An echo canceller for decreasing a deterioration factor of speech quality

(57) Abrégé/Abstract:

The echo canceller of the present invention comprises a means for estimating an echo gain from a reception signal and a transmission signal; first level calculating means for calculating the power of the reception signal; second level calculating means for calculating the power of the transmission signal; echo gain estimating means for estimating an echo gain based on the power of the reception signal and the power of the transmission signal; and echo level estimating means for adding the estimated echo gain to the power of the reception signal to obtain an estimated echo level, third level calculating means for calculating the power of a residual signal; and comparator means for comparing the estimated echo level with the power of the residual signal after the echo cancellation; wherein the echo of the reception signal contained in the transmission signal is cancelled when said estimated echo level exceeds the value obtained by adding a margin to the power of the residual signal. The various powers may be logarithmic values.

Canada http://opic.gc.ca · Ottawa-Hull K1A 0C9 · http://cipo.gc.ca OPIC · CIPO 191

O PI c

<img file="CA2199566C_D0001.tif" />

CIPO

CA 02199566 2004-08-24

ABSTRACT OF THE DISCLOSURE

The echo canceller of the present invention comprises a means for estimating an echo gain from a reception signal and a transmission signal; first level calculating means for 5 calculating the power of the reception signal; second level calculating means for calculating the power of the transmission signal; echo gain estimating means for estimating an echo gain based on the power of the reception signal and the power of the transmission signal; and echo level estimating means for adding the estimated echo gain to the power of the reception signal to obtain an estimated echo level, third level Calculating means for calculating the power of 10 a residual signal; and comparator means for comparing the estimated echo level with the power of the residual signal after the echo cancellation; wherein the echo of the reception signal contained in the transmission signal is cancelled when said estimated echo level exceeds the value obtained by adding a margin to the power of the residual signal. The various powers may be logarithmic values.

CA 02199566 2004-08-24

AN ECHO CANCELLER FOR DECREASING A DETERIORATION

FACTOR OF SPEECH QUALITY

BACKGROUND OF THE INVENTION

FIELD OF THE INVENTION

The invention relates to an echo canceller for cancelling an echo signal which is a factor contributing to deterioration of a speech quality in satellite communications and telephones with loudspeakers.

DESCRIPTION OF THE PRIOR ART

An example of conventional echo cancellers is disclosed in (1) Echo canceller technique¹⁴, Japan Industrial Technical Center, Dec. 20, 1986, ρρ. Ï29-I3O. An example of a filter coefficient substitution method of conventional echo cancellers is disclosed in (2) Japanese laid-open patent publication No. 62-24726. The construction of the conventional echo canceller will now be described with reference to Figures 8 - 14 of the accompanying drawings.

Figure 8 labelled PRIOR ART, is a block diagram of the conventional echo canceller 30 explained in the above literature (1). In the echo canceller shown in Figure 8, Rin (k) represents a reception signal at k (where, k represents a certain time of a digital signal). In other words, it represents the voice of the caller at the far end. In the case of an echo canceller used for a telephone with a loudspeaker, an echo path 1 corresponds to a path whereby the voice of the far-end caller, reproduced from the loudspeaker, goes around into the telephone's microphone via sound space. In the case of an echo canceller used for a satellite communication channel, the echo path 1 represents the path whereby the voice of the far-end caller ’’leaks” into that of the near end talker as a result of impedance mismatching of a hybrid transformer used for two-wire / four wire conversion.

Since the echo signal passes through the echo path 1, a portion of the reception signal Rin (k) is superposed on the near-end voice Nin (k). Therefore, a transmission signal Tin (k) contains not only the near-end voice but also the superposed fhr-end voice echo. If Tin (k) is transmitted to the far-end as it is, the communication quality deteriorates. In the echo canceller 30, an adaptive filter 3 generates a pseudo echo signal Tin¹ (k). An echo subtracter 2 subtracts the pseudo echo signal Tin* (k) from the transmission signal Tin (k) to generate a residual signal Res(k) in which the echo component is cancelled. The residual signal Res (k) is transmitted to the far-end.

An adaptive controller 4 controls the adaptive filter 3. When duplex speech is detected, in other words when the far-end caller and the near-end caller are both speaking simultaneously, the adaptive controller 4 stops the adaptation. When only the far-end caller

CA 02199566 2004-08-24 is speaking, the adaptive controller 4 outputs an adaptive control flag FLG to the adaptive filter 3 in order to activate the adaptation.

The adaptive controller 4 will hereinafter be described further in detail, with reference to Figure 9 which illustrates the adaptive controller 4 in the conventional echo canceller of Figure 8 disclosed in the above literature (1). A level calculator 5 calculates logarithmically converted power of the reception signal Rjn (k). In other words, it calculates a reception signal level Lrin (k) from a time k to a time prior to the L-th sampling according to formula (1) below, and outputs its result. A level calculator 6 calculates logarithmically converted power of the residual signal Res (k). In other words, it calculates the residual signal level Lres (k) from a time k to a time prior to the L-th sampling according to formula (2) below, and outputs its result. A level difference calculator 7 uses the signals output from level calculators 5 and 6 to calculate DL (k), a level difference between the reception signal level Lrin (k) and the residual signal level Lres (kJ, according to formula (3) below. In the following formulae, L denotes a duration from a time k to a time prior to the L-th sampling, and N denotes an average number to ensure the level difference DL (k).

