Blind single channel deconvolution using nonstationary signal processing

Blind deconvolution is fundamental in signal processing applications and, in particular, the single channel case remains a challenging and formidable problem. This paper considers single channel blind deconvolution in the case where the degraded observed signal may be modeled as the convolution of a nonstationary source signal with a stationary distortion operator. The important feature that the source is nonstationary while the channel is stationary facilitates the unambiguous identification of either the source or channel, and deconvolution is possible, whereas if the source and channel are both stationary, identification is ambiguous. The parameters for the channel are estimated by modeling the source as a time-varyng AR process and the distortion by an all-pole filter, and using the Bayesian framework for parameter estimation. This estimate can then be used to deconvolve the observed signal. In contrast to the classical histogram approach for estimating the channel poles, where the technique merely relies on the fact that the channel is actually stationary rather than modeling it as so, the proposed Bayesian method does take account for the channel's stationarity in the model and, consequently, is more robust. The properties of this model are investigated, and the advantage of utilizing the nonstationarity of a system rather than considering it as a curse is discussed.

[1]  T. Rao.  The Fitting of Non-stationary Time-series Models with Time-dependent Parameters , 1970 .

[2]  Yves Grenier,et al.  Time-dependent ARMA modeling of nonstationary signals , 1983 .

[3]  Shoji Makino,et al.  Common acoustical pole and zero modeling of room transfer functions , 1994, IEEE Trans. Speech Audio Process..

[4]  Heinrich Kuttruff,et al.  Room acoustics , 1973 .

[5]  A. Willsky,et al.  Time-varying parametric modeling of speech☆ , 1983 .

[6]  L. Zadeh,et al.  Time-Varying Networks, I , 1961, Proceedings of the IRE.

[7]  Alan Oppenheim,et al.  Time-varying parametric modeling of speech , 1977, 1977 IEEE Conference on Decision and Control including the 16th Symposium on Adaptive Processes and A Special Symposium on Fuzzy Set Theory and Applications.

[8]  Gavin C. Cawley,et al.  Fast Method of Channel Equalisation for Speech Signals and its Implementation on a DSP , 1999 .

[9]  James R. Hopgood.  Nonstationary signal processing with application to reverberation cancellation in acoustic environments , 2000 .

[10]  Charles W. Therrien,et al.  Discrete Random Signals and Statistical Signal Processing , 1992 .

[11]  Peter J. W. Rayner,et al.  Digital Audio Restoration: A Statistical Model Based Approach , 1998 .

[12]  Peter J. W. Rayner,et al.  Generalized feature extraction for time-varying autoregressive models , 1996, IEEE Trans. Signal Process..

[13]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[14]  O Punska,et al.  Bayesian segmentation of piecewise constant autoregressive processes using MCMC methods , 1999 .

[15]  John Mourjopoulos,et al.  Pole and zero modeling of room transfer functions , 1991 .

[16]  Ehud Weinstein,et al.  System identification using nonstationary signals , 1996, IEEE Trans. Signal Process..

[17]  S. Gudvangen,et al.  Comparison of pole-zero and all-zero modelling of acoustic transfer functions (echo cancellation) , 1992 .

[18]  J. J. Rajan,et al.  Bayesian approach to parameter estimation and interpolation of time-varying autoregressive processes using the Gibbs sampler , 1997 .

[19]  J. Makhoul,et al.  Linear prediction: A tutorial review , 1975, Proceedings of the IEEE.

[20]  M. Barlaud,et al.  Results on AR-modelling of nonstationary signals , 1987 .

[21]  L. A. Liporace.  Linear estimation of nonstationary signals. , 1975, The Journal of the Acoustical Society of America.

[22]  Eric Moulines,et al.  Simulation-based methods for blind maximum-likelihood filter identification , 1999, Signal Process..

[23]  Ta-Hsin Li,et al.  A blind equalizer for nonstationary discrete-valued signals , 1997, IEEE Trans. Signal Process..

[24]  S. J. Flockton,et al.  Modelling of acoustic transfer functions for echo cancellers , 1995 .

[25]  James R. Hopgood,et al.  Bayesian single channel blind deconvolution using parametric signal and channel models , 1999, Proceedings of the 1999 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics. WASPAA'99 (Cat. No.99TH8452).

[26]  Rong Chen,et al.  Blind restoration of linearly degraded discrete signals by Gibbs sampling , 1995, IEEE Trans. Signal Process..

[27]  Pjw Rayner,et al.  The effects of non-stationary signal characteristics on the performance of adaptive audio restoration systems , 1989, International Conference on Acoustics, Speech, and Signal Processing,.

[28]  Ta‐Hsin Li.  ESTIMATION AND BLIND DECONVOLUTION OF AUTOREGRESSIVE SYSTEMS WITH NONSTATIONARY BINARY INPUTS , 1993 .

[29]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1972 .

[30]  K OrJ.  Numerical Bayesian methods applied to signal processing , 1996 .

[31]  Peter J. W. Rayner,et al.  Separation of stationary and time-varying systems and its application to the restoration of gramophone recordings , 1989, IEEE International Symposium on Circuits and Systems,.

[32]  Simon J. Godsill,et al.  Considering non-stationarity for blind signal separation , 1999, Proceedings of the 1999 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics. WASPAA'99 (Cat. No.99TH8452).