JPH08125589A - Acoustic echo cancelling device - Google Patents

Abstract

PURPOSE: To permit high-speed property to be compatible with high stability and to enable acoustic controlling by dividing a pseudo impulse responce register into N-number of blocks, calculating the representative fluctuation rate of coefficient accumulative addition average power at every block and changing coefficient updating gain by means of the representative fluctuation rate. CONSTITUTION: An accumulative addition average power arithmetic circuit 11 calculates the single accumulative addition average power of respective variable coefficients which are stored in the pseudo impulse responce register 9 divided into N-number blocks. A block average accumulative power arithmetic circuit 12 calculates the individual average addition average power of the respective blocks. A block power fluctuation rate calculating circuit 13 calculates the fluctuation rate of the independent average accumulative addition average power of the respective blocks. A block coefficient updating gain selecting circuit 14 selects coefficient updating gain αwithin the range of zero to one corresponding to the fluctuation rate of the average accumulative addition average power of the respective blocks every time and transmits it to a coefficient correcting quantity arithmetic circuit 7. The circuit 7 calculates correcting quantity by an expression I through the use of the value. In the expression I, hSK is the estimated value of impulse responce (h) at a time kT, ek is residual acoustic echo and Xk is an input call receiving signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used in a communication line, a room sound field control device, and a high-quality voice communication conference device, and an acoustic echo component in which a signal on a receiving path appears in a transmitting path via an acoustic echo path. The present invention relates to an acoustic echo canceller that removes noise.

[0002]

2. Description of the Related Art Generally, an acoustic echo canceller is used in a long-distance telephone line using a communication satellite and a submarine cable.
A correction amount calculation circuit can be roughly divided into a device that removes reflection caused by impedance mismatch of a 2-wire to 4-wire converter and a device that removes reverberation due to acoustic coupling of speaker's voice in a loudspeaker telephone such as a video conference system. , A variable coefficient filter for generating pseudo acoustic echo and a subtraction circuit. The basic operation of the acoustic echo canceller will be described below.

FIG. 5 is a diagram showing the basic structure of an acoustic echo canceller. The reception signal input terminal 1 is connected to the reception signal output terminal 2, and the reception signal of the reception signal input terminal 1 is branched and supplied to the variable coefficient filter 3 to generate pseudo acoustic echo. The transmission signal from the transmission signal input terminal 4 and the pseudo-acoustic echo that is the output of the variable coefficient filter 3 are input to the subtraction circuit 5, the acoustic echo component in the transmission signal is removed, and the output of the subtraction circuit 5 is It is output to the transmission signal output terminal 6. The output of the transmission signal output terminal 6 and the signal of the reception signal input terminal 1 are input to the correction amount calculation circuit 7, and the filter coefficient of the variable coefficient filter 3 is corrected by the output of the coefficient correction amount calculation circuit 7. The reception signal is input to the reception signal input register 8 in the variable coefficient filter 3, and the sum of products of the reception signal of the reception signal input register 8 and the pseudo impulse response of the pseudo impulse response register 9 is obtained by the sum of products circuit 10. Sum of products circuit 10
Is output as a pseudo acoustic echo. The reception signal output terminal 2 and the transmission signal input terminal 4 are connected to a two-wire to four-wire converter in the case of a long-distance telephone line, and to a speaker and a microphone in the case of a public telephone system.

Assuming that the signal propagation characteristic of the acoustic echo path is linear and represented by an FIR type digital filter, if the impulse response h (t) and the input received signal x (t) are used, the sampling time interval is Is T and the acoustic echo y _{k at} time kT is represented by y _k = h ^T x _k (1). However, h = [h ₁ , h ₂ , ..., H _n ] ^T (2) x = [x _k−1 , ..., x _kn ] ^{T T} : transposition of the vector.

On the other hand, the estimated value of h at time kT is h
If s _k , the estimated value ys _k of y _k is given by ys _k = hs _k ^T x _k (3). In the acoustic echo canceller, when there is a voice signal in the reception signal input terminal 1 and there is no voice signal in the transmission signal input terminal 4 and only acoustic echo exists, the echo elimination operation is performed as an adaptive operation state. This adaptive motion algorithm is generally a learning identification method (Atsuhiko Noda, Jinichi Nagumo: “System Learning Identification Method” Measurement and Control, 7, 9, pp.597-605 (1
968)) is adopted. The sequential correction of hs _k by the learning identification method is performed by hs _{k + 1} = hs _k + α (x _k e _k ) / x _k ^T x _k (4). However, e _k = y _k −y s _k , 0 <α ≦ 1 (5), and e _k is called residual acoustic echo. Such a calculation operation is processed in the coefficient correction amount calculation circuit 7. The contents of the pseudo impulse response register 9 include the variable coefficient series h.
s _k is stored. α is a coefficient update gain for determining the sensitivity of estimation, and the closer it is to 1.0, the larger the amount of correction can be given, and high-speed acoustic echo removal can be performed. However, in actual use, near-end noise can be reduced. It is necessary to change the setting depending on the line status. The fact that the coefficient update gain α is currently determined depends on an empirical rule. There are also ones that variably control the coefficient updating gain α depending on the magnitude of the residual acoustic echo and ones that are set according to the indoor characteristics (for example, Shoji Makino and Nobuo Koizumi: “The adaptive characteristics of echo cancellers in the room sound field. Regarding improvement ”, J. A., J71-A, 12, pp.2212-2214 (1988-12).

