Telephonic communications – Echo cancellation or suppression – Residual echo cancellation
Reexamination Certificate
1999-04-07
2003-05-13
Isen, Forester W. (Department: 2644)
Telephonic communications
Echo cancellation or suppression
Residual echo cancellation
C379S388010, C379S406080, C379S406130, C370S289000, C455S570000
Reexamination Certificate
active
06563925
ABSTRACT:
BACKGROUND
This invention relates generally to a method and apparatus for echo cancellation. More particularly, this invention relates to a method and apparatus for space-time echo cancellation in a communication system.
Echo related problems are very common in communication systems such as cellular telephone systems. In such systems, speech originating from a far-end loud speaker echoes back to a microphone with a time delay causing perception problems. Perception is further impaired when the speaker is situated in a noisy environment, as in the case of a car telephone operating in a hands-free mode.
Echo cancellation has been used to decrease the echo from the far-end speaker during hands-free communication, and a considerable amount of effort has been spent in this field. One approach to cancelling echo is the single-microphone echo suppression technique which utilizes differences in the spectral characteristics of speech and noise. This type of method is disclosed, for example, in S. F. Boll, “Suppression of acoustic noise in speech using spectral subtraction”, IEEE Trans. on Acoustics, Speech, and Signal Processing, ASSP-27(2):113-120, April 1979; R. J. McAulay and M. L. Malpass, “Speech enhancement using a soft-decision noise suppression filter”, IEEE Trans. on Acoustics, Speech, and Signal Processing, ASSP-28:137-145, 1980; M. Sambur, “Adaptive noise canceling for speech signals, IEEE Trans. on Acoustics, Speech, and Signal Processing, ASSP-26:419-423, 1978; Y. Ephraim and D. Malah, “Speech enhancement using a minimum mean-square error short-time spectral amplitude estimator”, IEEE Trans. on Acoustics, Speech, and Signal Processing, ASSP-32:1109-1121, 1984; P. Sorqvist, P. Handel, and B. Ottersten, “Kalman filtering for low distortion speech enhancement in mobile communication”, Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, volume 2, 1219-1222, Munich, Germany, April 1997; S. H. Jensen, P. C. Hansen, and J. A. Sorensen, “A signal subspace approach for noise reduction of speech signals”, Proc. EUSIPCO European Signal Processing Conference, volume 2, 1174-1177, Lausanne, Switzerland, 1994, EURASIP; and Jesper Jansson and Peter Kaarle, “Noise cancelling by spectral magnitude subtraction”, Master Thesis, January 1992, Ericsson Mobile Communication A B, Department of Tele Transmissions Theory, Lund Institute of Technology, Sweden.
A conventional system for performing single microphone echo cancellation is illustrated in FIG.
1
. The system includes a filter
120
into which a signal directed to a loudspeaker
140
is fed. The output of the filter
120
represents the echo from the loudspeaker
140
. The output of the filter
120
is subtracted from a signal received from a microphone
110
in a subtractor
130
, resulting in a signal in which the echo from the loudspeaker is canceled.
A processing algorithm for single microphone echo cancellation can be described by the following equation:
U
out
(&ohgr;)=U(&ohgr;,r)−
K
(&ohgr;) U(&ohgr;,r
e
) (1)
where r
e
represents the spatial coordinates of the loudspeaker, r represents the spatial coordinates of the single microphone, U
out
(&ohgr;) is the Fourier transform of the resulting signal in which the echo is canceled, U(&ohgr;,r) is the Fourier transform of the signal u(t,r) observed at the output of the single microphone
110
, U(&ohgr;,r
e
) is the Fourier transform of a signal u(t,r
e
), e.g., a voltage, directed to the loudspeaker
140
, and K(&ohgr;) is the frequency response of the filter
120
, which may be calculated according to the equation:
K
(&ohgr;)=G(&ohgr;,r
e
,r) (2)
where G(&ohgr;,r
e
,r) is the Green function which describes signal propagation from the loudspeaker
140
to the single microphone
110
. The filter
120
simulates the frequency response of the noise generated by the loudspeaker
140
. This noise is subtracted from the signal received by the microphone
110
, resulting in a signal in which the echo is canceled.
