FM변조된 형태의 Kernel을 사용한 음성신호의 시간-주파수 표현 해상도 향상에 관한 연구

이희영,최승호,Lee, He-Young,Choi, Seung-Ho 한국음성학회 2005 음성과학 Vol.12 No.4

Time-frequency representation reveals some useful information about instantaneous frequency, instantaneous bandwidth and boundary of each AM-FM component of a speech signal. In many cases, the instantaneous frequency of each component is not constant. The variability of instantaneous frequency causes degradation of resolution in time-frequency representation. This paper presents a method of adaptively adjusting the transform kernel for preventing degradation of resolution due to time-varying instantaneous frequency. The transform kernel is the form of frequency modulated function. The modulation function in the transform kernel is determined by the estimate of instantaneous frequency which is approximated by first order polynomial at each time instance. Also, the window function is modulated by the estimated instantaneous. frequency for mitigation of fringing. effect. In the proposed method, not only the transform kernel but also the shape and the length of. the window function are adaptively adjusted by the instantaneous frequency of a speech signal.

레고블록 기반 재구성 주파수 선택표면 구조 설계

조성실(Sung-Sil Cho),홍익표(Ic-Pyo Hong) 한국정보기술학회 2021 한국정보기술학회논문지 Vol.19 No.5

In this paper, we design a reconfigurable frequency selective surface structure that can easily implement the desired frequency response characteristics using Lego blocks composed of a metal and a dielectric block. The proposed structure has the advantage of being able to easily reconfigure the operating frequency since various patterns can be easily fabricated by assembly using metal blocks and dielectric blocks. To confirm the possibility that the Lego block-based frequency selective surface structure has the desired frequency performance, a simple structure such as a dipole-type frequency selective surface structure and a cross-type frequency selective surface structure were designed and fabricated. The measured frequency response characteristics resonate frequency has a 9.32GHz for the dipole-type and 6.49GHz for the cross-type frequency-selective surface structure. Error compared with the simulation results is 0.64% for the dipole-type and 0.15GHz for the cross-type frequency-selective surface structure. Using the results of this study, it is possible to implement an indoor electromagnetic wave security space with a frequency selective shielding or absorption function with frequency reconfiguration characteristics.

언더 샘플링과 FFT를 이용한 정현파 신호의 실시간 주파수 검출

박진식(Jin-sik Park),주우용(Woo-yong Ju),정봉식(Bongsik Jeong) 한국정보기술학회 2011 한국정보기술학회논문지 Vol.9 No.12

This paper proposes an algorithm that detects in real time the frequency of a sinusoidal input signal whose frequency varies in a wide range using under-sampling technique and FFT algorithm. A sinusoidal input signal is under-sampled by three different frequencies below the input signal frequency, and then its frequency is detected by applying FFT algorithm to the sampled signals. The gap between these three different sampling frequencies, which is determined by the number of FFT points, sampling frequency and the maximum input frequency, is important for detecting exactly the input frequency. The proposed frequency detection algorithm was modeled and implemented using Mathworks Matlab and Xilinx System Generator, respectively. Through simulations with 25.6㎒ sampling frequency and 0.3㎒ gap of sampling frequency in 100~500㎒ band, the implemented circuit was verified to generate the same results with the Matlab model, and the frequency of input signal with SNR over 14㏈ was detected exactly by the proposed algorithm.

Frequency and Acoustic Reduction in English –ment Derivatives

성제현 한국영어학회 2019 영어학 Vol.19 No.4

This study investigates the influence of frequency on the production of bimorphemic words, and considers which frequency measure is most apt to explain the differences. Previous studies have reported that frequent words are produced faster and more casually than infrequent ones, and that medial segments will yield shorter durations. The present study examines the relation between frequency and the duration of medial segments in English derived words by conducting a production experiment with 6 native speakers of American English using 74 English –ment derivatives, and evaluates whole-word frequency, base frequency, and relative frequency (whole-word frequency divided by base frequency) against one another as predictors. The results show that models incorporating any of the three frequency measures strongly predict medial segment duration. Among the three frequency measures, whole-word frequency explained the most variance, across all consonant types. The duration of segments in highly frequent words tends to be shorter than that in relatively infrequent words. Overall, this study confirms that speakers are sensitive to the extralinguistic information associated with the words such as frequency, and in this case, traditional frequency measures (whole-word and base frequencies) are better predictors than relative frequency.

