JP2005142886A - Signal processor and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce errors in echo suppression processing in a communication device. <P>SOLUTION: The signal processor is used in the communication device using a transmitter and a receiver that are independent each other. The signal processor comprises a synchronous control unit for supplying a synchronous signal; a first buffer means for storing a receiving signal from one's party by a frame unit to output the receiving signal to the receiver, based on the synchronous signal; a second buffer means for storing a transmitting signal from the transmitter including an echo outputted from the receiver by the frame unit to output the transmitting signal, based on the synchronous signal; a pseudo echo signal generating means for generating a pseudo echo signal, based on the transmitting signal to be transmitted to one's party and the receiving signal stored in the first buffer means, and for outputting the pseudo echo signal synchronized with the transmitting signal outputted from the second buffer signal based on the synchronous signal; and a subtraction means for subtracting the pseudo echo signal from the transmitting signal outputted from the second buffer means to output the transmitting signal to be transmitted to one's party. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a signal processing device such as an echo suppressor or an echo canceller in a communication device, and a computer program that causes a computer to function as the signal processing device.

  With the spread of the Internet in recent years, development of TCP / IP related technologies is progressing. As a result, in addition to basic Internet communication forms such as e-mail and netnews, new communication forms such as Internet broadcasting using streaming technology and IP telephone using VoIP (Voice over IP) technology have been put into practical use. ing. Among the new forms of communication on the Internet, IP telephones using VoIP technology can construct a flexible network at a lower cost than conventional circuit-switched telephone networks. Application is expected.

  Since the video conference system assumes the participation of multiple users, it may take the form of a loudspeaker phone that uses a microphone and speaker on the desk to exchange voice without using a handset or headset. Many. In a loudspeaker telephone that does not use such a handset, the other party's voice flowing from the speaker is input to the microphone and is likely to generate an echo. The echo not only lowers the intelligibility and deteriorates the communication quality in voice communication, but also causes howling and makes communication impossible. Usually, in a loudspeaker telephone system such as a video conference system, an echo suppressor or an echo canceller is used to prevent echo (for example, Patent Document 1).

  The echo suppressor pays attention to the fact that transmission and reception are often performed alternately in voice communication, and the voice signal having the lower signal level (intensity level) of transmission or reception is regarded as an echo. It is a device that suppresses signals. The echo canceller is a device that generates a pseudo echo signal from a reception voice signal and a transmission voice signal and subtracts the pseudo echo signal from the transmission voice signal.

  The echo canceller disclosed in Patent Document 1 is an echo canceller installed on a digital communication network such as ATM that is transmitted and received in cell units, and temporarily stores each of the far-end side cell and the near-end side cell. The pseudo echo creation is executed in synchronization with a predetermined timing. By performing such processing, it is possible to prevent the occurrence of echoes during the echo canceling process and to prevent erroneous processing when creating pseudo echoes.

By the way, since an IP telephone based on the VoIP technology is a system using the Internet protocol (IP), processing such as A / D-D / A conversion and encoding is often realized by software on a computer. In a computer that performs such processing, an operating system (OS) capable of multitasking is often used, but each signal processing is still processed as an independent process. Then, even if the far-end side cell and the near-end side cell are temporarily accumulated and the timing is adjusted as in the echo canceller disclosed in Patent Document 1, a signal processing process such as pseudo echo signal generation and subtraction is performed. Since they are independent of each other, there is a problem in that the audio signals to be processed (audio data in frame units in the VoIP technology) do not match and erroneous processing occurs. In addition, even when signal processing is performed by hardware without performing software processing, there is a problem that erroneous processing similarly occurs when pseudo-echo signal generation and subtraction are realized by separate hardware elements. there were.
Japanese Patent No. 3123906

As described above, a conventional signal processing apparatus such as an echo canceller and a computer program have a problem that an audio signal and audio data to be processed do not match and erroneous processing occurs.
The present invention has been made to solve such a problem, and an object of the present invention is to provide a signal processing apparatus and a computer program that enable echo suppression with less erroneous processing even in software processing.

  In order to achieve the above object, a signal processing device according to a first aspect of the present invention is a signal processing device in a communication device using a transmitter and a receiver that are independent from each other, and a synchronization control unit that supplies a synchronization signal; A first buffer means for storing the received signal from the other party in units of frames and outputting the received signal based on the synchronization signal; and a transmitted signal from the transmitter including the echo output from the receiver in units of frames A second buffer means for storing and outputting the transmission signal based on the synchronization signal, a pseudo echo signal based on the transmission signal transmitted to the other party and the reception signal stored in the first buffer means, A pseudo echo signal generating means for outputting a pseudo echo signal in synchronization with a transmission signal output from the second buffer means based on the synchronization signal, and a transmission signal output from the second buffer means And it comprises a subtraction means for outputting a transmission signal to be transmitted to the other party by subtracting Luo pseudo echo signal.

