• Journal of KIISE:Software and Applications
• /
• v.36 no.3
• /
• pp.179-186
• /
• 2009

This paper proposes a target speech extraction which restores speech signal of a target speaker form noisy convolutive mixture of speech and an interference source. We assume that the target speaker is known and his/her utterances are available in the training time. Incorporating the additional information extracted from the training utterances into the separation, we combine convolutive blind source separation(CBSS) and non-negative decomposition techniques, e.g., probabilistic latent variable model. The nonnegative decomposition is used to learn a set of bases from the spectrogram of the training utterances, where the bases represent the spectral information corresponding to the target speaker. Based on the learned spectral bases, our method provides two postprocessing steps for CBSS. Channel selection step finds a desirable output channel from CBSS, which dominantly contains the target speech. Reconstruct step recovers the original spectrogram of the target speech from the selected output channel so that the remained interference source and background noise are suppressed. Experimental results show that our method substantially improves the separation results of CBSS and, as a result, successfully recovers the target speech.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.22 no.5
• /
• pp.501-509
• /
• 2011

Software defined radio(SDR) has a goal that places the analog-to-digital converter(ADC) as near the antenna as possible. But current technique actually can't do analog-to-digital converting about RF band signals. So one method is studying that samples RF band signals to IF band. One of the ways Sub-Sampling technique can convert signals from RF band to IF band without oscillator. If Sub-Sampling technique is used, over 2 bands can convert signals from RF band to IF band. But due to the filter performance in RF band, it is possible to generate interference between signals that is converted in low frequency band. The effect degrades performance. In this paper, we propose one method that uses time division multiplexing(TDM) method as a solution to avoid interference between signals. By doing TDM and Sub-Sampling at the same time that method can get signals without large changes of structures.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2003.11a
• /
• pp.730-734
• /
• 2003

A beamforming method is a kind of spatial filter that focus the amy's signal capturing abilities in a desired direction. In this paper, we detect the location of a source under reflective condition using the multi-dimensional MUSIC algorithm then, we can iud the image source locations from the experimental geometry, and we reconstruct the partial fields for direct wave and reflected wave by using Adaptive nulling algorithm. Numerical simulations are performed to verify its performance under various conditions.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.6
• /
• pp.98-105
• /
• 2003

We propose a new technique of deinterleaving multiple radar pulse sequences by means of genetic algorithm for threat identification in electronic warfare(EW) system. The conventional approaches based on histogram or continuous wavelet transform are so deterministic that they are subject to failing in detection of individual signal characteristics under real EW signal environment that suffers frequent signal missing, noise, and counter-EW signal. The proposed algorithm utilizes the probabilistic optimization procedure of genetic algorithm. This method, a time-of-arrival(TOA) only strategy, constructs an initial chromosome set using the difference of TOA. To evaluate the fitness of each gene, the defined pulse phase is considered. Since it is rare to meet with a single radar at a moment in EW field of combat, multiple solutions are to be derived in the final stage. Therefore it is designed to terminate genetic process at the prematured generation followed by a chromosome grouping. Experimental results for simulated and real radar signals show the improved performance in estimating both the number of radar and the pulse repetition interval.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.05a
• /
• pp.1003-1008
• /
• 2001

In beamforming method, source positions are predicted by MUSIC (Multiple Signal Classification) power method and composite sound fields can then be decomposed into each partial field by beamforming, detenninistically without restriction of the distance between reference microphones and sources. However, reference microphone array shape is important in both MUSIC and beamforming method. Thus the present paper describes the effect of the reference microphone array shape.