어휘 재인에서의 음절 빈도 효과: 한국어 단어와 어절에 대한 회귀 분석 연구

김준우,이솔빈,김상엽,남기춘 한국인지및생물심리학회 2023 한국심리학회지 인지 및 생물 Vol.35 No.4

기존 한국어 시각 어휘 재인 연구에서 음절 빈도 효과는 명사 및 동사와 같은 어휘의 품사, 단어 및 어절 등 어휘의 형태, 나아가 타입 및 토큰 빈도와 같은 음절 빈도의 다양한 계산 방식에 따라 혼재된 결과가 보고되었다. 이에 본 연구에서는 철자형 및 음운형 빈도 각각에 대하여, 타입 및 토큰, 길이 제한 및 길이 무관 빈도 등 네 가지 계산 방식에 따라 첫음절 빈도를 계산하여 기존 한국어 어휘 판단 데이터의 23,173개 단어 및 어절 대상 어휘 판단 데이터를 재분석하였다. 총 8가지 독립적 위계적 회귀 모형에서 용언의 원형 및 굴절형 모두에서 첫음절 빈도의 촉진 효과가 나타난 반면, 명사는 어절에서만 촉진 효과가 발견되었다. 이러한 첫음절 빈도의 촉진 효과는 철자 및 음운형 음절 빈도 모두 빈도 계산 방식과 무관하게 일관적으로 나타났다. 음절 빈도의 촉진적 효과는 한국어의 독특한 모아쓰기 체계의 영향으로 음절 간의 경계가 시각적으로 뚜렷해지고, 이로 인하여 시각 어휘 재인 시 직접적 철자 음절 경로를 통한 어휘 접속이 가능하기 때문으로 여겨진다. 또한, 명사 단어에서의 효과 부재는 단어 빈도 효과와 고빈도 이웃 수의 영향으로 유추된다. 따라서, 본 연구 결과는 음절의 음운형 및 철자형 표상 모두가 중요한 역할을 한다는 점을 시사한다. Previous studies on syllable frequency effect on visual recognition of Korean words have reported mixed results depending on their part of speech, morphological form (simple vs. inflected form), and measures of syllable frequency (e.g., type or token frequency). The present study reanalyzed lexical decision data from 23,173 words in three datasets (Yi et al., 2017; Kim et al., 2018; Kim et al., 2020). In eight independent hierarchical regression models, four frequency measures were employed for the orthographic and phonological first syllable, respectively. The first syllable frequency was measured in terms of type or token, or whether the length of syllabic neighbors was considered or not. The results showed robust facilitative effects of first syllable frequency in all forms of verbs, adjectives, and adverbs, while only in inflected form for nouns. The facilitative effects of first syllable frequency were demonstrated regardless of frequency measure, both in orthographic and phonological syllable frequency. This facilitative effect of syllable frequency could be due to the Korean writing system, which provides visually distinct syllabic boundaries, resulting in direct access to the mental lexicon via the orthographic route. The null effect of the syllable frequency in noun words could be attributed to their greater number of higher frequency syllabic neighbors (HFSN), compared to the other three categories. Thus, these results suggest that both the orthographic and phonological representation of a syllable plays an important role in visual word recognition.

주파수 보정 기법에 의한 디지털 주파수 판별기의 성능 개선

최연준,전중남 한국전자파학회 2024 한국전자파학회논문지 Vol.34 No.2

전자전 환경에서 레이다 신호의 주파수를 빠르게 판별하는 것이 매우 중요하다. DFD(digital frequency discriminator)는위상상관기를 거친 레이다 신호에서 IQ 데이터를 수집하고, 이로부터 500 nsec 이내에 주파수를 판별한다. 레이다 신호는 아날로그 및 전처리 과정을 거치면서, IQ 신호에 왜곡이 유입된다. 결과적으로 DFD의 출력주파수가 부정확해진다. 이 논문은 DFD 장치에서 주파수 정확도를 개선하는 기법을 제안한다. 기준 주파수를 인가하고 DFD 장치의 출력주파수를 측정한 다음에, 두 주파수 간의 오차를 선형 보간하여 출력을 보정한다. 이 기법을 적용하였을 때, 주파수 정확도가2.25 MHz에서 0.063 MHz로 향상되었고, 최대 주파수 오차는 6 MHz에서 2 MHz로 개선됨을 확인하였다. In an electronic warfare environment, radar signals must be rapidly discriminated. A digital frequency discriminator (DFD) collects I and Q data from a phase correlator and determines the radar signal frequency in 500 ns. Distortion is introduced into the IQ data as the detected radar signal passes through the analog and preprocessor parts, resulting in an inaccurate DFD output frequency. This paper proposes a technique to improve DFD frequency accuracy. The output is corrected by interpolating the difference between the DFD output frequency and the reference frequency. We confirmed that the proposed technique improves frequency accuracy from 2.25 MHz to 0.063 MHz and reduces the maximum frequency error from 6 MHz to 2 MHz.

