• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.10
• /
• pp.231-238
• /
• 2020

With a large number of documents appearing on the web, document-level novelty detection has become important since it can reduce the efforts of finding novel documents by discarding documents sharing redundant information already seen. A recent work proposed a convolutional neural network (CNN)-based novelty detection model with significant performance improvements. We observed that it has a restriction of using document-level information in determining novelty but assumed that the document-level information is more important. As a solution, this paper proposed two methods of effectively incorporating document-level information using a CNN-based novelty detection model. Our methods focus on constructing a feature vector of a target document to be classified by extracting relative information between the target document and source documents given as evidence. A series of experiments showed the superiority of our methods on a standard benchmark collection, TAP-DLND 1.0.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.11
• /
• pp.1341-1348
• /
• 2016

In this paper, we propose method to enhance security performance using HT-STBC with artificial noise under Wier-Tap channel model that exist with legitimate receiver and illegal eavesdropper. Conventional STBC with artificial noise scheme has a weakness that a limited increase in the BER of the difference between the receiver and an eavesdropper, when used over QPSK modulation. To solve this problem, we suggest HT-STBC combining hadamard transform and STBC with artificial noise for reduce BER of receiver than the conventional scheme and demonstrated through simulation that also increased BER difference between the receiver and an eavesdropper. By the simulation results, when used proposed scheme, showed approximately 3dB improvement in performance compared to the conventional scheme.

• Journal of the Korean Society for Railway
• /
• v.19 no.2
• /
• pp.159-169
• /
• 2016

The current railway control system in Korea is comprised of signaling, electric rail power, communication, and maintenance systems that are independent of each other. Further, these systems have different mediums and protocols for transmitting the field equipment data to the central control system. The Safety Supervision System has as its purpose the collecting of safety-related data from each system to predict and prevent accidents, this system utilizes standard protocol. Safety-related data need to be collected from field data transmission devices of the existing control system, the data should be collected without affecting the communication of the existing system. In this study, sniffing skill, which is typically used for network traffic monitoring or security, is used to collect data. The problems arising from the use of sniffing devices are noted, and the Packet Conversion Node is proposed as a solution to the problems. Further, functional and performance testing were completed for the prototype, and the software architecture and packet conversion process were verified.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.322-323
• /
• 2022

The most necessary thing for the optimal operation of a water purification plant is to accurately predict the pattern and amount of tap water used by consumers. The required amount of tap water should be delivered to the drain using a pump and stored, and the required flow rate should be supplied in a timely manner using the minimum amount of electrical energy. The short-term demand forecasting required from the point of view of energy optimization operation among water purification plant volume predictions has been made in consideration of seasons, major periods, and regional characteristics using time series analysis, regression analysis, and neural network algorithms. In this paper, we analyzed energy management methods through AI-based complex sensor applicability analysis such as LSTM (Long Short-Term Memory) and GRU (Gated Recurrent Units), which are types of cyclic neural networks.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.05a
• /
• pp.712-716
• /
• 2003

본 논문은 신경망을 이용한 간섭 신호 제어로써 복합 다중 퍼셉트론에서 DS/SS 이동 통신에서의 수신된 신호들을 역전파 학습알고리즘을 이용하여 검출하는 것에 대하여 연구한다. 수신 신호가 일정한 비트율을 갖는 채널에 전송하기 위하여 신경망을 이용한 새로운 탭 가중치 갱신 제어 방법을 제안한다. 적응 횡단선 필터는 심볼간의 채널에 발생하는 상호 심볼간 간섭을 억압하기 위해 LMS 알고리즘 사용한다. 이 알고리즘은 원하는 응답과 실제 출력간의 차인 에러를 이용하여 탭 가중치 조절 메카니즘을 통해 탭 가중치를 갱신함으로서 효과적으로 간섭을 제거하였다. 본 논문은 상호 심볼간 간섭을 효율적으로 억압해온 기존의 LMS 알고리즘에 다계층 퍼셉트론 신경망을 조합한 새로운 BISP 알고리즘을 제안하였으며, 제안된 알고리즘을 통해 탭 가중치 갱신이 보다 효율적으로 이루어짐을 알 수 있다. 시뮬레이션 결과를 통해 제안된 알고리즘의 평균 자승 에러의 수렴 특성이 기존 LMS 알고리즘을 이용한 수렴특성보다 우월하다는 것을 나타내었다.

• Proceedings of the KIEE Conference
• /
• 2008.07a
• /
• pp.27-28
• /
• 2008

For stable and sustainable energy supply, distributed generator (DG) has become an essential and indispensable element from environmental and energy security perspectives. However, installation of DG in distribution systems may cause negative affects on feeders because power outputs of DG could be changed irregularly. One of major negative affects is variation in voltage profile. In general, voltage regulation devices such as under load tap changer (ULTC) at distribution substation and step voltage regulator (SVR) along feeder in distribution system are used to maintain customers' receiving voltage within a predetermined range. These regulators are controlled by line drop compensation (LDC) method which calls for two parameters; the equivalent impedance and the load center voltage. Therefore, consideration of DG outputs in the LDC parameter design procedure may give large impact on the installable DG capacity. This paper proposes a method that estimates maximum Installable DG capacity considering LDC parameters of ULTC and SVR. The proposed algorithm is tested with model network.

