flow. All these characteristics suggest that the Lago Sofia conglomerate was formed in centripetally converging submarine channels, not in centrifugally diverging channels of submarine fans. The tributaries in the north were dominated by mass flows, probably affected by channel-bank failures or basin-marginal slope instability processes. In contrast, the trunk channel in the south was mostly filled by tractive processes, which resulted in the vertical and lateral accretion of gravel bars, deposition of gravel dunes and filling of scours and channels, similar to deposits of terrestrial gravel-bed rivers. The trunk channel developed along the axis of foredeep trough and its confinement within the trough is probably responsible for the thick, interconnected channel fills. The large-scale architecture of the trunk-channel fills shows an eastward offset stacking pattern, suggesting that the channel migrated eastwards most likely due to the uplift of the Andean Cordillera.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.24 no.3
• /
• pp.149-158
• /
• 2012

This study analyzed cross-sectional variations in residual current and strengths of stratification by observing cross-sectional velocity and salinity during spring tide and neap tide, respectively, for continuous 13-hour periods at 2 observation lines at northern and southern end of Seokmo Channel, which is located west of Ganghwado. Salinity distribution of channel depends on not only neap and spring tide but also impact of salinity. The residual current component was obtained by removing $M_2$ and $M_4$ tidal components that were extracted using the least squares method on 13-hour velocity component. Cross-section of residual velocity at northern and southern end of Seokmo Channel exhibited southward residual components at channel's surface layer, but northward residual current was observed at channel's bottom layer, clearly showing a 2-layer tidal circulation between surface and bottom layers. The variation in location of appearing northward residual current according to changes in spring and neap tidal cycle and its correlation with stratification were analyzed using the Richardson number and Simpsonhunter index. At northern and southern end of Seokmo Channel, northward residual current appears in the location where Richardson number is large, Simpson-hunter index appears as a value greater than 4.

• 한국정보통신설비학회:학술대회논문집
• /
• 2009.08a
• /
• pp.309-313
• /
• 2009

In this paper, the network performance of a turbo coded optical code division multiple access (CDMA) system with cross-layer, which is between physical and network layers, concept is analyzed and simulated. We consider physical and MAC layers in a cross-layer concept. An intensity-modulated/direct-detection (IM/DD) optical system employing pulse position modulation (PPM) for satellite broadcasting communications is considered. In order to increase the system performance, turbo codes composed of parallel concatenated convolutional codes (PCCCs) is utilized. The network performance is evaluated in terms of bit error probability (BEP). From the simulation results, it is demonstrated that turbo coding offers considerable coding gain with reasonable encoding and decoding complexity. Also, it is confirmed that the performance of such an optical CDMA network can be substantially improved by increasing the interleaver length and the number of iterations in the decoding process. The results of this paper can be applied to implement the satellite broadcasting communications.