國內銀行의 證券業 進出에 관한 硏究

이준석 한국외국어대학교 세계경영대학원 1997 국내석사

정규기저를 이용한 reed-solomon 부호의 직접복호법에 관한 연구

이준석 동의대학교 대학원 1998 국내석사

우리나라 권역별 유통 골재의 품질 및 이를 사용한 콘크리트의 특성 평가

이준석 청주대학교 2018 국내박사

본 연구에서는 현재 국내에서 유통되고 있는 각종 골재 자원의 전반적인 품질수준을 확인하고 양질의 골재 사용을 유도하기 위하여 권역별로 유통되고 있는 콘크리트용 골재의 물성평가를 통하여 품질이 KS 표준에 적합한지에 대한 여부를 파악하고자 한다. 또한 이들 골재를 사용한 콘크리트의 성능 평가 실험을 통하여, 불량골재로 인식되고 있는 발파석 및 EEZ 모래의 품질 현황 자료를 제공하여 골재 품질 확보의 필요성에 대한 실증적인 자료로 제시하고자 하였는데, 그 결과를 정리하여 요약하면 다음과 같다. (1) 각 권역의 골재원 및 KS 품질 적합 여부 등에 따른 골재의 기초적 물성을 검토함으로서 골재의 전반적인 품질 현황을 파악할 수 있었다. 골재가 유통되는 지역 및 골재원의 특성에 따라 골재의 품질은 다소 상이하게 나타났는데, 일부 골재의 품질 수준이 KS의 기준에 부적합한 골재인 것으로 나타났다. (2) 골재의 유통과정 중에 입도나 조립률을 검사하지 않는 골재채취법은 확연하게 낮은 불합격 비율이 나타났다. 이는 저품질의 불량 골재가 유통할 수 있는 근거가 되고, 이에 따라 기준에 부적합한 골재도 유통이 가능하기 때문에 이를 통해 국내 각 권역별로 유통되는 골재의 품질 실태가 매우 우려스러운 수준임을 확인하였다. (3) 각 골재원에 따른 골재의 KS 품질 확보 유무에 따른 콘크리트의 제반 성능은 골재원에 따라서 석산 굵은 골재 및 잔골재, 강사를 사용한 경우 우수한 것으로 나타났고, 반면 발파석 굵은 골재 및 잔골재, 마사를 사용한 경우는 골재의 품질 좋지 못하여 단위수량이 증가하고, 압축강도가 저하하게 된다는 것을 확인 할 수 있었다. 이상을 종합하면, 각 권역별로 유통되는 골재들의 품질실태가 매우 우려스러운 수준 이였다. 이러한 KS 표준에 부적합한 골재의 사용은 동일 압축강도 조건에서 물 시멘트비가 증가하게 되어 양질의 골재를 사용한 배합보다 압축강도가 현저히 감소하며 콘크리트 제반성능에 미치는 영향은 매우 큰 것으로 나타났다. 따라서 골재의 품질 확보를 위하여 건설 공사에 양질의 골재가 사용될 수 있도록 전반적인 관리 체계를 구축하고 지속적인 관리가 필요할 것으로 판단된다. With the purpose of confirming the general quality of the aggregate and inducing the usage of a good quality aggregate, in this research, the applicability of the aggregate used in various regions of South Korea for Korean standard (KS) is analyzed by properties evaluation of the aggregates for concrete. Additionally, the experiment on the concrete using these aggregates, the actual data about the necessary of quality control for exploded rock debris and sea sand (EEZ sand) is provided, and the results can be summarized as follow: (1) The general quality of the aggregate from various regions could be possible by evaluating the fundamental properties of the aggregate regarding KS parameters. The quality of aggregates were different depending on the regions used and sources from, and some aggregate were insufficient quality for KS. (2) The aggregate regulated by the aggregate acquisition act without inspecting gradation and fineness modulus during the distribution process showed higher rate of fail regarding the KS. Based on this, it is considered that the aggregate acquisition act provides the reason of distribution of low-quality aggregate. Therefore, it can be problem that the low quality level of the aggregate distributed in South Korea. (3) Regarding the general performance of concrete with the quality of aggregates which were passed KS or not, depending on the sources of the aggregate, coarse and fine aggregate from regular manufacturing process of rock and river caused favorable performance of concrete while exploded rock debris and grounded soil as a coarse and a fine aggregates caused insufficient performance with increased unit water content and decreased compressive strength. In summarizing, the general quality of the aggregate from various regions was relatively low. This insufficient quality level of the aggregate from KS causes increasing water-to-cement ratio to achieve similar level of fluidity, and thus generally low-quality concrete would be produced such as decreased compressive strength and durability. Therefore, to distribute favorable quality of aggregate for desirable performance of concrete, it is considered that the general quality controlling system should be established and continuous management is required.

