• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.2
• /
• pp.96-107
• /
• 1999

The generalized sidelobe cancellers(GSC's) ar used for suppressing an interference in array radar. The frequency-domain GSC's have a faster convergence rate than the time-domain GSC's because they remove the correlation between the interferences using a frequency-domain least mean square(LMS) algorithm. However, we have not fully used the advantage of the frequency-domain GSC's since we have always updated the weights of all frequency bins, even the interferer free frequency bin. In this paper, we propose a new frequency-domain GSC based on constant false-alarm rate(CFAR) detector, of which GSC adaptively determine the bin whose weight is updated according to the power of each frequency bin. This canceller updates the weight of only updated according to the power of each frequency bin. This canceller updates the weight of only the bin of which the power is high because of the interference signal. The computer simulation shows that the new GSC reduces the iteration number for convergence over the conventional GSC's by more than 100 iterations. The signal-to-noise ration(SNR) improvement is more than 5 dB. Moreover, the number of renewal weights required for the adaptation is much fewer than that of the conventional one.

• Journal of Biomedical Engineering Research
• /
• v.23 no.1
• /
• pp.1-8
• /
• 2002

Atrial fibrillation(AF). chronic or paroxysmal is the most frequent arrhythmia in human subjects Duration of P wave in signal-averaged electrocardiography(SAECG) reflects intra-atrial conduction time and therefore. could be used as an electrophysiological marker for atrial conduction chance at the earthy stave. So we apply the analysis method using SAECG to diagnose Paroxysmal atrial fibrillation(PAF) . Subjects Participated for the study consisted of two groups: a control group(n=34) of normal healthy volunteers and a group of AF Patients(n=38) with a documented history of PAF but no other history of cardiac disease. We evaluated the effect of several filtering and determination methods to find the starting and ending feints of the P wavy on its duration. To increase the measurement reliability of P wave duration. the automatic detection method was proposed. Also. to increase the detection rate for PAF risk, the decision threshold value was optimized using receiver operation characteristics(ROC) curve. Results showed that the highest statistical difference (p〈0.001) of the P wane duration between controls and subjects was obtained at the Processing condition, using absolute threshold vague(8.75 $\mu N$) , a least mean square(LMS) high pass filter and 30 Hz cutoff frequency. The most outstanding difference(sensitivity 88 % specificity 64.4 %) between controls and subjects was obtained at the decision threshold value of 112 ms.

• The Journal of the Acoustical Society of Korea
• /
• v.7 no.4
• /
• pp.54-76
• /
• 1988

In Part Ⅰ of the paper, we have developed various block least mean-square (BLMS) adaptive digital filters (ADF's) based on a unified matrix treatment. In Part Ⅱ we analyze the convergence behaviors of the self-orthogonalizing frequency-domain BLMS (FBLMS) ADF and the unconstrained FBLMS (UFBLMS) ADF both for the overlap-save and overlap-add sectioning methods. We first show that, unlike the FBLMS ADF with a constant convergence factor, the convergence behavior of the self-orthogonalizing FBLMS ADF is governed by the same autocorrelation matrix as that of the UFBLMS ADF. We then show that the optimum solution of the UFBLMS ADF is the same as that of the constrained FBLMS ADF when the filter length is sufficiently long. The mean of the weight vector of the UFBLMS ADF is also shown to converge to the optimum Wiener weight vector under a proper condition. However, the steady-state mean-squared error(MSE) of the UFBLMS ADF turns out to be slightly worse than that of the constrained algorithm if the same convergence constant is used in both cases. On the other hand, when the filter length is not sufficiently long, while the constrained FBLMS ADF yields poor performance, the performance of the UFBLMS ADF can be improved to some extent by utilizing its extended filter-length capability. As for the self-orthogonalizing FBLMS ADF, we study how we can approximate the autocorrelation matrix by a diagonal matrix in the frequency domain. We also analyze the steady-state MSE's of the self-orthogonalizing FBLMS ADF's with and without the constant. Finally, we present various simulation results to verify our analytical results.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.4
• /
• pp.691-698
• /
• 2003