<img file="CA2199566C_D0002.tif" />

Lres(k) = Iog_w

<img file="CA2199566C_D0003.tif" />

(1) (2)

DL(k) = £ iLrin(i)-Lrgs(f))/N (3)

The level difference DL (k) is equal to the sum of the amount of echo loss via the echo path and the amount of echo cancelling by the echo canceller. A level Compensator 8 calculates an estimated residual signal level Lres* (k) according to formula (4):

Zw'(*) = Lrin (*) - DL (*)

0)

Lres⁹ (k) represents an estimated value of the residual signal level Lres (k) at the time when only the echo signal of the reception signal exists in the transmission signal Tin (k), in other words when only the far-end caller is speaking.

The comparator 9 compares the residual signal level Lres (k) and the estimated residual signal Lres' (kJ. If the result meets the conditional formula (5) below, the

CA 02199566 2004-08-24 comparator 9 judges that both the far-end caller and the near-end caller are speaking, that is to say, that duplex speech condition is occurring. The comparator 9 then sets the adaptive control flag FLG to 0 in order to stop the adaptation of the adaptive filter.

Lres (A) &gt; Lres¹ (fc) (5)

FLG = 1 ” represents the order to activate the adaptation, and FLG = 0” represents the order to deactivate the adaptation.

Adaptation control is a process whereby activation or deactivation of the adaptation is controlled in dependence upon whether the near-end voice level is judged to have reached a level for stopping the adaptation processing. The activation and the deactivation of the adaptation should be controlled according to a relationship between the echo signal level in the transmission signal Tin (k) and the near-end voice level. When the level of near-end voice exceeds the level of the echo signal by the far-end signal, the adaptation processing is deactivated. On the other hand, as shown in Figures 10A - 10D, if the echo signal level exceeds the near-end voice level, or if the echo signal level exceeds the near-end voice level as increased by a certain margin, the adaptation should be activated. Figure 10A Illustrates a relationship between the near end voice level Nin (k) and the echo signal level Rin' (k). Figure 1ÛB illustrates near-end voice signal segments or periods in the reception signal Tin(k). Figure IOC illustrates an echo signal segment or period in the reception signal Tin (k). Figure 10D illustrates periods where the adaptation of the conventional adaptive filter should be executed.

Figure Π A illustrates a relationship between the near end voice level Nin (k) and the echo signal level Rin’(k). Figure 1 IB illustrates near-end voice signal segment or periods in the reception signals Tin (k), Figure 11C illustrates an echo signal segment in the reception signal Tin (k). Figure 1 ID illustrates a relationship between residual signal level Lres (k) and the estimated residual signal level Lrcs‘(k). Figure 11Ë illustrates periods where the adaptation of the conventional adaptive filter is executed.

In the conventional art, duplex speech is detected from the residual signal level Lres (k) and the estimated residual signal level Lres* (k) in the comparator 9. Therefore, if the residual signal level Lres (k) and the estimated residual signal level Lres’ (k) are nearly equal in value, as shown in Figure 1 ID, the adaptation is deactivated by partly mis-detecting the condition as a duplex speech condition, even in period (a) where the transmission signal Tin (k) comprises only the echo signal. In addition, if the residual signal level Lres (k) exceeds the estimated residual signal level Lres* (k), adaptation is deactivated, even in period (b) shown in Figure 11 where the adaptation should be carried out because echo signal level exceeds the near-end voice level.

Figures 12 A - 12E illustrate another example of the relationship between the near end voice signal level and the echo signal level and periods where the adaptation of the

CA 02199566 2004-08-24 conventional adaptive filter is executed. In Figures 12Λ - Figure 12E, if the residual signal level Lres (k) exceeds the estimated residual signal level Lres* (k), adaptation is deactivated, even in period (b) shown in Figure 12C where the adaptation should be carried out because echo signal level exceeds the near-end voice level.

Filter coefficient substitution in the conventional echo canceller will now be described with reference to Figure 13 and Figures 14A - 14C.

Figure 13 is a block diagram of a conventional echo canceller as disclosed in the above literature (2). The construction of the conventional example of Figure 13 is different from that of the conventional echo canceller shown by way of example in Figure 8, in that the adaptive filter 3 of Figure 8 comprises a first (adaptive) memory 10, a sum/ product calculator 11 and a filter coefficient updater 14 as shown in Figure 13 . Also, the echo canceller of Figure 13 comprises a switch means 12, a second (refuge) memory 13, and a coefficient substitution controller 15, but does not comprise the adaptive controller 4. Elements in Figure 13 which are the same portions or the corresponding portions in Figure 8 have the same reference numbers. Accordingly the detailed explanation ofthe same portions is omitted.

The adaptive memory 10 stores a first filter coefficient for updating which is generated from the transmitting signal and the residual signal after echo cancellation using an algorithm. The refuge memory 13 stores the first filter coefficient which is to be stored in the adaptive memory 10 in parallel for predetermined periods while the adaptation is judged to be proper. The coefficient substitution controller 15 controls whether the update of the filter coefficient used for the adaptation of the adaptive filter is proper or not, according to the difference between the transmitting signal level and the residual signal level.