[0006]

The method of least squares (LMS)
In the parameter estimation by the learning identification method based on, the estimation performance greatly depends on the amount of the coefficient update gain α. According to the equation (5), if the range of α is between 0 and 1, some performance can be obtained, but the convergence speed and the saturation echo removal amount differ due to the difference in the value. An example is shown in FIG. In the figure, a is the acoustic echo removal characteristic when the coefficient update gain α is fixed at 0.5, and b in the figure is the acoustic echo removal characteristic when the coefficient update gain α is set according to the indoor characteristic. It can be seen that a large difference appears in the erasing characteristic depending on the setting method of the coefficient update gain α. When a fixed gain is used, high speed performance can be maintained, but the risk of loss of operational stability increases. Further, if the setting is made according to the indoor characteristics, the operation stability is improved, but the adaptation speed is significantly deteriorated. As described above, the setting of the coefficient update gain α has a problem that conflicting conditions must be satisfied.

The present invention has been made in view of the above points, eliminates the above problems, and is excellent in high speed and operational stability.
An object of the present invention is to provide an acoustic echo canceller having high adaptive performance and performing acoustic control while always maintaining a large amount of acoustic echo cancellation.

[0008]

SUMMARY OF THE INVENTION The present invention is to solve these problems and comprises a reception signal input terminal, a reception signal output terminal, a transmission signal input terminal, and a transmission signal output terminal. A variable coefficient digital filter having a pseudo impulse response register that receives the reception signal of the reception signal input terminal, and the sound of the reception signal input from the reception signal output terminal to the transmission signal input terminal through an acoustic echo path. In an acoustic echo canceling apparatus in which a coefficient sequence is sequentially updated by a coefficient correction amount arithmetic circuit that minimizes a residual signal obtained by subtracting a pseudo acoustic echo generated by the variable coefficient digital filter from an echo component, The pseudo impulse response register is divided into N blocks, the representative variation rate of the coefficient cumulative addition average power for each block is calculated, and the representative variation rate is calculated. To provide an acoustic echo removal device coefficient update gain for each block is changed.

[0009]

According to the present invention, since the high speed and high stability of the estimation operation are ensured by the above means, the reverberation component is hardly mixed into the communication line, the deterioration of the communication voice sound quality is prevented, and the call itself. It is possible to reduce the risk of howling that would otherwise be impossible, and high-quality acoustic control becomes possible.

[0010]

Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of the acoustic echo canceller of the present invention. As shown in FIG. 1, according to the present invention, the conventional reception signal input terminals 1, 3 reception signal output terminals 2, variable coefficient digital filter 3, transmission signal input terminal 4, subtraction circuit 5, transmission signal output terminal 6, Coefficient correction amount calculation circuit 7,
A cumulative addition / average power calculation circuit 11 is provided in a device having the same configuration as the acoustic echo canceling device that employs a learning identification method as an adaptive algorithm, which is composed of a reception signal input register 8, a pseudo impulse response register 9, and a product-sum calculation circuit 10. ,
This is a configuration in which a block average cumulative power calculation circuit 12, a block power fluctuation rate calculation circuit 13, and a block coefficient update gain selection circuit 14 are added. The reception signal input terminal 1
, The reception signal output terminal 2 and the transmission signal input terminal 4
And the transmission signal output terminal 6 and the pseudo impulse response register 9 that receives the reception signal of the reception signal input terminal 1
And the variable coefficient digital filter 3 generated from the acoustic echo component of the received signal input to the transmitted signal input terminal 4 from the received signal output terminal 2 via the acoustic echo path. In the acoustic echo canceller in which the coefficient sequence is sequentially updated by the coefficient correction amount calculation circuit 7 so as to minimize the residual signal obtained by subtracting the pseudo acoustic echo, the pseudo impulse response register 9
The cumulative average power calculation circuit 11 for calculating the cumulative average power of each variable coefficient alone stored in, and the block average cumulative power calculation circuit 12 for calculating the average cumulative average power of each block, It is stored in the block power fluctuation rate calculation circuit 13 that calculates the fluctuation rate of the block-independent average cumulative addition average power, and in the block coefficient update gain selection circuit 14 that corresponds to the fluctuation rate of the average cumulative addition average power of each block. The coefficient update gain α corresponding to each block in the range of 0.0 to 1.0 is selected each time, the selected value is sent to the coefficient correction amount calculation circuit 7, and this value is used to correct by the equation (4). An acoustic echo canceller characterized by calculating a quantity. In the block average cumulative power calculation circuit 12, the average power P _BN of each block of the cumulative addition average power of the variable coefficient stored in the pseudo impulse response register is calculated by the equation (6). The size of each block may be equally spaced or may be unequally spaced.