Random Forest 기반 차량용 CAN BUS 다중 분류 침입 감지 시스템의 ASIC 설계

이준석 숭실대학교 대학원 2024 국내석사

Random Forest 기반 차량용 CAN BUS 다중 분류 침입 감지 시스템의 ASIC 설계 이준석 지능형반도체학과 숭실대학교 대학원 본 연구는 차량 내 네트워크에서 가장 널리 사용되는 CAN(Controller Area Network) 프로토콜의 보안 취약점을 해결하고, 외부 공격을 탐지 및 차단하기 위한 Random Forest 기반의 침입 감지 시스템(IDS)을 제 안한다. 제안하는 방법은 CAN 버스에서 나타날 수 있는 DoS, Fuzzy, Spoofing 공격 유형에 대해 실시간으로 탐지하고 차단할 수 있도록 하였 다. Random Forest 알고리즘은 경량 하드웨어 구현이 가능하며, 다수의 트리를 활용한 병렬 연산을 통해 탐지 지연 시간을 최소화할 수 있다는 장점을 가진다. 이를 통해 CAN 프레임 전송이 완료되기 전에 메시지의 공격 여부를 탐지하고 능동적으로 차단할 수 있도록 설계되었다. 추가적 으로 Random Forest 기반 IDS에 Rule-based 필터를 결합하여 탐지 성 능을 향상시켰으며, 이를 통해 99.18%의 탐지 정확도를 달성하였다. 평 균 탐지 시간은 1.8μs로 측정되었으며, 이는 500kbps CAN 네트워크에서 1비트가 전송되는 시간보다 짧은 수치로 CAN 프레임 전송이 완료되기 전에 차단이 가능함을 확인하였다. ASIC design and verification of Random Forest based CAN bus multi-class classification Intrusion Detection System LEE, JUN SEOK Department of Intelligent Semiconductors Graduate School of Soongsil University This paper proposes a Random Forest-based Intrusion Detection System (IDS) to address the security vulnerabilities of the Controller Area Network (CAN) protocol, which is the most widely used in-vehicle network communication standard, and to detect and block external attacks. The proposed method is designed to detect and block DoS, Fuzzy, and Spoofing attack types on the CAN bus in real-time. The Random Forest algorithm enables lightweight hardware implementation and minimizes detection latency through parallel processing using multiple trees. This design allows the system to identify and actively block malicious messages before CAN frame transmission is completed. Additionally, the integration of a Rule-based filter with the Random Forest-based IDS enhances detection performance, achieving a detection accuracy of 99.18%. The average detection time is measured at 1.8 μs, which is shorter than the time required to transmit one bit in a 500 kbps CAN network, confirming the ability to block attacks before the completion of CAN frame transmission.

