• Current Photovoltaic Research
• /
• v.4 no.2
• /
• pp.54-58
• /
• 2016

Laser-doped selective emitter process requires dopant source deposition, spin-on-glass, and is able to form selective emitter through SiNx layer by laser irradiation on desired locations. However, after laser doping process, the remaining dopant layer needs to be washed out. Laser-induced melting of pre-deposited impurity doping is a precise selective doping method minimizing addition of process steps. In this study, we introduce a novel scheme for fabricating highly efficient selective emitter solar cell by laser doping. During this process, laser induced damage induces front contact destabilization due to the hindrance of silver nucleation even though laser doping has a potential of commercialization with simple process concept. When the laser induced damage is effectively removed using solution etch back process, the disadvantage of laser doping was effectively removed. The devices fabricated using laser doping scheme power conversion efficiency was significantly improved about 1% abs. after removal the laser damages.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.6
• /
• pp.219-228
• /
• 2007

End-to-end path setup and maintenance are very important for mobile ad-hoc wireless communications, because of the mobility and the limited battery capacity of the nodes in the networks. the AODV routing protocol is the one of mary proposed protocols. However, there are route failure problem with the Proposed protocols between intermediate nodes due to such mobility and exhausted battery characteristics, and this is because only the shortest hop count is considered for the route setup. If route failure happens. Problem such as the waste of bandwidth and the increment of the energy consumption occur because of the discarding data packets in the intermediate nodes and the path re-setup process required by the source node. In addition, it obviously causes the network lifetime to be shortened. This paper proposes a routing protocol based on the AODV routing protocol that it makes use of the remaining energy, signal strength and SNR of the MAC layer resources to setup a path.

• Journal of Conservation Science
• /
• v.27 no.4
• /
• pp.451-458
• /
• 2011

This study aims at identifying wall painting production technique of Songsanri No. 6 tomb based on optical examination and scientific analysis along with the investigation of written records of ancient wall paintings. The result displays that bricks with mortar were used for the construction of the wall, and masonry joint were made of lime, sand and gypsum. The surface of brick were plastered with black and thereafter with white lime. Finely sifted clay on the top of the white surface, were applied not with plasterer's trowel but with a brush for making ground surface. The clay are mineral particles smaller than $30{\mu}m$ of silt. The wall paintings of Sonsanri No. 6 is the only example of making ground layer with clay for painting of Sasindo (literally four heavenly animals painting) among remaining wall paintings of ancient tombs. The historical value of the painting should be attributed not only in terms that it displays a rare production techniques which were never discovered in any other paintings in Korea and Asia, but also it might had been uncompleted when the tomb were closed, even though it considers the fact that the wall painting were discovered in damaged condition in 1933.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.10C
• /
• pp.1010-1015
• /
• 2009

In this paper, we propose a novel detection ordering technique for MIMO signal detection methods based on QR decomposition and successive interference cancellation (SIC). Recently, new signal detection methods for spatially multiplexed (SM) MIMO systems were proposed, where all the constellation points are tried as the first layer symbol, and the remaining layer symbols are estimated via SIC, producing candidate vectors. Finally, the ML metric values are calculated for the candidate vectors, that are again used to select the best symbol vector. It was also shown that the ordering method in the conventional V-BLAST is not suitable to these signal detection methods. In this paper, we propose a novel ordering method, and we show via computer simulations that the proposed ordering method improves the error performance.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.2C
• /
• pp.200-208
• /
• 2009

A macroblock layer rate control for H.264/AVC has the problem that allocated target bits for current frame occasionally are exhausted too fast due to inadequate quantization parameter assignment. In this case, the maximum permissible quantization parameter is used to encode for remaining macroblocks and it leads to degradation of the visual quality. In addition, the header bits length estimation algorithm used for quantization parameter assignment takes the average header bits length for the encoded macroblocks of the previous frame and the current frame. Therefore, it generates a big mismatch between the actually generated header bits length and the estimated header bits length. In this paper, we propose adaptive quantization parameter decision method to prevent early exhausting target bits during encoding the current frame by considering the number of macroblocks that have negative targets bits in previous frame and the improved header bits length estimation scheme for accurate quantization parameter decision.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.4
• /
• pp.139-148
• /
• 2002