P _B1 = (hs ₀ ² + ... + hs _m1-1 ² ) / m1 P _B2 = (hs _m1 ² + ... + hs _m2-1 ² ) / (m2-m1) (6) | P _BN = total ^{_{(hs mN 2 + ··· + hs}} L-1 2) / (L-mN) L is variable coefficient, m1, m @ 2, ..., mN coefficient number divides Then, the total number of N is divided block Is. Here, the average for each block is used as the evaluation value, but if it is a value that can represent that block, it may be the maximum value or the minimum value of the coefficient cumulative addition average power instead of this value. The configuration of the block average cumulative power calculation circuit 12 is shown in FIG. The block power fluctuation rate calculation circuit 13 uses each average power P _BN to perform the calculation using equation (7).

DP _BN (k + 1) = | P _BN (k + 1) −P _BN (K) | / P _BN (K) (7) The fluctuation rate shown in the equation (7) is an example, and is determined. It is not a calculation method. For example, the numerator may be the present value instead of the past value. The present invention also works effectively by using the difference value between dP _BN (k + 1) and dP _BN (K). In the block coefficient update gain selection circuit 14, the representative variation rate d of each block
Compare the value of P _BN with several thresholds interpolated in the circuit,
An operation of selecting a coefficient update gain corresponding to a suitable threshold range is performed. FIG. 3 shows this conceptual diagram. This time,
When it is detected that the variation rate of the block representative evaluation value becomes equal to or less than a certain value, the coefficient correction amount calculation circuit 7 of the block is detected.
By setting the correction amount calculation and the update calculation not to be performed, the calculation amount can be reduced, the load on the hardware can be reduced, and the risk of malfunction due to a calculation error can be reduced. FIG. 4 shows a result a of the adaptive processing operation according to the present invention when white noise is used as a reference input. As a comparison target, the result b of the model in which the coefficient update gain is fixed to 0.5 and the result c of the model in consideration of the indoor characteristics are shown in the same figure. The vertical axis represents the amount of acoustic echo cancellation, and the horizontal axis represents time. The maximum setting value of the coefficient update gain of the model according to the present invention is 1.0 and the minimum setting value is 0.05. The response speed in the initial cancellation transient region is 30 [dB] for the acoustic echo cancellation.
It can be seen that the model according to the present invention is superior to the model according to the present invention by about 2 times. Since the coefficient update gain is gradually set to a small value, it is not affected by disturbance.

[0013]

As described above, according to the present invention, the following excellent effects are expected. (1) By using the present invention, high speed and high stability can be realized at the same time, so that high quality voice communication can be maintained and the risk of howling can be suppressed to a low level. (2) In the adaptive operation process, a variable coefficient that does not need to be updated is generated. This reduces the amount of calculation,
The burden on the hardware can be reduced.

(3) Since the variable coefficient is not updated more than necessary, an arithmetic error or malfunction is less likely to occur when the digital signal processor or the like is used.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example of an acoustic echo canceller according to the present invention.

FIG. 2 is a block diagram showing a configuration of a block average cumulative power arithmetic circuit used for explaining the present invention.

FIG. 3 is a block diagram showing a concept of a state determination control unit.

FIG. 4 is a diagram showing an example of acoustic echo cancellation characteristics when white noise according to the present invention is used as a reference input.

FIG. 5 is a diagram showing an example of acoustic echo canceling characteristics by a conventional method when the white noise according to the present invention used in the present description is used as a reference input.

FIG. 6 is a block diagram showing an example of a basic configuration of an acoustic echo canceller using a conventional general learning identification method.

[Explanation of symbols]

 1 reception signal input terminal 2 reception signal output terminal 3 variable coefficient filter 4 transmission signal input terminal 5 subtraction circuit 6 transmission signal output terminal 7 correction amount calculation circuit 8 reception signal input register 9 pseudo impulse response register 10 sum of products calculation circuit 11 Cumulative average power calculation circuit 12 block average cumulative power calculation circuit 13 block power fluctuation rate calculation circuit 14 block coefficient update gain selection circuit

Claims (1)

[Claims] 1. A variable having a reception signal input terminal, a reception signal output terminal, a transmission signal input terminal, a transmission signal output terminal, and a pseudo impulse response register to which the reception signal of the reception signal input terminal is input. A coefficient digital filter and a pseudo acoustic echo generated by the variable coefficient digital filter are subtracted from the acoustic echo component of the reception signal input to the transmission signal input terminal from the reception signal output terminal through the acoustic echo path. In an acoustic echo canceller in which a coefficient series is sequentially updated by a coefficient correction amount calculation circuit that minimizes a residual signal to be obtained, the pseudo impulse response register is divided into N blocks, and coefficient accumulation for each block is performed. An acoustic echo characterized by calculating the representative fluctuation rate of the average power and changing the coefficient update gain for each block by the calculated representative fluctuation rate. Removal device.