First of all, the filter coefficient updater 14 updates the filter coefficient H(j) of the adaptive memory 10 from the transmission signal Tin (Ic) and the residual signal Res (k) according to a filter algorithm, such as the N-LMS method (learning identification method). The sum/product calculator 11 generates a pseudo echo signal by carrying out a convolution operation ofthe reception signal Rin (k) and the filter coefficient H(j) stored in the adaptive memory 10. On the other hand, the coefficient substitution controller 15 calculates the logarithmically converted power Ltin (k) of the transmission signal Tin (k) and the logarithmically converted power Lres (k) of the residual signal Res(k) according to formulae (6) below and (2), respectively, and farther calculates the difference Ds(k) between Ltin (k) and Lres (k) according to formula (7) below.

Ltin (k) = log,,· k

Σ

<img file="CA2199566C_D0004.tif" />

(6)

CA 02199566 2004-08-24

Ds (A) = Ltitf (it) - Lns (A) (7)

Figures 14À - Figure 14C show timing charts of Ds (k), of transfer processing and of substitution processing by the switching device 12, In Figure 13 and Figure 14B, H(j) represents the first filter coefficient of the adaptive memory 10, and Hm(j) represents the second filter coefficient of the refuge memory 13. If Ds (k) increases, as shown in Figure 14 A, transfer timing pulses in Figure 14B are generated in response to the increase of the Da (k) and the first filter coefficient of the adaptive memory 10 is transferred to the refuge memory 13 in order to store the first filter coefficient H (j) as a second filter coefficient Hm(j). On the other hand, if Ds (k) decreases, as shown in Figure 14A, substitution timing pulses in Figure 14C are generated in response to the decrease of Ds (k) and the second filter coefficient of the refiige memory 13 is transferred to the adaptive memory 10 in order to substitute the second filter coefficient HmQ) for the first filter coefficient H(j)

Next, coefficient substitution controller 15 monitors the time sequence of Ds (k)

When Ds (k) becomes larger, the coefficient substitution controller 15 judges that the adaptation of the adaptive filter is running correctly, and sends an order to the switching device 12 so as to cause it to transmit the first filter coefficient of the adaptive memory 10 to the refuge memory 13. The switching device 12 then transmits the first filter coefficient to the refùge memory 13 as ordered. As a result, the first filter coefficient of the adaptive filter 10 is stored in the refuge memory 13 as a second filter coefficient.

On the other hand, when Ds (k) becomes smaller, the coefficient substitution controller 2 5 judges that the adaptation of the adaptive filter is running incorrectly, and sends an order to the switching device 12 so as to substitute the first filter coefficient of adaptive memory 10 with the value of the second filter coefficient stored in the refuge memory 13. Then the switching device 12 transfers the second filter coefficient to the adaptive memory 10, as ordered.

In the conventional echo canceller discussed above, duplex speech cannot be detected properly, and as a result, adaptation is deactivated even in the period where the adaptation should be carried out. Since the adaptation is deactivated frequently, the processing of the adaptation is slow and therefore the echo still remains without being cancelled. Another problem is that, when the amount of echo cancellation by the adaptive filter is small, the residual echo prevents clear listening.

It is an object of the present invention to eliminate or at least mitigate these problems and to provide an echo canceller which determines duplex speech more precisely and performs adaptation more adequately.

CA 02199566 2004-08-24

SUMMARY OF THE INVENTION

According'to a first aspect of the invention, there is provided an echo canceller comprising an adaptive filter responsive to a reception signal and an adaptation control signal to produce a pseudo error signal, means for subtracting said pseudo error signal from the 5 transmission signal to produce a residual signal, and adaptation control means responsive to the residual signal and the reception signal to provide said adaptation control signal, wherein said adaptation control means comprises;

first level calculating means for calculating the logarithmically converted power ofthe reception signal;

second level calculating means for calculating the logarithmically converted power of the transmission signal;

echo gain estimation means for estimating an echo gain based upon the logarithmically converted power of the reception signal and the logarithmically converted power of the transmission signal;

echo level estimating means for adding the estimated echo gain to the logarithmically converted power of said reception signal to obtain an estimated logarithmic echo level; and comparator means for providing said adaptation control signal in dependence upon a comparison between said estimated logarithmic echo level and a signal derived from or comprising said transmission signal, wherein said first level calculating means calculates the logarithmically converted power ofthe reception signal according to the following formula,

I™ (*) = ^IoSio jj? R*» (O’ j where k represents a certain time of digital signal, L represents a time lor L-th sampling time, and Rin(k) represents a reception signal at k, said second level calculating means calculates the logarithmically converted power of the transmission signal according to the following formula,

Ltm($-logjJ £ 7^(0² where Tin(k) represents a transmission signal at k, said echo gain estimation means estimates said echo gain according to the following formula,

CA 02199566 2004-08-24 *

EG 53 (Ltin (f) - Lrin (i)} /1 where I represents a preselected average sampling number, Ltin(i) represents the logarithmically converted power of the transmission signal Tin(k) at i₉ and Lrin(i) represents the logarithmically converted power of the reception signal at i, and said echo level estimating means outputs said estimated logarithmic echo level according to the following formula;

Lech (A) = Inn (A) * EG

According to one preferred embodiment of the first aspect of the invention, the adaptation control means further comprises;

third level calculating means for calculating logarithmically converted power of the residual signal; and said comparator means compares the estimated logarithmic echo level with the logarithmically converted power of the residual signal and produces said adaptation control signal such that the echo of the reception signal contained in the transmission signal is 15 cancelled when said estimated logarithmic echo level exceeds said logarithmically converted power of the residual signal by a prescribed margin.