첫 음절 토큰 빈도와 타입 빈도가 단어 및 유사 단어 어휘 판단 시간에 미치는 영향

권유안 한국인지및생물심리학회 2012 한국심리학회지 인지 및 생물 Vol.24 No.4

Syllable frequency effect plays an important role in finding evidence supporting the hypothesis, incorporating syllable unit between letter and whole word level. The syllable frequency effect is defined as the inhibitory effect of words starting with high syllable frequency in lexical desition tasks. According to some studies, syllable frequency can be divided into two values. First is the type frequency which indicates the number of syllabic neighbors sharing the first syllable of a target word. Second is the token frequency which means the accumulated word frequency of the syllabic neighbors. Although type and token syllable frequency can make different effect, there is no evidence in Korean. This study conducted the meta-analyses to access whether the inhibitory syllable frequency effect could be expected by syllabic type or token frequency. The results showed that the syllabic token frequency played a critical role in the inhibitory effect in Korean syllable frequency effect and we explained the role of token and type frequency in visual word recognition model. 고빈도 음절로 시작되는 단어가 저빈도 음절로 시작되는 단어에 비해 어휘판단 시간이 느려지는 효과를 음절 빈도 효과라 하고, 이 효과는 시각 단어 재인에서 음절 처리를 포함한 네 수준 모형을 지지해 주는 중요한 효과로 알려져 있다. 그러나 음절 빈도는 계산 방식에 따라 타입 빈도(동일한 첫 음절을 포함한 음절 이웃 단어의 개수)와 토큰 빈도(음절 이웃 단어들의 누적 빈도)로 다르게 구분되고 각각이 단어 재인에서 다른 효과를 보일 수 있다는 주장에도 불구하고 한국어 음절 빈도 효과에서는 그 효과를 구분하지 않았다. 본 연구는 억제적 어휘판단 시간으로 대표되는 음절 빈도 효과가 음절 타입 빈도로 설명되는지 아니면 음절 토큰 빈도로 더 잘 설명되는 지를 기존 연구의 재분석을 통해 검증하였다. 분석 결과 어휘판단 시간의 증가는 음절 토큰 빈도의 증가에 의해 설명되었고, 반대로 유사 단어 자극의 어휘판단 시간의 증가는 음절 타입 빈도 및 표기 이웃 크기의 증가에 의해 설명되는 것으로 검증되었다. 이에 본 연구 결과를 바탕으로 음절 토큰 및 타입 빈도가 시각 단어 재인에서 이론적으로 어떤 의의를 갖는지를 설명하였다.

반복 이산 웨이브릿 변환을 이용한 주파수 추정 기법

박철원(Chul-Won Park) 대한전기학회 2011 전기학회논문지 P Vol.60 No.2

Power system frequency is the main index of power quality indicating an abnormal state and disturbances of systems. The nominal frequency is deviated by sudden change in generation and load or faults. Power system is used as frequency relay to detection for off-nominal frequency operation and connecting a generator to an electrical system, and V/F relay to detection for an over-excitation condition. Under these circumstances, power system should maintain the nominal frequency. And frequency and frequency deviation should accurately measure and quickly estimate by frequency measurement device. The well-known classical method, frequency estimation technique based on the DFT, could be produce the gain error in accuracy. To meet the requirements for high accuracy, recently Wavelet transforms and analysis are receiving new attention. The Wavelet analysis is possible to calculate the time-frequency analysis which is easy to obtain frequency information of signals. However, it is difficult to apply in real-time implementation because of heavy computation burdens. Nowadays, the computational methods using the Wavelet function and transformation techniques have been searched on these fields. In this paper, we apply the Recursive Discrete Wavelet Transform (RDWT) for the frequency estimation. In order to evaluate performance of the proposed technique, the user-defined arbitrary waveforms are used.