• Journal of Korea Multimedia Society
• /
• v.5 no.5
• /
• pp.582-589
• /
• 2002

In this paper, we propose a new method that combines antenna array, which employs any beamforming algorithm, and successive interference cancelling multiuser detector which improves the performance of the receiver and analyse its performance. Antenna away receivers can improve the performance by spatially discriminating among the users and reducing the multiple access interference. But the performance of the receivers degrades when the difference of the direction of arrival of the users are too small. Many interference cancelling multiuser detectors have been suggested for DS/CDMA systems which have low complexity for their performance. In this paper, we show how to mitigate such a degradation by combining successive interference cancelling multiuser detector.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.587-590
• /
• 2001

본 논문에서는 스마트 안테나를 적용하여 서로 다른 대역폭(1.25MHz, 5Mhz)과 서로 다른 환경(rural, urban)에서 W-CDMA 시스템의 성능을 분석하였다. 먼저 스마트 안테나 시스템의 성능을 분석하기 위해서 시·공간 상관특성을 갖는 채널이 필요하다. 이를 위해 시간 정보를 고려하기 위해 시변 선형 필터로 모델링하였고, 공간 정보를 고려하기 위하여 배열 응답 벡터를 삽입하였다. 이 때 공간적으로는 일정량의 각도 퍼짐을 갖는 부-다중경로 성분으로 구성된 한 개의 반사파가 한 방향으로부터만 수신된다고 가정하였다. 또한 한 칩 내에 들어오는 다중 경로 성분을 분리하여 가장 큰 신호 성분만을 찾아내어 원하는 사용자의 신호로 사용하는 방법으로 모의 실험을 실행하였다. 모의실험 결과, 협대역 CDMA를 사용하는 스마트 안테나 시스템보다 광대역 CDMA를 사용하는 스마트 안테나 시스템이 더 나은 성능을 보임을 확인할 수 있었다. 또한 스마트 안테나는 2D-RAKE receiver를 사용했을 때 urban area에서 보다 더 효과적임을 확인할 수 있었다.

• Journal of the Korean Magnetics Society
• /
• v.26 no.1
• /
• pp.24-30
• /
• 2016

MFL (Magnetic flux leakage) type nondestructive testing has been used for inspection of underground gas pipelines to find metal defects by detecting magnetic leakage signal. Because the underground gas pipeline is exposed by environment such as high pressure with great humidity, external defects are easily formed on the surface of pipelines and they are being grouped respectively. These adjacent defects cause the signal distortion of leakage flux so that it is hard to estimate the shape information of defects. In this paper, we performed to study of the signal distortion and compensating method for multiple defects in MFL type nondestructive testing system by using 3D FEM simulation. This paper proposes the basic algorithm of defect signal analysis on multiple defects on the surface of 30 inch diameter pipeline.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.10C
• /
• pp.965-974
• /
• 2007

In this paper, we proposed the feature selection algorithm for multi-class genre classification. In our proposed algorithm, we developed GMM separation score based on Gaussian mixture model for measuring separability between two genres. Additionally, we improved feature subset selection algorithm based on sequential forward selection for multi-class genre classification. Instead of setting criterion as entire genre separability measures, we set criterion as worst genre separability measure for each sequential selection step. In order to assess the performance proposed algorithm, we extracted various features which represent characteristics such as timbre, rhythm, pitch and so on. Then, we investigate classification performance by GMM classifier and k-NN classifier for selected features using conventional algorithm and proposed algorithm. Proposed algorithm showed improved performance in classification accuracy up to 10 percent for classification experiments of low dimension feature vector especially.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.12
• /
• pp.2303-2312
• /
• 1994

A high resolution algorithm is presented for resolving multiple narrowband plane waves that are incident on an equispaced linear array. To overcome the deleterious effects due to coherent sources, a number of noise-eigenvector-based approaches have been proposed for narrowband signal processing. For differing reasons, each f these methods provide a less than satisfactory resolution of the coherency problem. The proposed algorithm makes use of fundamental property possessed by those eigenvectors of the spatial covariance matrix that are associated with eigenvalues that are larger than the sensor noise level. This property is then used to solve the incoherent and coherent sources incident on an equispaced linear array. Simulation results are shown to illustrate the high resolution performance achieved with this new approach relative to that obtained with MUSIC and spatial smoothed MUSIC.