The comparator means may be operable to compare the estimated logarithmic echo level with the logarithmically converted power of the transmission signal to produce said adaptation control signal such that the echo of the reception signal contained in the 20 transmission signal is cancelled when said estimated logarithmic echo level exceeds a value obtained by subtracting a predetermined margin from said logarithmically converted power of the transmission signal.

According to a further preferred embodiment of the first aspect of the invention, the adaptation control means further comprises:

third level calculating means for calculating logarithmically converted power of a residual signal;

said comparator means comprises;

first comparator means for comparing the estimated echo level with the logarithmically converted power of the residual signal level after the echo cancellation and providing a first adaptive control signal in dependence thereupon; and second comparator means for comparing the estimated echo level with the logarithmically converted power ofthe transmission signal level and providing a second adaptive control signal in dependence thereon;

CA 02199566 2004-08-24 a selector for selecting the first adaptive control signal or the second adaptive control signal; and echo cancellation amount estimating means for estimating an amount of echo cancellation from the logarithmically converted power of the transmission signal and the logarithmically converted power of the residual signal and controlling the selector to output either the first adaptive control signal or the second adaptive control signal as said adaptation control signal in dependence upon the amount of echo cancellation, such that;

the echo of the reception signal included in the transmission signal is cancelled if the estimated echo level exceeds the value obtained by adding afirst margin to the logarithmically converted power of the residual signal in the event that the amount of the echo cancellation exceeds a threshold value, and the echo of the reception signal contained in the transmission signal is cancelled if the estimated echo level exceeds a value obtained by subtracting a second margin from the logarithmically converted power of the transmission signal in the event that the amount of the echo cancellation does not exceed the threshold value.

The logarithmically converted powers could be replaced by the actual power of the respective signals (i.e. no logarithmic conversion).

Thus, according to a second aspect of the invention, there is provided an echo canceller comprising an adaptive filter responsive to a reception signal and an adaptation control signal to produce a pseudo error signal, means for subtracting said pseudo error signal from the transmission signal to produce a residual signal, and adaptation control means responsive to the residual signal and the reception signal to provide said adaptation control signal, wherein said adaptation control means comprises:

first level calculating means for calculating the power of the reception signal; second level calculating means for calculating the power of the transmission signal; echo gain estimation means for estimating an echo gain based upon the power of the reception signal and the power of the transmission signal;

echo level estimating means for adding the estimated echo gain to the power of said reception signal to obtain an estimated echo level ; and comparator means for providing said adaptation control signal in dependence upon a comparison between said estimated echo level and a signal derived from or comprising said transmission signal, wherein said first level calculating means calculates the power of the reception signal according to the following formula, k

Lrin(k)-= £

CA 02199566 2004-08-24 where k represents a certain time of digital signal, L represents a time for L-th sampling time, and Rin(k) represents a reception signal at k, said second level calculating means calculates the power of the transmission signal according to the following formula, fc

Ltin(ti) = £ Tin (if /k-£+l where Tin (k) represents a transmission signal at k, said echo gain estimation means estimates said echo gain according to the following formula, k

EG* Σ iLtm(f)~lrin(fi}/J where 1 represents a preselected average sampling number, Ltin(i) represents the power of the transmission signal (Tin(k)) at /, and Lrin(i) represents the power of the reception signal (Rin(k)) at /, and said echo level estimating means outputs said estimated echo level (Lech (k)) according to the following formula.

Lech(k)=Lrin(k)*EG

Preferred embodiments of the second aspect of the invention may be similar to the preferred embodiment of the first aspect of the invention, but using non-logarithmic power rather than logarithmic power, as appropriate.

According to a third aspect of the invention, an echo canceller for deriving a pseudo echo signal (Tin’(k)) and subtracting the pseudo echo signal from a transmission signal (Tin(k)) to produce a residual signal (Res(k)), comprises:

a first memory (10) for storing a first filter coefficient (H(j)) for adaptation; a second memory ( 13) for storing a second filter coefficient (HmQ)) *ⁿ parallel during the adaptation process for a predetermined period of time during which cancellation is determined to be effective;

coefficient substitution means (15) forcalculatingthe logarithmically converted power of the transmission signal and the logarithmically converted power of the residual signal and for controlling whether the first filter coefficient in the first memory is updated or not according to the difference between the logarithmically converted power of the transmission signal (Tin(k)) and the logarithmically converted power of the residual signal(Res(k));