  A signal processing device according to a second aspect of the present invention is a signal processing device in a communication device using a transmitter and a receiver that are independent of each other, and includes a synchronization control unit that supplies a synchronization signal, and a reception signal from the other party in units of frames. A first buffer means comprising a storage section for storing and an attenuating section for attenuating the received signal and outputting the received signal to the receiver based on the synchronization signal; and a transmitter including an echo output from the receiver A second buffer means for outputting a transmission signal based on the synchronization signal, a storage section for storing the transmission signal in units of frames and an attenuation section for attenuating the transmission signal; The received signal and the transmission signal stored in the second buffer means are compared to determine the attenuation amount in the first buffer means and the attenuation amount in the second buffer means. And it includes a voice switch unit for setting the attenuation amount of the attenuation of each of Ffa means and the second buffer means.

  A computer program according to a third invention is a signal processing device in a communication device using a transmitter and a receiver that are independent from each other, and stores a synchronization control unit that supplies a synchronization signal and a reception signal from the other party in units of frames. The first buffer means for outputting the reception signal based on the synchronization signal and the transmission signal from the transmitter including the echo output from the reception unit are stored in units of frames, and the transmission signal is stored based on the synchronization signal. A second buffer means for outputting a speech signal; a pseudo-echo signal is generated based on the transmission signal transmitted to the other party and the received signal stored in the first buffer means; and from the second buffer means based on the synchronization signal. A pseudo echo signal generating means for outputting a pseudo echo signal in synchronization with an output transmission signal; and a pseudo echo signal from a transmission signal output from the second buffer means. Flip and characterized by causing a computer to function as a signal processing apparatus including a subtraction means for outputting a transmission signal to be transmitted to the other party.

  A computer program according to a fourth aspect of the present invention is a signal processing device in a communication device using a transmitter and a receiver that are independent from each other, and stores a synchronization control unit that supplies a synchronization signal and a reception signal from the other party in units of frames. First buffer means for outputting the received signal to the receiver based on the synchronization signal, and a transmitter from the transmitter including the echo output from the receiver. A second buffer means for outputting the transmission signal based on the synchronization signal, and a storage section for storing the speech signal in units of frames and an attenuation section for attenuating the transmission signal; and stored in the first buffer means The received signal and the transmission signal stored in the second buffer means are compared to determine the attenuation amount in the first buffer means and the attenuation amount in the second buffer means, and based on the synchronization signal. Characterized in that causes a computer to function as Te signal processing apparatus and a voice switch unit for setting the attenuation amount of the attenuation of each of the first buffer means and the second buffer means.

  In the present invention, a synchronization control unit that supplies a synchronization signal is provided to synchronize the operations of the first buffer means, the second buffer means, and the pseudo echo signal generation means, and the received signal, transmission signal, and pseudo echo signal Since each of them is synchronized, an echo cancellation process with few erroneous processes can be performed.

  In the present invention, the synchronization control unit for supplying the synchronization signal is provided to synchronize the reception signal and the transmission signal and to synchronize the operations of the first buffer unit, the second buffer unit, and the voice switch unit. Enables echo suppression with less processing.

  According to the present invention, it is possible to realize echo cancellation processing and echo suppression processing with less erroneous processing.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing the configuration of the signal processing apparatus according to the first embodiment of the present invention. A signal processing apparatus according to this embodiment encodes an A / D converted audio signal into audio data in units of frames, and in an IP telephone system that implements a voice call using the Internet Protocol (IP) by VoIP technology, a user terminal This is an echo canceller for suppressing an echo generated when sound output from a speaker (receiver) flows into a microphone (transmitter). In the following description, “near end side” refers to the side that produces echoes (the side that produces echoes from the microphone and speaker), and “far end side” refers to the other party to which the echoes are transmitted. doing.

  As shown in FIG. 1, the signal processing apparatus 1 according to this embodiment includes a reception buffer unit 11, a pseudo echo generation unit 12, a transmission buffer unit 13, a subtraction unit 14, and a synchronization control unit 15. These are realized by software on a computer. The reception signal input terminal 2 is a terminal for inputting a frame unit audio signal (reception audio data) transmitted from the far end side, and is connected to the reception buffer unit 11.