성곽유산의 변위 계측을 위한 UAV 기반 사진측량 적용 가능성 연구

이준석 한국전통문화대학교 국가유산전문대학원 2026 국내석사

성곽유산의 변위 계측을 위한 UAV 기반 사진측량 적용 가능성 연구 이준석 유산기술학과 건축·조경·도시 전공 한국전통문화대학교 국가유산전문대학원 본 연구는 대면적을 차지하여 지상 라이다 측량 적용에 한계가 있는 성 곽유산에 대해 드론 기반의 사진측량 기법을 활용한 성벽의 변위 계측 가 능성 검증을 목적으로 하였다. 연구 대상은 사적으로 지정된 해미읍성과 공 산성으로 선정하였다. 연구는 해미읍성을 대상으로 라이다 측량과 사진측량 데이터 비교를 통 해서 드론 기반 사진측량의 정확도를 검증하였다. 이후 공산성을 대상으로 성벽의 지형적 특성에 따라 굴곡형, 평지형, 경사형의 3개 구간을 설정하 고, 서로 다른 시점에 걸쳐 총 3회의 드론 기반 사진측량을 반복하여 오차 발생의 분포 경향 및 저감 방안을 모색하였다. 연구 결과는 다음과 같다. 첫째, 해미읍성을 대상으로 라이다 측량과 드론 기반 사진측량으로 제작 된 3개 쌍의 점군 데이터에 대해 위치를 정합한 결과 평균 RMS가 약 3.53mm로 나타났으며, 10mm 이내의 정합률은 평균 93.76%로 나타났다. 이 를 통해 드론 기반 사진측량 기법이 성곽유산의 계측에 활용 가능함을 확 인하였다. 둘째, 공산성을 대상으로 중복도 80% 조건에서 다시점에 걸쳐 촬영된 드 론 기반 사진측량 점군 데이터의 좌표 정렬 결과, 평균 RMS는 굴곡형 5.15mm, 평지형 5.34mm, 경사형 6.10mm이 산출되었다. 10mm 이내의 평균 정합률은 97.78%로 확인되었다. 이는 다시기에 걸친 드론 기반 사진측량으 로 취득된 데이터가 오차범위인 10mm 내에서 좌표 정렬이 매우 정밀하게 이루어질 수 있음을 시사한다. 셋째, 좌표 정렬 과정에서 상대적으로 높은 오차가 발생한 요인을 저감하 기 위해 순방향 및 역방향 촬영 데이터를 병행하고 중복도를 90%로 향상시 켰다. 그 결과, 평균 RMS는 3.99mm로 감소하였으며, 전체 점군의 99.83%가 10mm 이내의 오차 범위에서 정합되는 것으로 확인되었다. 특히 기존에 오 차가 집중되었던 성벽 상부 및 성돌 틈 영역에서도 뚜렷한 오차 저감 효과 가 나타났다. 이러한 결과는 대면적을 차지하는 성곽유산에 대해 드론 기반 사진측량 기법이 라이다 측량에 근접한 수준의 정확도를 확보할 수 있으며, 다시기에 걸친 반복 촬영 시에도 좌표 정합과 거리 오차 등을 계측할 수 있어서 변 위 계측 등의 모니터링에 활용이 가능함을 밝혔다. 특히, 성벽의 지형적 조 건에 따른 오차 특성을 규명함으로써 향후 대면적 성곽유산의 정기적 모니 터링을 위한 최적 촬영 조건 설정에 사용될 수 있는 기초자료를 제공하였 다는 점에서 의의가 있다. 주제어: 기후변화, 구조물, 붕괴, 문화유산, 안전관리, 3D 스캔

Characteristics of nanowire and nanoparticle framework for functional nano-electronic applications