CA 02199566 2004-08-24

9/1 interpolating means (23) for interpolating between the value of the first filter coefficient (H(j)) in the first memory and the value of the second filter coefficient (Hm(j)) in the second memory;

a filter coefficient updater ( 14) for updating the first filter coefficient (H(j)) in the first memory according to the transmission signal (Tin(k)) and the residual signal (Rcs(k));

calculator means (11) for generating the pseudo echo signal (Tin'(k)) by carrying out a convolution operation upon the reception signal (Rin(k) and the first filter coefficient (H(j)) stored in the first memory; and an echo subtractor for subtracting the pseudo echo signal (Tin'(k)) from the transmission signal (Tin(k)) to produce the residual signal (Res(k));

wherein said coefficient substitution means (15) controls substitution of the first filter coefficient (H(j)) by the second filter coefficient (Hm(j)) in the event that the adaptation does not provide a desired level of echo cancellation, and said interpolation means (23) carries out interpolation and updates for a second predetermined period of time so that the value of the first filter coefficient (H(j) is continuous to the value of the second filter coefficient (Hm(j)).

Embodiments of the three aspects of the present invention will now be described, by way of example only, with reference to Figures 1 - 7 of the accompanying drawings, in which Figures 8 - 14, labelled PRIOR ART, illustrate construction and operation of a known, conventional echo canceller.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a simplified block schematic diagram of an adaptive controller of an echo canceller, constituting a first embodiment of the present invention;

Figures 2 A - 2E illustrate operation of the adaptive controller in the first embodiment of this invention;

Figure 3 is a simplified block schematic diagram of the adaptive controller of an echo canceller constituting a second embodiment of this invention;

Figures 4À - 4E illustrate the operation of the adaptive controller in the second embodiment of this invention;

Figure 5 is a simplified block schematic diagram of the adaptive controller of an echo canceller constituting a third embodiment of this invention;

Figure 6 is a simplified block schematic diagram of an echo canceller which constitutes a fourth embodiment of this invention and employs filter coefficient substitution;

Figures 7 A- 7C illustrate the operation of the coefficient substitutcr of the fourth embodiment of this invention;

Figure 8 is a simplified block schematic diagram of a PRIOR ART conventional echo canceller;

Figure 9 is a simplified block schematic diagram of an adaptive controller of the conventional echo canceller;

Figures 10A - 10D illustrate periods in which adaptation of the adaptive filter of the conventional echo canceller should perform its adaptation;

Figures 11A - 11E illustrate periods in which adaptation of the conventional adaptive filter takes place;

Figures 12A - 12E illustrate other periods in which the adaptation of the conventional adaptive filter takes place.

Figure 13 is a simplified block schematic diagram of a known echo canceller which uses conventional filter coefficient substitution; and

Figures 14A - 14C illustrate performance of the conventional coefficient substitution method.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1

The first embodiment which improves the adaptive control of the adaptive filter will be described with reference to Figure 1, in which elements which are the same or correspond to portions in Figure 9 have the same reference numbers. Accordingly the detailed explanation of the same portions is omitted.

The adaptive controller 4 is an element of an echo canceller similar to that shown in Figure 8. Referring to Figure 1, the adaptive controller 4 receives the digital signal reception signal Rin (k), the transmission signal Tin (k) and the residual signal Res (k) and outputs an adaptive control flag FLGa.

Three level calculators 5, 16, 6 calculate and output a reception signal level Lrin (k), a transmission signal level Ltin (k), and a residual signal level Lres (k), respectively. The word level here denotes a logarithmic conversion value of the signal power.

When the reception signal level Lrin (k) exceeds the sound threshold value, an echo gain estimator 17 calculates an estimated echo gain EG according to formula (8) below from the reception signal level Lrin (k) and the transmission signal level Ltin (k). In other words, the echo gain estimator 17 calculates the equivalent transmission gain of the reception signal appearing at the transmission end via the echo path, as an average value of the reception signal over several consecutive time periods. This value is normally fixed in a system, but in embodiments of the present invention it is calculated occasionally, for example, once a several hundred milliseconds.

The adaptive controller 4 can estimate the echo gain precisely, because it uses the average value of the level differences between the reception signal level and the transmission signal level at the time when the reception signal has a clearly detectable signal level because someone is speaking.

k

EG = 52 (0 (0)// (8)

1=4+/+1

An echo level estimator 18 calculates an estimated echo level Lech(k) according to formula (9) below. That is to say, the echo level estimator 18 estimates the amount of reception signal appearing as the echo in the transmission signal in the event that the near-end signal Nin (k) does not exist.

Lech (k) = Lrin (k) + EG (9)

The comparator 19 compares the value of the residual signal level Lres (k), increased by the margin a, with the estimated echo level Lech (k), and sets the adaptive control flag FLGa to 1 when such comparison meets the conditional formula (10) below. It sets the adaptive control flag to 0 when it does not meet the conditional formula (10). The comparator 19 then outputs the adaptive control flag FLGa.