In many situations, speech and noise tend to have similar spectral distributions. In such situations, the single-microphone echo suppression technique does not yield substantial improvement in speech intelligibility. On the other hand, the signal and the echo in a car environment are acoustical fields which have different spatial characteristics. Taking this into account, the spatial separation of the speech and the echo can be exploited to reduce the noise level without any bad effects on the speech.
It is known that spatial signal processing requires arrays that combine the outputs of several microphones. Techniques which utilize arrays in conjunction with signal processing have been developed and applied in other fields such as sonar and seismic focus searching. This type of technique, called “matched-field processing”, can achieve effective rejection of underwater noise (ambient noise and ocean reverberation), as described in L. G. Krasny, “Spatial processing of acoustic signals in a plane-parallel waveguide”, Sov. Phys. Acoust., 30, 4, 495-501, 1984 and A. B. Baggeroer, W. A.Kuperman, and H. Shmidt, “Matched-field processing: source localization in correlated noise as an optimum parameter estimation problem”, J. Acoust. Soc. Am. 83, 571-587, 1988.
FIG. 2
illustrates a conventional system for performing matched field processing. The system includes N filters
150
which filter N signals received from microphones
110
to remove noise from the signals, where N=1, 2, 3, . . . The filters
150
simulate the frequency response of the noise, including the echo from the loudspeaker
140
. The filtered results are summed in a summer
160
, and the resulting sum is a signal in which the echo is canceled.
A matched-field processing algorithm can be described by the following equation:
U
out
(
ω
)
=
∑
i
=
1
N
U
(
ω
,
r
i
)
H
*
(
ω
;
r
i
)
(
3
)
where r
i
represents the spatial coordinates of the ith microphone, U
out
(&ohgr;) is the Fourier transform of the signal output from the summer
160
, U(&ohgr;,r
i
) is the Fourier transform of the field u(t,r
i
) observed at the output of the i-th microphone
110
, and H(&ohgr;;r
i
) is the amplitude-phase distribution at the aperture of the array which satisfies the system of equations:
∑
k
=
1
N
g
N
(
ω
;
r
i
,
r
k
)
H
(
ω
;
r
k
)
=
exp
(
ⅈ
ω
&LeftBracketingBar;
r
i
-
r
o
&RightBracketingBar;
/
c
)
,
(
4
)
where g
N
(&ohgr;;r
i
,r
k
) is the spatial correlation function of the background noise, r
o
represents the spatial coordinates of the talker, and c represents the speed of sound.
There are some difficulties which become apparent when the matched-field processing technique is applied for echo cancellation in a car telephone environment. First, matched-field processing is based on an assumption that the microphone array is located in a free-field propagation channel. However, the free-field propagation model does not take into account the effects of the waveguide sound propagation in a car cabin and is thus unrealistic for a car environment. Secondly, matched-field processing does not take into account a priori information about the spatial structure of the echo field. Since it is known that the echo field is the spatial-coherent acoustic field, it would be worthwhile to explore the possibility of suppressing spatial-coherent acoustic noise by means of the algorithms, in which technical feasibility would be combined with a reasonably high performance. Third, matched-field processing does not account for the speech signal output by a loudspeaker. Including this signal would substantially improve performance of the echo cancellation.
There is thus a need for a method and apparatus for echo cancellation which avoids the problems of the prior art.
SUMMARY
It is therefore an object of the present invention to provide a method and apparatus for echo cancellation that takes into account a signal output from a loudspe
Coats & Bennett P.L.L.C.
Ericsson Inc.
Isen Forester W.
Swerdlow Daniel
LandOfFree
Method and apparatus for space-time echo cancellation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for space-time echo cancellation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for space-time echo cancellation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3079034