이준석 한양대학교 대학원 2013 국내박사

This dissertation describes an experimental study on fabrication and characteristics of semiconductor and metal nanoparticle (NP), metal oxide nanowire (NW), and its application for the unique functional nano-electronic devices such as non-volatile nano floating gate memory (NFGM), nanowire field effect transistor (NWFET), and nano generator. Part I : Semiconductor and Metal Nanoparticels At first, in order to develop the optimum fabrication for semiconductor and metal NPs, we invented efficient method for the NPs formation by using a Nd: YAG LASER (wavelength=355 nm) irradiation method. A LASER irradiation method is the simple and direct fabrication of the NPs without performing any micro- or nano-pattering process. Scanned LASER irradiation of low power causes localized segregation of ultrathin as-deposited films by the LASER-induced heat, resulting in the fabrication of NPs. The size of NPs can be easily controlled by initial deposition thickness of materials. In addition, this technique gives narrow spatial distribution of the NPs, resulting in an excellent degree of scalability for the size. Experimentally, observations by scanning electron microscopy (SEM), atomic force microscopy (AFM), and transmission electron microscopy (TEM) clearly confirm highly uniform controllable nanoscale NPs. Second, semiconductor and metal NPs were investigated for use in charge storage for metal oxide semiconductor (MOS) devices with thin HfO2 tunneling and control oxide layer. The CoSi2 NPs were fabricated without a post-annealing process by exposure of Co/Si/HfO2 tunneling oxide/Si stacks to an external UV LASER. The thickness of the Co and Si layers were intentionally controlled to obtain ideal CoSi2 NPs. Observations from X-ray photoelectron spectroscopy (XPS) and high resolution transmission electron microscopy clearly confirm the formation of CoSi2 in our process. These CoSi2 NPs in MOS devices exhibited a memory window of 3.4 V under +11V / -9 V in capacitance-voltage curves as well as good retention and endurance times, thereby demonstrating that they show promise for nonvolatile applications. In addition, the memory behavior of natively oxidized AlOx shell – Al NPs in MOS structure were investigated for high retention characteristics. Among many of the discussions regarding improvements in retention characteristics, an attempt of tunneling barrier engineering (TBE) using a native metal oxide shell was chosen. TEM measurements clearly demonstrate the formation of an AlOx shell (thickness 1 – 1.5 nm), surrounding Al (size 5 – 7 nm) core NPs in the MOS structure. Electrical measurements exhibited a memory window of 3.6 V, together with promising charge retention characteristics of about 10 years. A possible band model needed for enhanced retention characteristics was given by considering the electron / hole barrier width and the additional interface states through the AlOx shell as a method of tunneling barrier engineering. Third, spin filtered charge trap devices called as spin capacitor which modifying a conventional MOS structure to incorporate spin trap sites in the oxide layer and ferromagnetic electrode was experimentally investigated. The basic concept of spin capacitor is starting from a modified by MOS capacitor with floating gate and subsequently detected thorough change of the capacitance curve in accordance with external magnetic field. We consider an identical structure, but having ferromagnetic NPs as floating gate and ferromagnetic thin film as gate electrode. A ferromagnetic cobalt (Co) NPs were used to trap the spin filtered charges. As magnetic properties, the switching magnetic fields are 60 Oe for the Co NPs. The fundamental characteristics of the spin capacitor were analyzed using C-V curve as a function of external magnetic field. Consequently, the C-V and I-V curve reliably exhibited spin dependent MOS behaviors with floating gate. Part II : Metal Oxide Nanowires and Doping Process One-dimensional (1D) nanostructures such as nanowire (NW) and nanotube have attracted great interest over the past decade because of their specific physical properties and their potential as building blocks for next-generation nano-electronic devices. Among various 1D materials, zinc oxide (ZnO), which has a direct and wide band gap, is a promising candidate for light-emitting diodes, UV/gas sensors, transistor channels, and other devices that can utilize the unique vertical alignment characteristics and highly ordered single crystalline properties of NW structures. First, structural properties of ZnO NWs synthesized by hydro thermal process were investigated. The ZnO NWs were synthesized on ZnO thin film seed layers via an aqeous solution method with zinc nitrate hexahydrate (Zn(NO3)2•6H2O) and hexamethylenetetramine (HMT). The growth speed and the shape of the ZnO NWs were determined for various mole concentrations. The structural analysis of the ZnO NWs was performed using X-ray diffraction, SEM, and TEM measurements. The correlation of structural results wit growth conditions, such as the mole concentration and the growth temperature of chemical precursors based on Gibbs free energy. Second, doping process of ZnO nanowires for fuctional nano-electronic applications was investigated. The undoped ZnO NWs are intrinsically n-type, their use in practical devices has been hindered, and much effort has been dedicated toward the development of p-type ZnO NWs. In particular control and manipulation of the doping process is increasingly becoming a key approach that has been taken for the realization of p-type ZnO NWs. To realze p-type ZnO NWs, the initial dopant candidates tested included group-V elements substitute for O and group-III elements to substitute for An, despite the large size mismatches in both cases. Therfore, group I species as lithium (Li) and silver (Ag) have been used to synthesize p-type ZnO NWs based on the expectation thate these elements would function as shallow acceptors in ZnO host materials. In case of Li dopants, Li has the smallest ionic radius (0.76 Å) that is very close to that of Zn (0.74 Å). TEM meausrements of the c-axis oriented and highly vertically aligned NWs demonstrated that Li defect in as-grown ZnO:Li NWs can occupy the empty cages of the wurtzite structure at octahedral sites, and that Li substitution of Zn occurred because of thermally-induced migration due to post-annealing in the presence of oxygen. The stable formation of p-type ZnO:Li NWs using a NWFET and a simple n-type thin film / p-type annealed ZnO:Li NWs homojunction diode. In case of Ag dopants, zinc nitrate hexahydrate (Zn(NO3)2•6H2O) and hexamethylenetetramine (HMT) solution was prepared in deionized water at 25 mmol and silver nitrate AgNO3 solution was also prepared in deionized water at 1 mmol. Low temperature photoluminescence (PL) spectra at 10 K gained significant insight into the effects of Ag-doping. The ZnO:Ag NWs showed four dominant peaks at 3.357, 3.318, 3.252 and 3.180 eV, representing neutral acceptor-bound exciton peak (AoX), free electron to the acceptor transition peak (FA), and donor-to-acceptor pair transition peak (DAP) and donor-to-acceptor pair transition phonon replica peak (DAP-1LO), respectively. Finally, we demonstrate simple and effective approaches to improve the efficiency of piezoelectric energy harvesting through lithium and silver dopants. The output voltage and current of an optimized nano generator (NG) are drastically improved by over 40 times compared to those of a NG based on undoped ZnO NWs, respectively. Furthermore, we demonstrate that output power of our NG is sufficient to drive electrophoretic displays. This approach is expected to be an attractive potential strategy, providing the high possibility for realizing multi-functional, self-powered devices with high performance.