  The reception buffer unit 11 includes a storage unit 16 including a memory that stores received voice data transmitted from the far end side, and a control unit 17 that controls reading of the stored data. Then, the reception buffer unit 11 acts to prevent erroneous processing in units of frames in the echo cancellation processing by temporarily storing the received speech data and reading it at a predetermined timing. The reception buffer unit 11 is connected to the speaker 4 via a pseudo echo generation unit 12, a synchronization control unit 15, a D / A conversion unit (not shown), and the like.

  The pseudo-echo generation unit 12 uses an FIR filter as a tapped delay line filter that creates a pseudo-echo for removing the echo and an adaptive algorithm such as a learning identification method (NLMS: Normalized Least Mean Square). And an adaptive filter that calculates filter coefficients, and acts as a pseudo-echo generating means generally used as an echo canceller. As the pseudo echo generation unit 12, any adaptive algorithm may be used as long as the pseudo echo signal is generated based on the received voice data and the voice signal (transmitted voice data) in units of frames transmitted to the far end side. Any delay line filter may be used. For example, as an adaptive algorithm, a least mean square algorithm (LMS), a recursive least square algorithm (RLS), a projection method, or the like may be used instead of the learning identification method. Further, an IIR filter (infinite length impulse response filter) or the like may be used instead of the FIR filter. Alternatively, a subband type or a frequency domain type may be used. The pseudo echo generation unit 12 is connected to one input and output of the reception buffer unit 11 and the subtraction unit 14 and the synchronization control unit 15.

  The transmission buffer unit 13 includes a storage unit 18 including a memory that stores transmission voice data transmitted from the near end side, and a control unit 19 that controls reading of the stored data. The transmission buffer unit 13 works in cooperation with the reception buffer unit 11 to prevent erroneous processing in units of frames in echo cancellation processing. The transmission buffer unit 13 is connected to the microphone 5 via the other input of the subtraction unit 14, the synchronization control unit 15, an A / D conversion unit (not shown), and the like.

  The subtracting unit 14 is, for example, a subtractor, and has an action of subtracting the pseudo echo signal generated by the pseudo echo generating unit 12 from the transmission voice data output from the transmission buffer 13 and removing the echo signal component. To do. The subtraction output of the subtraction unit 14 is connected to the transmission signal output terminal 3 and the pseudo echo generation unit 12.

  The synchronization control unit 15 includes, for example, a reference signal oscillator or a function for acquiring internal time information of the system, and controls reading of received voice data and transmitted voice data to be processed and determines signal processing timing. . The synchronization control unit 15 is connected to the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13.

  In the signal processing device 1, the synchronization control unit 15 supplies a synchronization signal to the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13. Then, the signal processing operations of the reception buffer unit 11, the pseudo echo generation unit 12, the transmission buffer unit 13, and the subtraction unit 14 are synchronized to process each of the reception voice data, the transmission voice data, and the pseudo echo signal. Since the timing is synchronized, erroneous processing can be reduced even when echo cancellation processing is realized by software.

  Next, the operation of this signal processing apparatus will be described with reference to FIG. 1 and FIG. FIG. 2 is a flowchart showing the operation of the signal processing apparatus according to this embodiment. In the signal processing apparatus of this embodiment, the echo cancellation processing for the transmission audio data in units of frames is realized by software. Hereinafter, one cycle of this echo cancellation processing will be described.

  First, the synchronization control unit 15 processes reception voice data, transmission voice data, and pseudo echo signal, and operation timings of the reception buffer unit 11, the pseudo echo generation unit 12, the transmission buffer unit 13, and the subtraction unit 14. And is supplied to the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13 (step S21; hereinafter abbreviated as "S21"). The synchronization signal may be generated by using a reference oscillator output on a computer in which the signal processing apparatus is realized, or a reference signal may be supplied from the outside.

  When reception voice data in units of frames is input to the reception voice signal input terminal 2, the reception buffer unit 11 stores the reception voice data in the storage unit 16. And the control part 17 of the reception buffer part 11 reads the memorize | stored data, and outputs it to the speaker 4 via a D / A conversion part etc. (S22).

  When the reception voice data is stored in the storage unit 16 of the reception buffer unit 11, the pseudo echo generation unit 12 is simulated based on the reception voice data stored in the storage unit 16 based on the synchronization signal transmitted from the synchronization control unit 15. An echo signal is generated (S23). Specifically, a pseudo echo signal is generated by convolving the coefficient of the delay line filter provided in the pseudo echo generation unit 12 with the received voice data referred to from the storage unit 16. The generation timing of the pseudo echo signal is controlled by the synchronization signal from the synchronization control unit 15.

