Telephonic communications – Echo cancellation or suppression
Reexamination Certificate
2001-08-13
2004-11-09
Kuntz, Curtis (Department: 2643)
Telephonic communications
Echo cancellation or suppression
C379S347000, C379S406070, C379S406160
Reexamination Certificate
active
06816591
ABSTRACT:
BACKGROUND OF THE INVENTION
(i) Field of the Invention
The present invention relates to a voice switching system for and a voice switching method of removing sound echoes and reducing occurrence of howling in sound reinforced communication systems such as hands-free telephones and teleconference systems which are on their ways to becoming popular in recent years.
(ii) Description of the Related Art
There have so far been developed a wide variety of sound reinforced communication systems each of which makes it possible for many people to talk with one another each with one set of hands-free speaker and microphone.
The sound reinforced communication system of this type is in general equipped with at least one pair of hands-free telephone units, i.e., first and second hands-free telephone units positioned in respective closed near-end and far-end rooms distant far away from each other and each having one set of hand-free speaker and microphone, thereby making it possible for two operators to communicate with each other by the speakers and microphones of the first and second hands-free telephone units through a special transmission line having the first and second hands-free telephone units electrically connected with each other. The speaker and microphone forming parts of the hands-free telephone unit begins to be operated when the operator inputs his or her voice to the microphone, so that the reinforced communication unit has another name called “a voice switching system”.
One of typical examples among those conventional voice switching systems is disclosed, for example, in Japanese Patent Application Laid-Open No. 62151/1990.
The conventional voice switching system disclosed in this Japanese publication is shown in
FIG. 7
as comprising a reception voice attenuator
702
designed to attenuate the reception signal inputted from a signal input terminal
701
to output the attenuated signal to a speaker
703
, a transmission voice attenuator
705
adapted to attenuate the transmission signal inputted from a microphone
704
to output the attenuated signal to an output terminal
706
, a receiving-side voice detection processing section
707
operative to carry out voice detection processing on the inputted reception signal, a transmitting-side voice detection processing section
713
functioning to carry out voice detection processing on the inputted transmission signal, and a loss controlling section
719
serving to control the losses of the reception voice attenuator
702
and the transmission voice attenuator
705
.
The receiving-side voice detection processing section
707
thus constructed comprises a signal level computing section
708
which computes the amplitude level of the reception signal inputted from the signal input terminal
701
, a time constant selecting section
709
which selects a time constant used when a minimum reception signal level is computed, a minimum signal level computing section
710
which computes the minimum reception signal level, a threshold computing section
711
which computes a threshold from the minimum reception signal level computed in the minimum signal level computing section
710
, and a voice detecting section
712
which carries out voice detection according to the reception signal level and the threshold. The amplitude level computed in the signal level computing section
708
is a reception signal level obtained by rectifying and smoothing the reception signal.
The transmitting-side voice detection processing section
714
further comprises a signal level computing section
714
which computes the amplitude level of the transmission signal inputted from the microphone
704
, a time constant selecting section
715
which selects a time constant used when a minimum transmission signal level is computed, a minimum signal level computing section
716
which computes the minimum transmission signal level, a threshold computing section
717
which computes a threshold from the minimum transmission signal level computed in the minimum signal level computing section
716
, and a voice detecting section
718
which carries out voice detection according to the transmission signal level and the threshold. The amplitude level computed in the signal level computing section
714
is a transmission signal level obtained by rectifying and smoothing the transmission signal.
The following description will be directed to the operation of the above conventional voice switching system with reference to FIG.
7
.
When a reception signal, i.e., the voice signal of a far-end speaker is inputted to the signal input terminal
701
, the reception signal is outputted to a near-end speaker from the speaker
703
by way of the reception voice attenuator
702
. The microphone
704
is then operated to collect the voice outputted from the speaker
703
and the voice of the near-end speaker speaking to the microphone
704
to output a transmission signal. This transmission signal becomes a transmission output signal via the transmission voice attenuator
705
, and the transmission output signal is outputted to the far-end speaker from the signal output terminal
706
.
Description will then be given to the receiving-side voice detection processing section
707
and transmitting-side voice detection processing section
713
required to compute the losses to be inserted into the reception voice attenuator
702
and the transmission voice attenuator
705
by the loss controlling section
719
. Only the receiving-side voice detection processing section
707
will appear because the receiving-side voice detection processing section
707
and the transmitting-side voice detection processing section
713
are operated in the same manner.
In the signal level computing section
708
, the amplitude level of a reception signal in each sample or frame (multiple samples) is computed to obtain a signal level Lri(k) in which the legend “k” represents a sample number or a frame number. In the time constant selecting section
709
, a time constant “Tr” is determined according to the amplitude level of the reception signal. In the minimum signal level computing section
710
, a minimum reception signal level Nr(k) is computed by the smoothing processing of the following equation 1 using this time constant.
Nr
(
k
)=
Nr
(
k
−1)+
Tr
(
Lr
(
k
)−
Nr
(
k
−1)) (equation 1)
In the threshold computing section
711
, a threshold “Thr” for voice detection is computed by the following equation 2 based on the minimum reception signal level Nr(k),
Thr=&agr;·Nr
(
k
) (equation 2)
wherein the legend “&agr;” is indicative of a coefficient for computing the threshold.
In the voice detecting section
712
, the reception signal level Lri(k) is compared with the threshold “Thr”, and when the reception signal level is higher than the threshold, it is determined that a voice is present, while when the reception signal level is lower than the threshold, it is determined that no voice is present.
The methods of computing and controlling the losses in the loss controlling section
719
will then be described hereinafter.
The loss controlling section
719
is firstly operated to have a transmission signal level Lsi(k) compared with a reception output signal level Lro(k) obtained by multiplying the reception signal level Lri(k) by a receiving-side loss “Gr”, and a sound echo path gain “&agr;h” is computed by the following equation.
&agr;h=Lsi
(
k
)/
Lro
(
k
) (equation 3)
The loss controlling section
719
is similarly operated to have a reception signal level Lri(k) compared with a transmission output signal level Lso(k) obtained by multiplying the transmission signal level Lsi(k) by a transmitting-side loss Gs, and a circuit echo path gain “&bgr;h” is computed by the following equation.
&bgr;h=Lri
(
k
)/
Lso
(
k
) (equation 4)
Computed by the following equation with the sound echo path gain “&bgr;h” and the circuit echo path gain “&bgr;h” is an insertion loss “G”,
G=Hm
/(
Mc·&agr;h·&bgr;h
)
Terada Yasuhiro
Ura Takefumi
Kuntz Curtis
Matsushita Electric - Industrial Co., Ltd.
Pearne & Gordon LLP
Pham Tuan
LandOfFree
Voice switching system and voice switching method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Voice switching system and voice switching method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Voice switching system and voice switching method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3347400