We extend the sue of the method of least square to develop a recursive algorithm for the design of adaptive transversal filters such that, given the least-square estimate of this vector of the filter at iteration n-l, we may compute the updated estimate of this vector at iteration n upon the arrival of new data. We begin the development of the RLS algorithm by reviewing some basic relations that pertain to the method of least squares. Then, by exploiting a relation in matrix algebra known as the matrix inversion lemma, we develop the RLS algorithm. An important feature of the RLS algorithm is that it utilizes information contained in the input data, extending back to the instant of time when the algorithm is initiated. In this paper, we propose new tap weight updated RLS algorithm in adaptive transversal filter with data-recycling buffer structure. We prove that convergence speed of learning curve of RLS algorithm with data-recycling buffer is faster than it of exiting RLS algorithm to mean square error versus iteration number. Also the resulting rate of convergence is typically an order of magnitude faster than the simple LMS algorithm. We show that the number of desired sample is portion to increase to converge the specified value from the three dimension simulation result of mean square error according to the degree of channel amplitude distortion and data-recycle buffer number. This improvement of convergence character in performance, is achieved at the B times of convergence speed of mean square error increase in data recycle buffer number with new proposed RLS algorithm.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2002.11a
• /
• pp.535-542
• /
• 2002

최근 IT 산업이 급성장하면서 화상 회의, 게임, 채팅 등에서의 아바타(avatar) 제어를 위한 자연스러운 인터페이스 기술이 요구되고 있다. 본 논문에서는 동적 윤곽선 모델(active contour models; snakes)을 이용하여 복잡한 배경이 있는 컬러 CCD 카메라 영상에서 얼굴과 눈, 입, 눈썹, 코 등의 얼굴 요소에 대해 윤곽선을 추출하거나 위치를 파악하는 방법을 제안한다. 일반적으로 스네이크 알고리즘은 잡음에 민감하고 초기 모델을 어떻게 설정하는가에 따라 추출 성능이 크게 좌우되기 때문에 주로 단순한 배경의 영상에서 정면 얼굴의 추출에 사용되어왔다 본 연구에서는 이러한 단점을 파악하기 위해, 먼저 YIQ 색상 모델의 I 성분을 이용한 색상 정보와 차 영상 정보를 사용하여 얼굴의 최소 포함 사각형(minimum enclosing rectangle; MER)을 찾고, 이 얼굴 영역 내에서 기하학적인 위치 정보와 에지 정보를 이용하여 눈, 입, 눈썹, 코의 MER을 설정한다. 그런 다음, 각 요소의 MER 내에서 1차 미분과 2차 미분에 근거한 내부 에너지와 에지에 기반한 영상 에너지를 이용한 스네이크 알고리즘을 적용한다. 이때, 에지 영상에서 얼굴 주변의 복잡한 잡음을 제거하기 위하여 색상 정보 영상과 차 영상에 각각 모폴로지(morphology)의 팽창(dilation) 연산을 적용하고 이들의 AND 결합 영상에 팽창 연산을 다시 적용한 이진 영상을 필터로 사용한다. 총 7명으로부터 양 눈이 보이는 정면 유사 방향의 영상을 20장씩 취득하여 총 140장에 대해 실험한 결과, MER의 오차율은 얼굴, 눈, 입에 대해 각각 6.2%, 11.2%, 9.4%로 나타났다. 또한, 스네이크의 초기 제어점을 얼굴은 44개, 눈은 16개, 입은 24개로 지정하여 MER추출에 성공한 영상에 대해 스네이크 알고리즘을 수행한 결과, 추출된 영역의 오차율은 각각 2.2%, 2.6%, 2.5%로 나타났다.해서 Template-based reasoning 예를 보인다 본 방법론은 검색노력을 줄이고, 검색에 있어 Feasibility와 Admissibility를 보장한다.매김할 수 있는 중요한 계기가 될 것이다.재무/비재무적 지표를 고려한 인공신경망기법의 예측적중률이 높은 것으로 나타났다. 즉, 로지스틱회귀 분석의 재무적 지표모형은 훈련, 시험용이 84.45%, 85.10%인 반면, 재무/비재무적 지표모형은 84.45%, 85.08%로서 거의 동일한 예측적중률을 가졌으나 인공신경망기법 분석에서는 재무적 지표모형이 92.23%, 85.10%인 반면, 재무/비재무적 지표모형에서는 91.12%, 88.06%로서 향상된 예측적중률을 나타내었다.ting LMS according to increasing the step-size parameter $\mu$ in the experimentally computed. learning curve. Also we find that convergence speed of proposed algorithm is increased by (B+1) time proportional to B which B is the number of recycled data buffer without complexity of computation. Adaptive transversal filter with proposed data recycling buffer algorithm could efficiently reject ISI of channel and increase speed of convergence in avoidance burden of computational complexity in reality when it was experimented having the same condition of

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.05a
• /
• pp.231-234
• /
• 2002