American Option Pricing with Dual Approach

이준석 서울대학교 대학원 2007 국내석사

건강염려 성향자의 질병에 대한 파국적 사고 경향성

이준석 서울대학교 대학원 2004 국내석사

소포클레스 『필록테테스』의 데우스 엑스 마키나 연구 : 플롯과의 연결을 중심으로

이준석 서울대학교 대학원 2003 국내석사

영어 학습자들의 연령별 학습동기와 태도에 관한 연구

이준석 동국대학교 교육대학원 2002 국내석사

The purpose of this study is to investigate the motivation of English learners and to compare their attitude toward the English language and American people according to various age groups (i. e. Middle school students, High school students, the twenties group, and the thirties group). A questionnaire is like belows. 1. What is the purpose of studying English? 2. What is the attitude to English? 3. What is the attitude to American as a native speaker? 4. What do you think about learning English? 5. What do you think about teaching method in Middle & Highschool? 200 subjects of each age were selected and divided into 4 groups by ages. The collected data was classified and analyzed statically. The major findings of this study are follows: In the motivation part of the questionnaire, most of age groups are instrumentally motivated for the examination or needs for their works. Intrinsic motivation could help them to study their learning and teacher should raise their consciousness to remind of it. In the attitude about the English language section, almost age groups recognize that English is indispensable, that is to say, important to succeed in an international society. As the ages are growing, learners have strong desire to study English. However, their current English ability is not being satisfied. In the attitude about American section, the most of age groups have negative attitude even though they'd like to go America if they have a chance. therefore, it is hard to assume that this negative attitude for American could be affected badly to learn English. In the opinion about learning English section, the opinions were somewhat different among age groups, and the score was statistically meaningful. To try to speak aloud is generally effective on all age groups, in the opinion about native language teacher, Middle & High school student groups are doubtful. In the opinion about Korean Middle & High school English classroom teaching, even though the public syllabuses have been developed continually, which has not reflected appropriately on the real classroom teaching method. This study shows that Korean English learners have not only some different motivation but also common preference in English learning, therefore teacher should consider common preferences first and then study how they teach each of students' age groups accordingly.