• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.11
• /
• pp.2547-2552
• /
• 2013

To achieve better performance in dynamic spectrum access networks, each secondary user tracks channels and chooses a good channel to transmit its packet. When all secondary users adopt the same dynamic spectrum access policy, they have similar channel information, which leads secondary users to choose the same channel and more collisions among them. To relieve this problem, we propose an access policy using belief vector tracking method for other competing secondary users. Simulation results are provided to show that the proposed policy yields better performance than the existing policies which do not take into account what other secondary users are doing.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.16 no.6
• /
• pp.610-617
• /
• 2011

This paper proposes a novel and direct dead-time compensation method of the 3-phase inverter using space vector pulse width modulation(SVPWM) topology. The proposed dead-time compensation method directly compensates the dead-time to the turn-on time of the effective voltage vector according to the current direction of the medium voltage reference. Each phase voltages are determined by the switching times of the effective voltage vectors, and the practical switching times have loss according to the current direction by the dead-time effect in the 3-phase inverter. The proposed method adds the dead-time to the switching time of the effective voltage vector according to the current direction, so it does not require complex d-q transform and controller to compensate the voltage error. The proposed dead-time compensation scheme is verified by the computer simulation and experiments of 3-phase R-L load.

• Journal of the Korean Society for information Management
• /
• v.22 no.3 s.57
• /
• pp.261-287
• /
• 2005

The purpose of this paper is to explore the ways to improve the performance of SVM (Support Vector Machines) text classifier using inter-document similarities. SVMs are powerful machine learning systems, which are considered as the state-of-the-art technique for automatic document classification. In this paper text categorization via SVMs approach based on feature representation with document vectors is suggested. In this approach, document vectors instead of index terms are used as features, and vector similarities instead of term weights are used as feature values. Experiments show that SVM classifier with document vector features can improve the document classification performance. For the sake of run-time efficiency, two methods are developed: One is to select document vector features, and the other is to use category centroid vector features instead. Experiments on these two methods show that we can get improved performance with small vector feature set than the performance of conventional methods with index term features.

• Proceedings of the KIPE Conference
• /
• 2007.07a
• /
• pp.153-155
• /
• 2007

본 논문에서는 3상 저주파 변압기를 이용한 절연형 멀티레벨 인버터를 위한 새로운 공간 벡터 제어 방식을 제안한다. 제안된 제어 방식은 기존의 미리 계산된 표를 이용한 방식이 아닌 완전히 프로그래밍에 의한 방식으로 구현 시간은 기존 방식에 비해 늘이지 않으며, 확장성이 용의하다는 장점이 있다. 제안된 제어 방식을 공통암을 이용한 3상 IHCML 인버터에 적용하였으며, Matlab을 이용한 시뮬레이션 및 실험실 차원의 인버터 제작을 통하여 제안한 방식의 타당성을 검증 하였다.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.9
• /
• pp.38-48
• /
• 1998

Generating test patterns for asynchronous sequential circuits remains to be a very difficult problem. There are few algorithms for this problem, and previous works cut feedback loops, and insert synchronous flip-flops in the feedback loops during ATPG. The conventional algorithms are similar to the algorithms for synchronous sequential circuits. This means that the conventional algorithms generate test patterns by modeling asynchronous sequential circuits as synchronous sequential circuits. So, test patterns generated by those algorithms nay not detect target faults when the test patterns are applied to the asynchronous sequential circuit under test. In this paper an algorithm is presented to generate test patterns for asynchronous sequential circuits. Test patterns generated by the algorithm can detect target faults for asynchronous sequential circuits with the minimal possibility of critical race problem and oscillation. And it is guaranteed that the test patterns generated by the algorithm will detect target faults.

• Journal of Advanced Marine Engineering and Technology
• /
• v.35 no.8
• /
• pp.1105-1110
• /
• 2011

This paper presents an improved direct torque control based on artificial neural networks technique. The major problem that is usually associated with DTC drive is the high torque(speed) ripple. To overcome this problem a torque hysteresis band with variable amplitude is proposed based on artificial neural networks. The artificial neural networks proposed controller is shown to be able to reducing the torque(speed) ripple and dependency on motor parameter and to improve performance DTC especially at high speed and reversal running.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.06a
• /
• pp.79-80
• /
• 2014

본 논문에서는 다방향 패턴을 이용한 구조광(structured light)방식의 3차원 스캔 기법을 적용하여 3차원 스캐너의 3차원 형상 취득의 정밀도를 높이는 방법을 제안한다. 기존의 단방향 패턴을 이용한 3차원 구조광 방식의 스캔 기법으로 3차원 형상을 취득하는 경우, 스캔 대상 표면의 법선벡터가 패턴의 방향과 수평 할수록 부정확한 형상이 복원된다. 본 논문에서는 스캔 대상의 경사면에 따른 법선 벡터(normal vector) 검출 및 검출된 벡터를 통한 최적의 패턴 방향 선출(quantization), 그리고 각 화소의 최적의 패턴방향을 이용한 선별적인 구조광 방식의 3차원 스캔 기법을 통하여 3차원 형상 취득의 정확성을 높이는 기법을 제안한다.

• Journal of the Korea Computer Industry Society
• /
• v.5 no.9
• /
• pp.987-992
• /
• 2004

There is the temporal correlation of the video sequence between the motion vector of current block and the motion vector of previous block. In this paper, we propose the improved diamond search pattern using an motion vector prediction candidate search point by the predicted motion information from the same block of the previous frame. Simulation results show that PSNR(Peak-to-Signal Noise Ratio) values are improves as high as high as 14~24% in terms of average number of search point per motion vector estimation and improved about 0.02~0.37dB on an average except the full search(FS) algorithm.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2011.07a
• /
• pp.403-405
• /
• 2011

모바일 영상 기기들에 사용되는 저전력, 고화질의 이미지 센서는 Rolling Shutter 방식을 사용하는 CIS(CMOS Image Sensor)이다. 구조적 특성에 의해 Rolling Shutter 방식을 사용하는 CIS는 촬영 환경에 의해 결과물에 왜곡을 일으키게 된다. 본 논문은 Rolling Shutter에 의한 왜곡을 분석하고 촬영된 영상의 Rolling Shutter 왜곡 모션벡터를 예측하는 알고리즘을 제안하였다. 또한 제안 알고리즘의 정확성을 높이기 위해 Sub-block과 Kalman-filter를 적용하여 Sub-pixel 단위로 계산하여 기존의 Lucas-Kanade 알고리즘 보다 효율적인 모션벡터 예측 알고리즘을 제안한다. 또한 보다 정확한 성능 변화를 추적하기 위하여 MSE 방식을 사용하여 비교 분석 하였다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1998.11a
• /
• pp.146-150
• /
• 1998

In this paper, we propose a nonlinear system identification method using the genetic algorithm. We represent the nonlinear system as a parameter vector and a measurement vector of ARMAX model. In order to identify the nonlinear system, we find the parameter vector using the genetic algorithm. The parameter vector is regarded as a chromosome of gene. The error between the desired output and estimated output every sampling period is used to calculate the fitness of one gene. The simulation results showed the effectiveness of using the genetic algorithm in the nonlinear system identification.