Pilot study was performed to examine the feasibility of the pond system for further polishing of treatment wetland effluent from December 2000 to June 2001. The wetland system used for the experiment was highly effective to treat the sewage during the growing season, but it was less effective and its effluent was still high to discharge to the receiving water body. Therefore, the wetland effluent may need further treatment to prevent water quality degradation. Pond system could be used to hold and further polish the wetland effluent during the winter season and ots feasibility was evaluated in this study. Additional water quality improvement was apparent in the pond system during winter season, and the pond effluent could be good enough to meet the effluent water quality standards if it is properly managed. Timing of the pond effluent discharge appears to be critical for pond system management because it is a closed system and whole water quality constituents are affected by physical, chemical, and biological pond environments. Once algae started to grow in mid-April, constituents in the pond water column interact each other actively and its control becomes more complicated. Therefore, upper layer of the pond water column which is clearer than the lower layer my need be discharged in March right after ice cover melted. In the experiment, water quality of the upper water column was markedly clear in March than ant other times probably because of freezing-thawing effect. The remaining lower water column could be further treated by natural purification as temperature goes up or diluted with better quality of wetland effluent for appropriate water uses. This study demonstrated the feasibility of pond system for subsequent management of wetland effluent during the winter season, however, more study is needed for field application.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2008.04a
• /
• pp.1001-1004
• /
• 2008

This study reviewed the results of utilization of insulation heat curing method using double layer bubble sheet in slab concrete and mass concrete in cold weather environment. First of all, when double layer bubble sheets are applied, it was shown that slab concrete was protected from early freezing by remaining between 6 and $10^{\circ}C$ even in case outside temperature drops $10^{\circ}C$ below zero until the 2nd day from piling, and in the case of mass concrete, with the maximum temperature difference between the center and surface less than $6^{\circ}C$, crack occurrence index was close to 2 and no hydration heat crack occurred by internal constraint. The insulation heat preservation curing method using the double bubble sheet applied in this field prevented early freezing owing to stable curing temperature management, deterring concrete strength development delay at low temperature, and obtained the needed strength. Also, it was proven that the method is highly effective and economic for cold weather concrete quality maintenance through curing cost reduction like construction period shortening and labor cost reduction, etc by reducing the process of temporary equipment installation and disassembling.

• Restorative Dentistry and Endodontics
• /
• v.23 no.1
• /
• pp.515-524
• /
• 1998

The purposes of this study were to observe the cavity wall morphology and to evaluate the degree of cavity cleanliness when root-end retrograde cavity preparation was done with ultrasonic diamond instruments. To observe the morphology of retrocavity and to evaluate the degree of cavity cleanliness. root-end resections were done on 20 palatal roots of extracted maxillary first molars after canal filling with gutta-percha. Retrocavities were prepared using either ultrasonic diamond instruments or stainless steel ones of medium power setting of level 6 ($Miniendo^{TM}$, EIE, CA, U.S.A.). Morphology of the cavity. degrees of the remaining canal debris and smear layer were evaluated under the scanning electron microscope. The results were as follows: Cavities prepared with ultrasonic diamond instruments showed scratched appearance of wall, while ultrasonic stainless steel preparation showed hatcheted appearance. Ultrasonic diamond instruments induced more smear layer than stainless steel ones did (p<0.01) in the cavity. However, there was no significant difference in canal debris (p>0.05).

• Korean Journal of Materials Research
• /
• v.29 no.11
• /
• pp.703-708
• /
• 2019

In this study, we investigate the relationship between the peeling behavior of the backsheet of a photovoltaic(PV) module and its surface temperature in order facilitate removal of the backsheet from the PV module. At low temperatures, the backsheet does not peel off whereas, at high temperatures, part of the backsheet remains on the surface of the PV module after the peeling process. The backsheet material remaining on the surface of the PV module is confirmed by X-ray diffraction(XRD) analysis to be poly-ethylene(PE). Differential scanning calorimetry(DSC) is also performed to investigate the interfacial characteristics of the layers of the PV module. In particular, DSC provides the melting temperature($T_m$) of laminated ethylene vinyl acetate(EVA) and of the backsheet on the PV module. It is found that the backsheet does not peel off below the $T_m$ of ethylene of EVA, while the PE layer of the backsheet remains on the surface of the PV module above the $T_m$ of the PE. Thus, the backsheet is best removed at a temperature between the $T_m$ of ethylene and that of PE layer.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.10
• /
• pp.935-943
• /
• 2017

CNN (Convolution neural network), which is used for image classification and speech recognition among neural networks learning based on positive data, has been continuously developed to have a high performance structure to date. There are many difficulties to utilize in an embedded system with limited resources. Therefore, we use GPU (General-Purpose Computing on Graphics Processing Units), which is used for general-purpose operation of GPU to solve the problem because we use pre-learned weights but there are still limitations. Since CNN performs simple and iterative operations, the computation speed varies greatly depending on the thread allocation and utilization method in the Single Instruction Multiple Thread (SIMT) based GPGPU. To solve this problem, there is a thread that needs to be relaxed when performing Convolution and Pooling operations with threads. The remaining threads have increased the operation speed by using the method used in the following feature maps and kernel calculations.