  A part of the sound output from the speaker 4 is input to the microphone 5 as an echo component EC together with the sound to be input. The voice and echo component EC input to the microphone 5 are input to the transmission buffer unit 13 via an A / D conversion unit or the like. The transmission buffer unit 13 stores the voice and the echo component EC in the storage unit 18. Then, the control unit 19 of the transmission buffer unit 13 reads the stored data based on the synchronization signal from the synchronization control unit 15 and outputs it to one input of the subtraction unit 14 (S24).

  The synchronization control unit 15 supplies a synchronization signal to the control unit 17 of the reception buffer unit 11, the pseudo echo generation unit 12, and the control unit 19 of the transmission buffer unit 13, and the timing at which the received voice data is read from the reception buffer unit 11, The timing at which the pseudo echo generation unit 12 generates a pseudo echo and outputs it to the subtraction unit 14 is synchronized with the timing at which the transmission voice data including the echo component EC is read from the transmission buffer unit 13. That is, the synchronization control unit 15 controls the transmission voice data read from the transmission buffer unit 13 and the pseudo echo signal corresponding to the transmission voice data to be correctly input to the subtraction unit 14. In this way, when the transmission voice data including the echo component EC read from the transmission buffer unit 13 and the pseudo echo signal generated by the pseudo echo generation unit 12 are synchronized and input to the subtraction unit 14, The subtracting unit 14 subtracts the pseudo echo signal from the transmission voice data and outputs the residual signal from which the echo component EC is removed as the transmission voice data (S25).

  When the transmission voice data from which the echo component EC is removed is output from the subtraction unit 14, the adaptive filter provided in the pseudo echo generation unit 12 is provided in the pseudo echo generation unit 12 based on the transmission voice data. The coefficient of the obtained delay line filter is updated to prepare for echo cancellation processing of the next frame (S26).

  As described above, the signal processing apparatus 1 according to this embodiment includes the synchronization control unit 15 and performs control so that the transmission voice data to be subtracted and the pseudo echo signal correspond to each other correctly. Processing is eliminated and communication quality can be improved.

  Next, the flow of signals (data) in this signal processing apparatus will be described with reference to FIG. FIG. 3 is a diagram showing the data flow in the signal processing apparatus according to this embodiment in time series.

  When reception voice data is input to the reception signal input terminal 2, the reception voice data is stored in the storage unit 16 of the reception buffer unit 11 (arrow 31 in the figure, hereinafter referred to as "31").

  When the reception buffer unit 11 receives the reception voice data, the synchronization control unit 15 supplies a synchronization signal to the control unit 17 to instruct when and how many frames the reception voice data is read from the storage unit 16 ( Synchronous processing A). The control unit 17 reads the received voice data from the storage unit 16 based on an instruction from the synchronization control unit 15. By this synchronization processing A, fluctuations in time series of received voice data are absorbed. The received voice data is converted into a voice signal through a D / A converter or the like and output to the speaker 4 (32).

  A part of the sound output from the speaker 4 is input to the microphone 5 with the echo component EC added to the original transmitted voice (echo delay). That is, a time delay from when the echo component EC is output from the speaker 4 to when it is input to the microphone 5 occurs as an echo delay. The audio signal including the echo component EC is converted into transmission audio data in units of frames via an A / D conversion unit or the like and input to the transmission buffer unit 13 (33).

  When the transmission buffer unit 13 receives the transmission voice data, the synchronization control unit 15 supplies a synchronization signal to the control unit 19 regarding when and how many frames the transmission voice data is read from the storage unit 19. An instruction is given (synchronization process B). At the same time, the synchronization control unit 15 refers to the received voice data in frame units and supplies the synchronization signal to the pseudo echo generation unit 12 to instruct when to generate the pseudo echo signal (synchronization). Process C). As a result, the transmission voice data read from the storage unit 18 and the corresponding pseudo echo signal are input to the subtraction unit 14 in synchronization with each other (35, 36).

  When receiving the synchronized transmission voice data and the corresponding pseudo echo signal, the subtracting unit 14 subtracts the echo component EC from the transmission voice data and transmits the transmission voice data to be transmitted to the far end side. Output to the transmission signal output terminal 3 (37).

  As described above, in the signal processing apparatus according to this embodiment, the synchronization control unit 15 controls the operation timing of the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13. It is possible to perform echo cancellation processing using the pseudo echo signal.