Lech (k) z Lres (k) + a (10)

Operation of the adaptive controller 4 will be described with reference to Figure 2 for the case where the amount of the echo cancellation is large, in which case the adaptation is fully activated and the echo is adequately cancelled. In order to set the adaptive control flag FLGa of the invention, the residual signal level Lres (k) is compared with the estimated echo level Lech (k), as shown in the conditional formula (10).

Figure 2A illustrates a relationship between the near-end voice level Nin (k) and the echo signal level Rin’ (k). Figure 2B illustrates a near-end voice signal segment or period in the transmission signal Tin (k). Figure 2C illustrates an echo signal segment in the transmission signal Tin (k). Figure 2D illustrates a relationship between residual signal level Lres (k) + margin a and the estimated echo signal level Lech (k). Figure 2E illustrates periods where the adaptation of the conventional adaptation filter is executed.

When the amount of echo cancellation is large, the residual signal level Lres 0c) is almost equal to the near-end voice level, and the estimated echo level Lech (k) is almost equal to the echo signal level. As stated above, the near-end voice level and the echo signal level are ideal parameters for use in controlling the adaptation. Therefore, embodiments of the present invention, use more preferable parameters, as compared with the conventional example using the residual signal level Lres 0c) and the estimated residual signal level Lres’ (k).

Hence, adaptation can be activated over more periods than the adaptation of the conventional example depicted in Figure 11. That is to say, in the period where the transmission signal Tin (k) comprises only the echo signal, as shown in period (a) in Figure 2C, the estimated echo level Lech (k) always exceeds the residual signal level Lres (k) + margin a , i.e. the approximation of the near-end voice level, as shown in Figure 2D. Therefore, the adaptive control flag FLGa is continuously set to 1 as shown in Figure 2E. Moreover, even in periods such as period (b) in Figure 2C, where the echo signal level exceeds the near-end signal level during duplex speech, the estimated echo level Lech (k) exceeds the residual signal level Lres (k) + margin a in the same manner as in period a. Therefore, the adaptive control flag FLGa is set to 1 as shown in Figure 2E, during period (b) in Figure 2C.

The margin a in the conditional formula (10) is used to absorb an error which is derived by approximating the near-end voice level by the residual signal level Lres(k), and an error which is deviated from the true echo level, since the echo level Lech(k) is obtained by the estimation.

As described above, according to this invention, a better adaptation control can be achieved by the comparison of the above signals when the amount of the echo cancelling is large.

Embodiment 2

A second embodiment of the invention which improves adaptive control of the adaptive filter will now be described with reference to Figure 3 which is a block diagram of the adaptive controller 4 of the second embodiment of this invention. Those elements in Figure 3 which are the same portions or corresponding portions as in Figure 1 and Figure 9 have the same reference numbers. Accordingly the detailed explanation of the same portions is omitted.

This adaptive controller 4 corresponds to the adaptive controller 4 in the echo canceller of Figure 8. The adaptive controller 4 of Figure 3 receives the reception signal Rin (k) and the transmission signal Tin (k), both of which are digital signals, and outputs an adaptive control flag FLGb.

Two level calculators 5, 16 calculate and output the reception signal level Lrin 0c) and the transmission signal level Ltin (k), respectively, in the same way as the corresponding elements in the first embodiment. The echo gain estimator 17 and the echo level estimator 18 calculate the estimated echo gain EG and the estimated echo level Lech (k), respectively, in the same way as in the first embodiment. However, in this second embodiment, a comparator 20 compares the estimated echo Lech (k) with the transmission signal level Ltin (k).

When the transmission signal level Ltin (k), margin β, and the estimated echo level Lech (k) meet the conditional formula (11) below, the comparator 20 sets the adaptive control flag FLGb to 1 . When they do not meet the conditional formula (11), the comparator 20 sets the adaptive control flag FLGb to 0.

Lech £ Ltin (Æ) - β (11)

The operation of the adaptive controller 4 of Figure 3 will now be described with reference to Figures 4A - 4E for the case where the amount of the echo cancellation is small; in other words for the case where the adaptation is not fully activated and the echo is not adequately cancelled. In order to set the adaptive control flag FLGb of the

CA 02199566 2004-08-24 invention, the transmission signal level Ltin (k) is compared with the estimated echo level Lech (k), as shown in the conditional formula (11) above.

Figure 4À illustrates a relationship between the near-end voice level Nin (k) and the echo signal level Rin* (k). Figure 4B illustrates near-end voice signal segments or periods in the reception signals Tin (k). Figure 4C illustrates an echo signal segment in the reception signal Tin (Ic). Figure 4D illustrates a relationship between transmission signal level Ltin (k), margin β and the estimated echo signal level Lech (k). Figure 4E illustrates periods where the adaptation of the adaptation filter is executed.

When the transmission signal Tin (k), does not contain, near-end voice, the transmission signal level Ltin (k) is the same as the echo signal level. Therefore the transmission signal level Ltin (k) becomes almost equal to the estimated echo level Lech (k). During duplex speech, where the near-end signal and the far-end signal overlap, since the transmission signal level Ltin (k) includes an overlapped near-end voice and echo signal, the transmission signal level (Ltin (k) - margin β) exceeds the estimated echo level Lech (k). On the other hand» in the periods where the transmission signal Tin (k) includes only the echo signal, the estimated echo level Lech (k) exceeds the transmission signal level (Ltin (k) margin β) as shown in Figure 4D.