• Journal of Korean Society of Water and Wastewater
• /
• v.11 no.3
• /
• pp.59-66
• /
• 1997

Although it produces well-treated water in water treatment plant, water quality at the tap can be changed depending on the state of pipes. It is because water quality deteriorates as plant water passes through pipeline networks. Therefore, the improvement of not only water treatment technology but also O & M of water pipelines is required to supply good water to consumers. The purpose of the study was to obtain the basic data of control technology for water quality in pipes through investigating water quality in distribution system. We selected 11 sampling sites and investigated water quality from plant to endpoint of distribution system. we also simulated decreasing tendency of free chlorine through pipeline network. As the result of water quality test, all parameters were below allowable levels, but some parameters had the possibility of being over levels. So there must be more work to set up proper countermeasure for violable parameters.

• Korean journal of communication and information
• /
• v.56
• /
• pp.194-218
• /
• 2011

Focusing on the individuals' structural positions and roles in the internet discussion network, this research explores whether and how the opinion leaders' network characteristics are associated with the message quality and interpersonal influence in terms of attention-drawing and response-generation, which prior studies often failed to fully explicate. Findings suggest that discussion participants with high message quality occupy more central positions in the discussion network, thus enjoy more attention and responses of other following participants. However, opinion leader's network centralities, which tap the structural positions and unique roles in the online discussion network, systematically mediate the effect of the message quality on interpersonal influence. Moreover, significant interaction between opinion perception and network centrality was found only on the majority opinion group, rendering the entire discussion structure toward more enclaved deliberation and group polarization. Taken together, the results imply that the influence of the online opinion leader can only be substantiated with participant's central positions in the discussion network, which has been ignored by the prior opinion leadership research.

• International Journal of Computer Science & Network Security
• /
• v.23 no.10
• /
• pp.1-10
• /
• 2023

This paper proposes a method to extend Inter-Carrier Interference (ICI) canceling Orthogonal Frequency Division Multiplexing (OFDM) receivers for 5G mobile systems to spatial multiplexing 2×2 MIMO (Multiple Input Multiple Output) systems to support high-speed ground transportation services by linear motor cars traveling at 500 km/h. In Japan, linear-motor high-speed ground transportation service is scheduled to begin in 2027. To expand the coverage area of base stations, 5G mobile systems in high-speed moving trains will have multiple base station antennas transmitting the same downlink (DL) signal, forming an expanded cell size along the train rails. 5G terminals in a fast-moving train can cause the forward and backward antenna signals to be Doppler-shifted in opposite directions, so the receiver in the train may have trouble estimating the exact channel transfer function (CTF) for demodulation. A receiver in such high-speed train sees the transmission channel which is composed of multiple Doppler-shifted propagation paths. Then, a loss of sub-carrier orthogonality due to Doppler-spread channels causes ICI. The ICI Canceller is realized by the following three steps. First, using the Demodulation Reference Symbol (DMRS) pilot signals, it analyzes three parameters such as attenuation, relative delay, and Doppler-shift of each multi-path component. Secondly, based on the sets of three parameters, Channel Transfer Function (CTF) of sender sub-carrier number n to receiver sub-carrier number l is generated. In case of n≠l, the CTF corresponds to ICI factor. Thirdly, since ICI factor is obtained, by applying ICI reverse operation by Multi-Tap Equalizer, ICI canceling can be realized. ICI canceling performance has been simulated assuming severe channel condition such as 500 km/h, 8 path reverse Doppler Shift for QPSK, 16QAM, 64QAM and 256QAM modulations. In particular, 2×2MIMO QPSK and 16QAM modulation schemes, BER (Bit Error Rate) improvement was observed when the number of taps in the multi-tap equalizer was set to 31 or more taps, at a moving speed of 500 km/h and in an 8-pass reverse doppler shift environment.

• Proceedings of the IEEK Conference
• /
• 2002.06b
• /
• pp.305-308
• /
• 2002

An 1.8V 6bit IGSPS ADC for high speed data acquisition is discussed in this paper. This ADC is based on a flash ADC architecture because the flash ADC is the only practical architecture at conversion rates of IGSPS and beyond. A straightforward 6bit full flash A/D converter consists of two resistive ladders with 63 laps, 63 comparators and digital blocks. One important source of errors in flash A/D converter is caused by the capacitive feedthrough of the high frequency input signal to the resistive reference-lauder. Consequently. the voltage at each tap of the ladder network can change its nominal DC value. This means large transistors have a large parasitic capacitance. Therefore, a dual resistive ladder with capacitor is employed to fix the DC value. Each resistive ladder generates 32 clean reference voltages which alternates with each other. And a two-stage amplifier is also used to reduce the effect of the capacitive feedthrough by minimizing the size of MOS connected to reference voltage. The proposed ADC is based on 0.18${\mu}{\textrm}{m}$ 1-poly 6-metal n-well CMOS technology, and it consumes 307㎽ at 1.8V power supply.