  Next, the initial operation of pseudo echo generation in this signal processing device will be described with reference to FIG. 4 and FIG. 4 and 5 are diagrams showing an initial operation of the pseudo echo generator 12 of the signal processing apparatus according to this embodiment.

  In the signal processing apparatus of this embodiment, the synchronization control unit 15 synchronizes the operations of the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13, so that the received voice data, the pseudo echo signal, and the transmission buffer data are synchronized. Are synchronized. In the first stage of the synchronization operation, either the received voice data or the transmitted voice data may be used as a reference for synchronization. FIG. 4 shows an example of synchronization based on received voice data. For example, the received voice data arrives from the far end before transmitting from the near end.

  When receiving voice data is received from the far end side, the reception buffer unit 11 stores the received voice data in the storage unit 16. The synchronization control unit 15 supplies a synchronization signal to the reception buffer unit 11, reads the received voice data at a predetermined timing, and sends it to the pseudo echo generation unit 12 (S41). The echo component EC of the audio signal output to the speaker 4 is input to the microphone 5 together with the audio to be transmitted.

  When the reception voice data is read from the reception buffer unit 11, the synchronization control unit 15 instructs the pseudo echo generation unit 12 to generate a pseudo echo signal. Since there is no transmission voice data as a sample in the initial state, the pseudo echo generation unit 12 generates a pseudo echo signal by referring only to the reception voice data and outputs the pseudo echo signal to the subtraction unit 14 (S42). Specifically, for example, a preset coefficient or a coefficient obtained from previously stored transmitted voice data is applied to the delay line filter of the pseudo echo generation unit 12 to generate a pseudo echo signal.

  When the reception voice data is read from the reception buffer unit 11, the transmission buffer unit 13 stores the transmission voice data input from the microphone 5 and converted by the A / D conversion unit or the like in the storage unit 18. The synchronization control unit 15 supplies a synchronization signal to the control unit 19 of the transmission buffer unit 13 so as to synchronize with the pseudo echo signal generation processing in the pseudo echo generation unit 12, reads out the transmission voice data, and sends it to the subtraction unit 14. Output (S43).

  When the transmission voice data and the pseudo echo signal synchronized with each other are received, the subtracting unit 14 subtracts the pseudo echo signal from the transmission voice data (S44).

  The pseudo echo generation unit 12 refers to the subtracted transmission voice data and determines whether the previously generated pseudo echo signal is appropriate (S45).

  As a result of the determination, if the pseudo echo signal is not appropriate (No in S45), the pseudo echo generation unit 12 does not generate a new pseudo echo signal and outputs the previously generated pseudo echo signal to the subtraction unit 14 again. The transmission buffer unit 13 reads the transmission voice data of the next frame (S46). As a result, the subtracting unit 14 subtracts the same pseudo echo signal from the transmission voice data of the next frame of the transmission voice data output previously (S44). This loop is repeated until the correspondence between the received voice data that generated the pseudo echo signal and the transmitted voice data corresponding thereto is optimized (S45).

  As a result of the determination, if the pseudo echo signal is appropriate (Yes in S45), that is, if the received voice data processed first and the transmission voice data to be subtracted correspond to each other in an optimal state, the pseudo echo is generated. The unit 12 determines a combination of the received voice data and the corresponding transmitted voice data, and thereafter generates a pseudo echo signal by synchronizing the received voice data and the transmitted voice data based on the combination (S47). ).

  In addition, as a method for determining whether or not the pseudo echo signal is optimal, a method for determining a correspondence relationship based on sequence number and time information attached to received voice data and transmitted voice data in frame units, transmitted voice data For example, a method of selecting a signal having the maximum level obtained by subtracting the pseudo echo signal from (a pseudo echo signal correctly reflecting an echo component) can be considered. Moreover, these determination methods are not limited to one, You may determine combining several methods.

  Next, an example in which synchronization is performed based on transmitted voice data will be described. FIG. 5 shows an example of synchronization based on transmission voice data. For example, the operations of the reception buffer unit 11, the pseudo echo generation unit 12, and the transmission buffer unit 13 are not synchronized for some reason. Is the case.

  When the voice including the echo component EC is input from the microphone 5, the transmission buffer unit 13 stores the transmission voice data including the echo component EC in the storage unit 18 via the A / D conversion unit or the like. The synchronization control unit 15 supplies a synchronization signal to the transmission buffer unit 13 and reads the transmission voice data at a predetermined timing (S51).

