• 한국멀티미디어학회논문지
• /
• 제11권7호
• /
• pp.986-996
• /
• 2008

전력선 통신을 이용한 홈 네트워크 환경에서는 가전 기기의 노이즈 영향으로 인해 네트워크의 일부 구간에 부하가 증가되어 송신 노드와 수신 노드 간에 갑자기 통신이 불가능한 상황이 발생될 수도 있다. 본 논문에서는 전력선 기반 홈 네트워크 환경에서 네트워크 부하 증가로 인한 통신 장애가 발생하는 경우 자동적으로 중계 가능한 노드들을 찾아 스패닝 트리 알고리즘에 따른 필요 최소한의 중계 노드들을 선발해 이들 노드들의 중계를 통해 수신 노드까지 패킷을 전달 가능하도록 하는 자동 중계 방식을 제안한다. 이 방식은 시뮬레이터 개발을 통한 성능 분석 결과 미국의 에쉴론 사의 자동 리피터 선택 기술에 비해 중계 노드들을 찾는 과정에서 노드간의 패킷 교환 횟수는 다소 증가하나 중계를 통해 수신 노드를 발견할 수 있는 확률은 거의 두 배 가까이 높아짐을 알 수 있다.

• 한국정보통신학회논문지
• /
• 제23권9호
• /
• pp.1060-1069
• /
• 2019

보행자 기반의 카메라 자가 보정 방법들은 복잡한 보정 장치나 절차가 필요하지 않기 때문에 비디오 감시 시스템에 적합하다. 하지만 임의 보행자를 보정 대상으로 사용하는 경우 보행자들의 키를 모르기 때문에 보정 정확도가 저하될 수 있다. 본 논문은 실제 감시 환경에서 이 문제를 해결하기 위한 베이지안 보정 방법을 제안한다. 제안하는 방법에서는 감시 지역 사람들의 키에 대한 통계가 있다고 가정하고, 발-머리 호몰로지(foot-head homology)를 사용하여, 발과 머리의 좌표와 보행자 키의 불확실성을 모두 고려하는 확률 모델을 구성한다. 이 확률 모델을 직접 푸는 것은 난해하므로, 본 연구에서는 근사적 방법인 변분 베이지안 추론(variational Bayesian inference)을 사용한다. 따라서, 이를 통해 관측된 보행자들의 키를 추정함과 동시에 정확한 카메라 파라미터를 구할 수 있다. 다양한 실험을 통해 제안된 방법이 노이즈에 강하며, 보정에 대한 정확한 신뢰도를 제공함을 보였다.

• 정보보호학회논문지
• /
• 제33권6호
• /
• pp.907-917
• /
• 2023

최근 자율주행 자동차는 운전자 지원 시스템에 딥러닝 기술을 적용하여 운전자에게 편의성을 제공하고 있지만, 딥러닝 기술이 적대적 회피 공격(adversarial evasion attacks)에 취약함이 밝혀졌다. 본 논문에서는 객체 인식 알고리즘인 YOLOv5(You Only Look Once)를 대상으로 MI-FGSM (Momentum Iterative-Fast Gradient Sign Method)를 포함한 5가지 적대적 회피 공격을 수행하였으며 객체 탐지 성능을 mAP(mean Average Precision)로 측정하였다. 특히, 본 논문에서는 모폴로지 연산을 적용하여 적대적 공격으로부터 노이즈를 제거하고 경계선을 추출하여 YOLO가 객체를 정상적 탐지할 수 있는 방안을 제안하고 이를 실험을 통해 그 성능을 분석하였다. 실험 결과, 적대적 공격을 수행했을 때 YOLO의 mAP가 최소 7.9%까지 떨어져 YOLO가 객체를 정확하게 탐지하지 못하는 것을 87.3%까지 성능을 개선하였다.

• 오토저널
• /
• 제14권6호
• /
• pp.48-59
• /
• 1992

This study is on the performance of the perforated tube muffler when it operates as an exhaust silencer with through-flow, steady or pulsating. Theoretical estimation of the insertion loss was made by means of transfer matrix and by using the impedance equation for the perforated tube obtained for the case of low-speed steady through-flow. Experiment was performed for the measurement of the insertion loss at two flow conditions. The one is a steady flow from the exhaust pipe of an idling diesel engine. The effect of the through-flow velocity and steadiness on the muffler performance was obtained. By comparing the theoretical prediction with the experimental result, the validity of the impedance equation in the theoretical model was discussed. It has been found that steadiness as well as magnitude of the through-flow has a significant effect on the performance of the perforated tube muffler. Especially, the self-noise due to the pulsating flow in the engine exhaust system must be taken into account for the prediction of the muffler performance.