고 신뢰성 Audio Target 주파수 발생장치에 대한 연구

박창식,한은택,김익재,신동규 한국인터넷정보학회 2023 인터넷정보학회논문지 Vol.24 No.3

The frequency synthesizer performs a simple function of generating the necessary frequency by manipulating the input frequency signal, but stable and precise frequency generation is essential for reliable operation in mechanical equipment such as communication, control, surveillance, medical and military fields. In various fields, analog and digital methods or hybrid methods have been used for frequency synthesis. Especially in the field of communication, a precise frequency synthesizer required for each frequency band from low frequency AF (Audio Frequency) to high frequency microwave is used. In this paper, a highly reliable frequency synthesizer is designed and implemented using FPGA(Field Programmable Gate Array) without using a microprocessor for application to a railway track circuit system using AF frequency. As a result of the experiment, it showed the performance of generating the target frequency with an accuracy of more than 99.999%. The AF-class frequency synthesizer with such high reliability is expected to play an important role in enhancing convenience, such as reducing the waiting time of citizens who use braking and signaling systems and reducing errors in planned operation when used in transportation such as railways and subways. 주파수 합성기는 입력된 주파수 신호를 조작하여 필요한 주파수를 생성하는 단순한 기능을 수행하지만 통신, 제어, 감시, 의료 및군사 분야 등의 기계장비에서 신뢰성 있는 동작을 위해서는 안정적이고 정밀한 주파수 생성이 필수적이다. 다양한 분야에서 주파수합성은 아날로그 방식과 디지털 방식 또는 이를 혼용한 하이브리드 방식이 사용되어 왔으며, 특히 통신 분야에서는 저주파 AF(Audio Frequency)부터 고주파 마이크로파까지 각 주파수 대역에서 필요한 정밀한 주파수 합성기를 사용하고 있다. 본 논문은 AF 주파수를사용하는 철도 궤도회로 시스템에 적용하기 위한 고도의 신뢰성이 보장된 주파수 합성기를 마이크로프로세서 사용 없이 FPGA(Field Programmable Gate Array)의 로직 회로만으로 설계하고 구현하였다. 실험결과 99.999%이상의 정확도로 Target 주파수를 발생시키는성능을 나타내었다. 이러한 고도의 신뢰성을 갖는 AF급 주파수 합성기는 철도, 지하철 등의 교통운송 수단에 사용될 때 제동 및 신호시스템의 안전한 운영과 계획된 운행의 오차가 줄어들어 이를 이용하는 시민의 대기시간 감소와 편의성을 높이는데 중요한 역할을할 것으로 판단된다.

등척성 운동 시 요추의 각도에 따른 중앙주파수와 토크의 특성

박경희,권오윤,장근,강성재,김영호 한국전문물리치료학회 2001 한국전문물리치료학회지 Vol.8 No.2

Fatigue is the decline in force produced as a result of prolonged muscle activity. Localized muscle fatigue can be identified by a shift toward low in the frequency components of the EMG signal, typically represented by a fall in the median frequency. Previous studies show that a shortened muscle developes a higher fatigue than elongated muscles. The purpose of this study was to investigate the time-related change of median frequency and torque during maximal isometric back extension exercises at different exercise angles (0˚, 12˚, 36˚, 72˚). Twenty healthy subjects (mean age = 24.35 ± 2.70) were evaluated in this study. Median frequency was extracted from the EMG signals by fast Fourier transform (FFT). Initial median frequency and the slope of median frequency change over time were computed from linear regression analysis. Pearson's product moment correlation was used to quantify the relationship between scope of median frequency and torque. The results were as follows: 1) Significant differences in y-intercepts of torque regression equation with respect to exercise angle were shown. However, there were no differences in the slopes of the median frequency and torque, and y intercept of the median frequency among exercise angles. 2) There was no significant correlation between slope of median frequency and torque. 3) But there was moderate correlation between median frequency and torque at each exercise angle. In conclusion, the exercise angle during maximal isometric back extension exercise is not a direct effect on slope of median frequency and torque. But results showed that median frequency and torque shift were highly correlated in all subjects.