Accordingly, by comparing the transmission signal level (Ltin (k) - margin β) with the estimated echo level Lech (k), the periods where the transmission signals Tin (k) includes only the echo signal can be accurately estimated, and the adaptive control flag FLGb is set to 1 as shown in Figure 4Ë. Thus, the adaptation of the present invention can be activated over more periods than that of the conventional examples depicted in Figure 12E.

The margin β in the conditional formula (11) is used to absorb an error which i$ a deviation from the true echo level, since the echo level Lech (k) is obtained by the estimation.

As described above, according to this invention, a better adaptation control can be achieved by the comparison of the above signals when the amount of the echo cancelling is small.

Embodiment 3

A third embodiment of the invention, which improves the adaptive control of the adaptive filter, will now be described with reference to Figure 5 which is a block diagram of the adaptive controller 4 in the third embodiment of this invention. Those elements which are the same portions or the corresponding portions in Figures 1, 5 and 9 have the same reference numbers. Accordingly the detailed explanation of the same portions is omitted.

The adaptive controller 4 in Figure 5 also corresponds to the adaptive controller 4 in the echo canceller of Figure 8. The input and the output of the adaptive controller 4 are the same as those of the adaptive controller ofFigure 1. Three level calculators 5,16,6 calculate and output the reception signal level Lrin (k), the transmission signal level Ltin (k) and the

CA 02199566 2004-08-24 residual signal level Lres (k), respectively, in the same way as in the first embodiment. The word ’'level·* here denotes the logarithm conversion value of power.

The echo gain estimator 17 and the echo level estimator 18 calculate the estimated echo gain EG and the estimated echo level Lech (k), respectively, in the same way as in the first embodiment. Since the adaptive controller 4 uses the average value of the level difference between the reception signal and the residual signal when the reception signal is obviously recognized as sound, the amount of the echo cancellation can be estimated precisely. A main difference between the adaptive controller 4 in Figure 5 and the adaptive controller of the first and second embodiments (Figures 1 and 3 respectively) is that the adaptive controller of Figure 5 comprises two comparators 19 and 20 that set the adaptive control flags FLGa and FLGb, respectively, and output the two control flags to a selector 22.

When the reception signal level Lrin (k) exceeds the sound threshold value, an echo cancellation amount estimator 21 calculates an estimated echo cancellation ERLE from the transmission signal level Lrin (k) and the residual signal level Lres (k) according to formula (12):

k

ERLE= Σ ^Ltinfi) - Lres (1)}/K (12) f=A-r+i where K represents an average sampling number which makes the estimated echo gain ERLE reliable.

When the estimated echo cancellationERLE exceeds the sound threshold value which is a level when the reception signal includes a voice signal, in other words when the estimated echo cancellation ERLE is relatively large, the flag selector 22 outputs the adaptive control flag FLGa outputted from the comparator 19 as a final adaptive control flag FLG. When the estimated echo cancellation ERLE does not exceed the threshold value, in other words when the estimated echo cancellation ERLE is relatively small, the flag selector 22 outputs the adaptive control flag FLGb outputted from the comparator 20 as a final adaptive control flag FLG. FLG 1 represents the adaptation activation order to the adaptive filter, and FLG =”0 represents the deactivation order to the adaptive filter. The two adaptive control flags (FLGa, FLGb) are used so as to improve the adaptation control regardless of the amount of echo cancellation.

As already explained above in relation to the first and the second embodiments, the adaptive control flag FLGa is effective for better adaptation control when the echo cancellation is large, and adaptive control flag FLGb is effective when the echo canceUation i$ small.

As explained above, this embodiment of the invention makes it possible to perform the adaptation control better regardless of the amount of the echo cancellation, because both

CA 02199566 2004-08-24 adaptive control flags FLGa and FLGb are employed according to the amount of the echo cancellation.

In addition, margins ol and β can absorb an error derived by approximating the nearend voice level by the residual signal level Lres(k), and an error which is deviated from the true echo level, since the echo level Lcch(k) is obtained by the estimation.

Embodiment 4.

À fourth embodiment of the invention, which deactivates the noise generation when substituting the filter coefficient of the sum/ product calculator according to the adaptation control, will be described with reference to Figure 6, which is a block diagram of an echo canceller equipped with the filter coefficient substitution device of the fourth embodiment of the invention. The same or corresponding elements in Figure 6 and Figure 13 have the same reference numbers and are not described in detail here, since those elements are described above in connection with Figure 13.

Referring to Figure 6, upon receiving an order from the coefficient substitution controller 15 to transfer the value of the first filter coefficient in the adaptive memory 10 to the refuge memory 13, the switching device 12 transfers the value of the filter coefficient according to the procedure used by the conventional echo canceller described with reference to Figure 13. However, in the present embodiment, upon receiving an order to replace the value of the first filter coefficient in the adaptive memory 10 with the value in the refuge memory 13, the switching device 12 does not respond at all. Thia is a significant difference as compared with the prior art.