• 대한전기학회논문지
• /
• 제39권12호
• /
• pp.1306-1316
• /
• 1990

A controller using a multilayer neural network is proposed to the dynamic control of a PUMA 560 robot arm. This controller is developed based on an error back-propagation (BP) neural network. Since the neural network can model an arbitrary nonlinear mapping, it is used as a commanded feedforward torque generator. A Proportional Derivative (PD) feedback controller is used in parallel with the feedforward neural network to train the system. The neural network was trained by the current state of the manipulator as well as the PD feedback error torque. No a priori knowledge on system dynamics is needed and this information is rather implicitly stored in the interconnection weights of the neural network. In another experiment, the neural network was trained with the current, past and future positions only without any use of velocity sensors. Form this thim window of position values, BP network implicitly filters out the velocity and acceleration components for each joint. Computer simulation demonstrates such powerful characteristics of the neurocontroller as adaptation to changing environments, robustness to sensor noise, and continuous performance improvement with self-learning.

• 대한전기학회논문지
• /
• 제36권5호
• /
• pp.367-376
• /
• 1987

In this paper we have disigned and implemented a real time ALE (adaptive line enhancer) using a high speed digital processor,NEC 7720. For the ALE system, we have employed an adaptive LMS(least mean square) algorithm proposed by Widrow and Hoff and a 32-order FIR(finite impulse response) filter. Extensive computer simulations have been performed to investigate the peformance of the ALE and to determine necessary parameters for hardware design. The developed software for an NEC 7720 was tested in real time operation using an NEC7720 hardware emulator. The ALE has been tested by sinusoidal waves and real CW (continuous wave) signals. It was found that the experimental results were well agreed with the computer simulation results. Thus it can be concluded that the ALE is useful for detection and enhancement of a sinusoidal signal which is corrupted by an additive Gaussian noise.

• Journal of the Korean Data and Information Science Society
• /
• 제13권2호
• /
• pp.307-315
• /
• 2002

Suppose one is given a vector X of a finite set of quantities $X_i$ which are independent Poisson signals. A null hypothesis $H_0$ about E(X) is to be tested against an alternative hypothesis $H_1$. A quantity $$\sum\limits_{i}\omega_ix_i$$ is to be computed and used for the test. The optimal values of $\omega_i$ are calculated for three cases : (1) signal to noise ratio is used in the test, (2) normal approximations with unequal variances to the Poisson distributions are used in the test, and (3) the Poisson distribution it self is used. A comparison is made of the optimal values of $\omega_i$ in the three cases as parameter goes to infinity.

• 소음진동
• /
• 제4권3호
• /
• pp.365-375
• /
• 1994

Periodic longitudinal vibrations of a video and audio tape and the like used for transducing pictures and sounds may be induced by self-excited vibrations which are caused by frictions against the heads and guides, and their eccentricity, noncircularity or irregularity. It is important to analyze the longitudinal vibration of a tape because it causes distortions of a reproduced signal. It is difficult to measure directly the longitudinal vibration. In this paper a method estimating longitudinal vibration using impulse responses in the transverse direction is presented. And, the parameters boundary domain where the transverse and longitudinal vibration can be decoupled is introduced. In the domain where the both vibrations are decoupled, analytic methods using frequency characteristics and transient responses of the transverse vibration, respectively, are presented. The time domain method predicted more exactly the instantaneous phase of the longitudinal vibration than frequency domain method did.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2008년도 추계학술대회 논문집 전력기술부문
• /
• pp.141-143
• /
• 2008

Due to the demand for high-output motors in the limited space between the wheels in an electric train, self-ventilating traction motors have been used for many years. periodical disassembly maintenance is necessary to remove the small quantities of dust that enter the motor from the open-air ventilation. Reducing this burden, as well as increasing efficiency and reducing noise, would benefit the next generation of moters To address these needs, KRRI is developing a fully enclosed type traction motor, a fully enclosed type traction motor with outer fans, a high-efficiency permanent-magnet synchronous motor and a direct-drive motor(DDM) as traction motors for the next generation of trains.

• 진공이야기
• /
• 제3권1호
• /
• pp.16-21
• /
• 2016

FinFETs are able to be scaled down to 22 nm and beyond while suppressing effectively short channel effect, and have superior performance compared to 2-dimensional (2-D) MOSFETs. Bulk FinFETs are built on bulk Si wafers which have less defect density and lower cost than SOI(Silicon-On-Insulator) wafers. In contrast to SOI FinFETs, bulk FinFETs have no floating body effect and better heat transfer rate to the substrate while keeping nearly the same scalability. The bulk FinFET has been developed at 14 nm technology node, and applied in mass production of AP and CPU since 2015. In the development of the bulk FinFETs at 10 nm and beyond, self-heating effects (SHE) is becoming important. Accurate control of device geometry and threshold voltage between devices is also important. The random telegraph noise (RTN) would be problematic in scaled FinFET which has narrow fin width and small fin height.