A. Sano

Adaptive active noise control schemes in time-domain and transform-domains

A general structure of adaptive active noise control systems is investigated to propose stability-assured adaptive algorithms to update parameters of an adaptive feedforward controller. In order to deal with a general case where all of the sound transmission channel dynamics are uncertainly changeable, which cannot be solved by conventional approaches, a new fully adaptive algorithm is also presented on a basis of robust stabilization by a fixed feedback controller and the strictly positive real property of a newly given error system. The proposed adaptive algorithms can be implemented by not only the time-domain approaches, but also a frequency-domain approach using an adaptive frequency sampling filter and a wavelet transform domain approach using an adaptive wavelet packet filter bank. Comparison of the three proposed approaches is given through numerical simulations and experiments.

Separable estimation of discrete and continuous spectra of signal with mixed spectrum

The authors investigate a method for estimating the mixed spectra of a signal composed of multiple sinusoids and an autoregressive (AR) process. An inverse filtering approach is taken in which they apply the high-resolution Toeplitz approximation method to the filtered signal to estimate the frequencies of the sinusoids. The optimum inverse filter is adjusted in a data-adaptive way, using only received signals, so as to minimize a specified criterion. Thus the coefficients of the inverse filter give the estimate of the AR parameter. Two kinds of criterion are proposed. The first one utilizes the received signal itself. The other one consists of the differences between the autocorrelation function of the filtered signal and that of the reconstructed sinusoids with the estimated frequencies and amplitudes. The number of sinusoids and the order of the AR part are estimated on the basis of the minimum of the criterion. The authors show that even when the additive noise is white, the inverse filtering approach can improve the resolution of the frequency estimate by compensating the errors in the calculated autocorrelation function.<<ETX>>

Directions-of-arrival tracking of coherent cyclostationary signals without eigendecomposition

By exploiting the cyclostationarity exhibited by most communication signals, we propose a computationally efficient method for estimating the directions-of-arrival (DOA) of coherent narrowband signals impinging on a uniform linear array without eigendecomposition and spatial smoothing processes. An online implementation of this method is proposed for tracking the DOA of moving signals in a multipath propagation environment. In the proposed algorithm, a combination of the cyclostationarity and a subarray scheme are used to decorrelate the signal coherency and to suppress the noise and interfering signals, while the noise subspace is estimated from the resulting cyclic correlation matrix through a linear operation. As a result, the proposed algorithm has the advantages of computational simplicity and robust detection capability.

Simultaneous optimization method of regularization and singular value decomposition in least squares parameter identification

In order to attain stabilized convergence, the authors propose a generalized regularization scheme using multiple regularization parameters and an a priori estimate, and they obtain analytically the parameter values that minimize the mean square error (MSE) or the estimated MSE using only accessible data signals. They show that method can give simultaneously the optimal regularization parameters and the optical truncation of smaller eigenvalues in the singular value (or eigenvalue) decomposition (SVD or EVD). The proposed schemes for the optimized regularization and SVD are exemplified in impulse response identification using low-pass input and optimized extrapolation of the bandlimited signal.<<ETX>>

Adaptive recursive scheme for spectral analysis of sinusoids in signals with unknown colored spectrum

A new technique is given to estimate frequencies and amplitudes of sinusoids embedded in background signals with unknown colored spectrum. In such colored noise cases, Pisarenkos harmonic retrival gives only biased estimates. The present approach can eliminate the biases by modeling the colored signals as the AR process, the parameters of which are determined so as to minimize a specified criterion, which also plays an important role of deciding the number of sinusoids. This paper discusses the adaptive implementation of the above algorithm based on a steepest descent method.

Minimum MSE-based detection of cyclostationary signals in array processing

On the motivation that many modulated communication signals exhibit a cyclostationary property, we propose a minimum mean squares error (MSE) based approach for detection of the signal of interest (SOI) impinging on a uniform line array to improve the signal detection capability. Moreover, a data-based iterative regularization algorithm is given, in which the number of the SOI is determined from measured data only and their directions of arrival (DOA) can be decided from the stabilized estimate of the linear prediction coefficients. The effectiveness of the presented approach is demonstrated through numerical examples.

Adaptive and Robust Control Scheme for Active Noise Attenuation

Abstract The proposed adaptive active noise cancelling system is composed of two controllers: an adaptive feedforward controller and a robust feedback controller. By detecting the noise from the primary noise source, the adaptive controller can generate an artificial sound from the secondary source to reduce the noise level to zero at the observation point. New stability-guaranteed adaptive algorithms are given in time-domain and frequency-domain on a basis of the strictly positive real property of the error model in adaptive control theory, and it is implemented to adjust the coefficient parameters in the HR adaptive controller structure. The role of the robust feedback controller is to compensate multi-path effects and disturbances. The robust controller is designed via the H ∞ control theory by taking into account the sensitivity to the disturbances and the model uncertainties in the transfer functions of sound propagation. Numerically and experimentally obtained results exhibited significant improvement on convergency and stability of the adaptive control schemes.

Direct closed-loop identification approach to unstable plant

Deals with the problem of direct closed-loop identification for unstable plant models disturbed by stochastic noise. The unstable plant is stabilized by a digital feedback controller. Then by introducing the output inter-sampling scheme, the plant model is identified from the inter-sampled input-output data of the plant even though the external reference or the test signal does not hold a persistently exciting property. Both time and frequency domain approaches are developed and numerical examples are performed to demonstrate the effectiveness of the proposed approaches.

Array beamforming based on cyclic signal detection

In mobile communication systems, in order to extract the desired signal while suppressing the interference and noise, array beamforming techniques are usually employed. However in most beamforming methods, a priori knowledge of the directions of arrival (DOA) or waveform of the signals of interest (SOI) is often required. In this paper, the cyclostationary property of most communication signals is exploited to develop a new method of beamforming. The DOA is estimated by a signal selective direction finding scheme, and then the optimum weights are determined according to multiple linear constraints. The effectiveness of the proposed approach is demonstrated through numerical examples.

Stabilized Identification Via GSVD Optimized Based on Bayesian Information Theoretic Criterion and Minimum Mean Squares Error

Abstract Identification of an impulse response by using strongly correlated input signals is a typical ill-posed least squares estimation problem. In the present paper, multiple number of regularization parameters are employed to regularize the inversion of a rank-deficient matrix consisting of input signals. We give two approaches for determining these regularization parameters by using only accessible signal data. The first approach is based on minimization of a new Bayesian information theoretic criterion (ABIC) which can be derived on a basis of the Bayesian statistical approach. The second one is based on minimization of an estimated mean squares error (EMSE) of the parameter estimate. An iterative algorithm is given to calculate the optimal regularization parameters and the noise variance estimate simultaneously. By applying the proposed schemes, one can identify oscillatory peaks outside of the frequency range of the input power spectrum, even if the PE condition is not satisfied or even when the number of input-output data is rather less than the length of impulse response to be estimated.