When an interpolator 23 receives the order from the coefficient substitution controller 15 to replace the value of the filter coefficient in the adaptive memory 10 with the value in the refuge memory 13, in other words when the replacement of the filter coefficient takes place because the adaptation of the current value is not proper, the interpolator 23 compares the first filter coefficient H(j) of the adaptive memory 10 and the second filter coefficient Hm(j) of the refuge memory 13, calculates the difference and obtains an interpolation width DLO), where, j “ 1, 2.,., J, in the event that the interpolation is performed for M divisions according to formula (13). The resultant interpolation width DL(j) is stored in the refuge memory 13. J represents the number of the taps of the adaptive filter.

DLtf) = wherej ^-1,2,.....J (13)

Figure 7A illustrates a timing chart of Ds (k) for the present fourth embodiment. Figure 7B illustrates a timing signal for controlling transfer of the signal from HQ) to Hm G)

CA 02199566 2004-08-24 by the switching device 12. Figure 7C illustrates a tuning signal for interpolating HQ) at the interpolator 23.

In Figure 6 and Figure 7, HQ) represents the first filter coefficient of the adaptive memory 10, and HmQ) represents the second filter coefficient of the refiige memory 13. If 5 Ds(k) increases as shown in Figure 7 A, transfer timing pulses in Figure 7B are generated in response to the increase of Ds (k) and cause the first filter coefficient of the adaptive memory to be transferred to the refuge memory 13 in order to store the first filter coefficient H(j) as a second filter coefficient Hm (j). On the other hand, ifDs (k) decreases as shown in Figure 7A, many substitution timing pulses, as shown in Figure 7C, are generated in response to the 10 decrease of the Ds (k) and cause the second filter coefficient of the refuge memory 13 to be transferred to the adaptive memory 10 in order to substitute the second filter coefficient HmQ) for the first filter coefficient HQ). In this case, the second filter coefficient of the refiige memory 13 is transferred to the adaptive memory 10 using many interpolation timing

<img file="CA2199566C_D0005.tif" />

In the prior art, as stated above, if the first filter coefficient value and the second filter coefficient value differ considerably, the first filter coefficient changes considerably at the moment when the substitution is carried out. Therefore, the output of the sum/product calculator 11 becomes discontinuous because the substitution is carried out by only one-time substitution. Alternatively, in this embodiment, since the substitution of the first filter coefficient is carried out by M-time interpolation, that is, the substitution is carried out gradually, the discontinuity does not occur, even if the first filter coefficient and the second filter coefficient value are considerably different. The interpolator 23 controls addition of the interpolation width DL(j) to the first filter coefficient of the adaptive memory 10 to obtain the second filter coefficient Hm(j) of the second, refuge memory 13, for every j, where, j = 1, 2..., J. The second filter coefficient Hm(j) of the second or refuge memory 13 is obtained by carrying out the following calculation:

Hm(j) = H(j) + DL(j) : (j = 1, 2, .....J) where H(j) is the first filter coefficient of the adaptive memory 10 and DL(j) is the interpolation width.

As a result, after the interpolator 23 has carried out the interpolation for M times, the value of the filter coefficient in the adaptive memory 10 becomes equal to the value in the adaptive memory Hm(j), where, j=l, 2........J, before the interpolation begins.

Embodiment 5.

In the first, second and third embodiments, the reception signal level Lrin (k), the transmission signal level Ltin (k) and the residual signal level Lres (k), which are calculated in the calculating means 5, 16 and 6, respectively, are the logarithmic conversion values of the power. However, in a fifth embodiment, the reception signal level Lrin (k), the transmission signal level Ltin (k) and the residual signal level Lres (k) can be replaced by the power level itself. In this fifth embodiment, in order for the adaptive controller 4 then to operate in the same manner as in the first, second and third embodiments, the various calculations according to formulae (8),(9),(10),(11) and (12) are carried out instead according to formulae (14),(15),(16),(18) and (20), respectively, as given below. The values for a and β are given by formulae (17) and (19), respectively.

2197566 k

EG= Σ {Ltin (fi/Lrin (ί)}/1 (14)

1=4-/+1

<td> Lech (k) = Lrin (k) x EG</td><td> (15)</td>

<td> Lech (k) £ Lres (it) x a^z</td><td> (16)</td>

<td> a'^lO*'¹⁰</td><td> (17)</td>

<td> Lech (Jfc) £ Lting</td><td> (18)</td>

<td> p/ _{= lo}«/io</td><td> (19)</td>

<td> 4 ERLE = 52 {Ltin (ij / Lres (i) }{K</td><td> (20)</td>

i=4-£+l

Embodiment 6.

Furthermore, the formula (14) for estimating the echo gain EG and the formula (20) for estimating the echo cancellation ERLE may be replaced by the formula (21) and the formula (22), respectively.

4 eg⁼ Σ $ i Σ (0 (²1) /=4-/+1 i=4-/+l

4

ERLE = 52 ^(0/ Σ ^(0 (²²) i=4-£+l i=4-JT+l

In echo cancellers embodying the present invention, the substitution value is not discontinuous and does not produce significant noise at a moment of substitution, even when the filter coefficient in the adaptive memory is substituted by the refuge value.

CA 02199566 2004-08-24