셀룰라 신경회로망의 연상 메모리를 이용하여 시각적인 입력 데이터의 연산을 통하여 영상 패턴의 분류와 인식을 수행한다. 셀룰라 신경회로망은 일반적인 신경회로망과 같이 비선형 데이터의 실시간 처리가 가능하고, 세 포자동자와 같이 격자구조의 셀로 이루어져 인접한 셀과 직접 정보를 주고받는다. 응용 분야로는 최적화, 선형/비선형화, 연상 메모리, 패턴인식, 컴퓨터 비젼 등에 적용할 수 있다. 영상의 이미지 픽셀을 셀룰라 신경회로망의 셀에 대응하여 전체 이미지 영상을 모든 셀룰라 신경회로망의 셀에서 동시에 병렬로 처리할 수 있어 2-D 이미지 처리에 적합하다 본 논문은 셀룰라 신경회로망에 의한 연상 메모리 구조를 설계하고, 학습된 하중값 메모리에서 가장 적당한 하중값을 선택하여 학습된 영상과 정확히 일치하는 출력을 얻는 방법을 제시한다. 학습을 통한 연상 메모리 구현에는 각각의 뉴런에서 일정하지 않은 다른 템플릿을 사용한다. 각각의 템플릿은 뉴런들 간의 연결 하중값을 나타내고 학습011 따라 갱신된다. 학습방법으로는 템플릿 하중값 학습에 뉴런들 간의 연결 하중값을 조정하는 가장 단순한 규칙인 Hebb의 학습방법이 사용되었고 분류값 학습에 LMS 알고리즘이 사용되었다

• Journal of the Korean Society for Precision Engineering
• /
• v.26 no.7
• /
• pp.51-57
• /
• 2009

In this paper, we studied on an attenuation effect of automobile exhaust noise according to the direction of canceling speaker in duct-acoustic ANC system. Automobile exhaust noise was recorded at 800rpm, 3500rpm and 5000rpm of a diesel engine. Directions of canceling speaker can be set to $30^{\circ}$, $90^{\circ}$ and $150^{\circ}$ against the primary noise flow by acrylic ducts to be made for the experimentation. DSP board used to control the ANC system. The algorithm of this ANC system applied the Filtered-x-LMS algorithm that is modified to compensate for a property of DSP input signal and the secondary-path effect. As an experiment result, the direction of canceling speaker was proved to influence the reduction effect of noise. The $150^{\circ}$ duct in the attenuation effect of noise showed a better result than the $90^{\circ}$ or $30^{\circ}$ duct.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.19 no.3
• /
• pp.288-293
• /
• 2009

This study presents the active noise control of duct noise. The duct was excited by a steady-state harmonic and white noise force and the control was performed by one control speaker attached to surface of the duct. An adaptive controller based on filtered x LMS(FXLMS) algorithm was used and controller was defined by minimizing the square of the response of the error microphone. The assemble controller, which is called a hybrid ANC(active noise control) system, was combined with feedforward and feedback controller. The feedforward ANC attenuates primary noise that is correlated with the reference signal, while the feedback ANC cancels the narrowband components of the primary noise that are not observed by the reference sensor. Furthermore, in many ANC applications, the periodic components of noise are the most intense and the feedback ANC system has the effect of reducing the spectral peaks of the primary noise, thus easing the burden of the feedforward ANC filter.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.15 no.2
• /
• pp.127-133
• /
• 2004

A feedforward amplifier, which is composed of several components, is an open loop system. Therefore, feedforward amplifiers are apt to deteriorate its performance according to the environmental changes even though the cancellation performance and the linearization bandwidth of feedforward systems are superior to other linearization methods. A control method is needed for maintaining the original performance of feedforward amplifiers or to keep the desired performance within a little error bounds. In this paper, an adaptive control method using the steepest descent algorithm, which has a good convergence characteristic and is easy to implement, is suggested. The characteristics of the suggested control method compare with the characteristics of other control methods and the simulation results are presented.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.12C
• /
• pp.1660-1668
• /
• 2004

This paper proposes a Voice Activity Detection (VAD) algorithm based on Radial Basis Function (RBF) network using dual threshold. The k-means clustering and Least Mean Square (LMS) algorithm are used to upade the RBF network to the underlying speech condition. The inputs for RBF are the three parameters in a Code Exited Linear Prediction (CELP) coder, which works stably under various background noise levels. Dual hangover threshold applies in BRF-VAD for reducing error, because threshold value has trade off effect in VAD decision. The experimental result show that the proposed VAD algorithm achieves better performance than G.729 Annex B at any noise level.