• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.2 s.308
• /
• pp.87-92
• /
• 2006

In array signal processing, high order statistics can be used to estimate parameters from signal of sums of complex exponential. This paper presents a high order Matrix Pencil method(MPM) using the fourth order cumulant. Since the fourth order cumulant can suppress the Gaussian noise, the response of MPM has better noise immunity than the conventional approaches. We successfully formulate the high order MPM with all the benefits of MPM along with higher accuracy. In the numerical simulations we demonstrated that the proposed method with forth order cumulant has better resolution to find degree of arrival(DOA) in the presence of the Gaussian noise.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.2 s.344
• /
• pp.39-43
• /
• 2006

This paper deals with the AR Identification of unknown system using cumulant, which is the 3rd order statistics of output signal in the presence of the noise signal. The algorithms for identification of unknown system we applies to the AR identification method using the cumulant which is possible to the guarantees of global convergence and the representation of amplitude and phase information of system among with the method of parametric modeling. In the process of identification, we considered unknown system to the one of AR system. After the generation of input signal, it was being passed through the system then We use the its output signal that the noise is added. As a result of identification of AR system by changing the signal to noise ratio, we get the fairly good results compared to original system output values and confirmed that the pole was located in the unit circle of z transform.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.6C
• /
• pp.629-636
• /
• 2006

In array signal processing, high order statistics can be used to estimate parameters from signal of sums of complex exponential. In this paper, we derive two types of direction finding algorithms which use the fourth-order cumulant and moment of the received array data. Since the fourth order cumulant can suppress the Gaussian noise, the response of MPM has better noise immunity than the conventional approaches. The performance of each method in regard to the probability of resolution and SNR in the presence of the Gaussian noise is investigated. As a result, the proposed method applied to the fourth-order statistic can find DOA more correctly in the presence of the Gaussian noise.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.2
• /
• pp.50-59
• /
• 1994

This paper presents a speech recognition method utilizing third-order cumulants as a feature vector and a neural network for recognition. The use of higher-order cumulants provides desirable uncoupling between the gaussian noise and speech, which enables us to estimate the coefficients of AR model without bias. Unlike the conventional method using second-order statistics, the proposed one exhibits low bias even in SNR as low as 0 dB at the expense of higher variance. It is confirmed through computer simulation that recognition rate of korean single-vowels with the cumulant-based method is much higher than the results with the conventional method even in low SNR.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.3 no.2
• /
• pp.157-160
• /
• 1999

We show the cumulants of a minimal sufficient statistics in k parameter natural exponential family by parameter function and partial parameter function. We nd the cumulants have some merits of central moments and general cumulants both. The first three cumulants are the central moments themselves and the fourth cumulant has the form related with kurtosis.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1995.06a
• /
• pp.55-60
• /
• 1995

고차 통계 방법을 이용하여 잡음이 섞인 음성 신호에서 피치를 구하는 방법과 이를 이용하여 유성음 및 무성음 구간을 구분하는 방법을 구현하고 그 결과를 기술하였다. 고차 통계의 일종인 3차 cumulant 함수의 경우 Gaussian 또는 대칭적인 분포를 갖는 잡음 신호를 음성신호로부터 효과적으로 분리하여 제거시키는 특징을 가지고 있으며, 이러한 특징을 이용하면 잡음 환경에서 여러 가지 음성 특징 파라메터들을 보다 신뢰도 높게 추정할 수 있다. 본 논문에서는 dam성 신호의 3차 cumulant 함수의 자기상관함수로부터 음성의 피치 주기를 추정하였으며, 피치 위치에서의 normalized peak 크기에 의해 유성음과 무성음을 구분하였다. 또한 성능 비교를 위해 음성 신호 자체의 자기 상관 함수로부터 역시 피치 주기 및 유성음/무성음 구분을 수행하였다. 백색 및 유색 Gaussian 잡음 환경에서의 음성의 피치 주기 추정 실험 결과 SNR가 낮은 경우에 3차 cumulant를 이용한 방법이 2차 통계에 비해 우수한 성능을 나타내었다. 또한 동일한 잡음 환경에서의 유성음/무성음 판별 시험에서도 3차 cumulant를 이용한 방법이 기존의 2차 통계를 이용한 방법에 비해 성능이 크게 향상된 결과를 얻었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.9
• /
• pp.2009-2019
• /
• 2013

In this paper, we propose an automatic modulation recognition algorithm based on cyclic moment and new modified cumulant for analog and digital modulation signals. It is noteworthy that each modulated signal has different cycle frequency characteristics according to its order of cyclic moment. By means of this characteristics as classification features, various modulated signals can be efficiently classified. Also, to identify modulated signals having the same cycle frequency characteristics, we take advantage of the additional classification factors such as variations of envelope and phase as well as modified cumulant. The proposed algorithm was evaluated by considering the number of symbols, SNR, and frequency offset. In the simulation condition where the number of gathered symbols was about 819, and SNR and frequency offset were above 10dB and below 25%, respectively, the average accuracy of the proposed algorithm was more than 95%.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.3
• /
• pp.705-712
• /
• 2005

Entropy is the basic concept of information theory. It is well defined for random varibles with known probability density function(pdf). For given data with unknown pdf, entropy should be estimated. Usually, estimation of entropy is based on the approximations. In this paper, we consider a kernel based approximation and compare it to the cumulant approximation method for several distributions. Monte carlo simulation for various sample size is conducted.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2000.11a
• /
• pp.225-226
• /
• 2000

• Journal of the Korean Chemical Society
• /
• v.42 no.4
• /
• pp.398-410
• /
• 1998

For non-linear solution of the Boltzmann equation, the cumulant moment method has been studied. To apply the method to the normal shock wave problem, we restricted ourselves to the monatomic Maxwell molecular gases. The method is based on the iterative approach developed by Maxwell-Ikenberry-Truesdell (MIT). The original MIT approach employs the equilibrium distribution function for the initial values in beginning the iteration. In the present work, we use the Mott-Smith bimodal distribution function to calculate the initial values and follow the MIT iteration procedure. Calculations have been carried out up to the second iteration for the profiles of density, temperature, stress, heat flux, and shock thickness of strong shocks, including the weak shock thickness of Mach range less than 1.4. The first iteration gives a simple analytic expression for the shock profile, and the weak shock thickness limiting law which is in exact accord with the Navier-Stokes theory. The second iteration shows that the calculated strong shock profiles are consistent with the Monte Carlo values quantitatively.