  On the other hand, the reception buffer unit 11 receives reception voice data and stores it in the storage unit 16. The synchronization control unit 15 supplies a synchronization signal to the reception buffer unit 11 and reads the reception voice data at a predetermined timing (S52).

  When the reception voice data is read from the reception buffer unit 11, the synchronization control unit 15 instructs the pseudo echo generation unit 12 to generate a pseudo echo signal. Here, since the state is not synchronized, the pseudo echo generation unit 12 generates a pseudo echo signal based on the previously referred transmission voice data and reception voice data and outputs the pseudo echo signal to the subtraction unit 14 (S53).

  When the read transmission voice data and the pseudo echo signal are received, the subtracting unit 14 subtracts the pseudo echo signal from the transmission voice data (S54).

  The pseudo echo generation unit 12 refers to the subtracted transmission voice data and determines whether the pseudo echo signal generated previously is appropriate (S55).

  As a result of the determination, if the pseudo echo signal is not appropriate (No in S54), the pseudo echo generation unit 12 refers to the reception voice data corresponding to the next frame from the storage unit 16 of the reception buffer unit 11 (S56), A new pseudo echo signal is generated and output to the subtracting unit 14 (S53).

  When the new pseudo echo signal is received, the subtracting unit 14 subtracts the new pseudo echo signal from the transmission voice data read in step S51 (S54).

  That is, in this loop, the pseudo echo generator 12 does not change the coefficient of the delay line filter determined by the adaptive filter, but changes the received voice data to be referred to in units of frames to generate a new pseudo echo signal. . Then, the subtracting unit 14 subtracts the new pseudo echo signal from the previously received transmission voice data. This loop is repeated until the correspondence between the previous transmitted voice data and the pseudo echo signal, in other words, the correspondence between the previous transmitted voice data and the received voice data that generated the pseudo echo signal is optimized (S55). ).

  As a result of the determination, if the pseudo echo signal is appropriate (Yes in S55), that is, if the transmitted voice data to be subtracted and the received voice data that has generated the pseudo echo signal are correctly associated and become an optimal state, The echo generation unit 12 determines a combination of the transmission voice data and the corresponding reception voice data, and generates a pseudo echo signal by synchronizing the transmission voice data and the reception voice data based on the combination thereafter. (S57).

  Note that the method for determining whether or not the pseudo echo signal is optimal corresponds to the sequence number and time information attached to the received voice data and the transmitted voice data in units of frames, as in the method shown in FIG. A method of determining the relationship, a method of selecting the one having the maximum level as a result of subtracting the pseudo echo signal from the transmission voice data (the one in which the pseudo echo signal correctly reflects the echo component), and the like can be considered. Moreover, these determination methods are not limited to one, You may determine combining several methods.

Further, the determination of whether or not the pseudo echo signal described with reference to FIGS. 4 and 5 is optimal may be performed only in the initial state of the echo cancellation process, or may be performed constantly during the execution of the echo cancellation process. Alternatively, the determination may be performed only for a predetermined number of frames, and the determination may be stopped after the determination is performed for the number of frames.
Furthermore, although it has been described that the filter coefficient is not updated in the loops of FIGS. 4 and 5, it may be performed while updating the filter coefficient. However, in this case, pseudo echo generation, subtraction, and coefficient update are looped in units of samples, and whether the pseudo echo signal is optimal is determined in units of frames, and a buffer that separately stores the filter coefficients and the subtraction result signal is required. Become. Further, in the case of No in step S45 of FIG. 4, it is necessary to start again from step S42.

  As described above, in the signal processing device of this embodiment, the pseudo echo generator 12 outputs the same pseudo echo signal until the reception voice data and the transmission voice data are synchronized, or different reception voice data until the synchronization is established. The pseudo echo signal is continuously generated based on the above, and after synchronization is established, the pseudo echo signal is generated based on the synchronized reception voice data and transmission voice data, and the output signal of the subtracting unit 14 is transmitted. Since the signal is supplied to the signal output terminal 3, an optimal pseudo echo signal can be generated even in the initial state where synchronization is not achieved, or when synchronization is lost after synchronization is achieved, and the accuracy of echo cancellation processing is improved. Can do.

  Next, a second embodiment according to the present invention will be described in detail with reference to FIG. FIG. 6 is a diagram showing a configuration of a signal processing device according to the second embodiment of the present invention. The signal processing apparatus according to this embodiment is an echo suppressor for suppressing echoes in the same IP phone as in the first embodiment. In addition, about the same component as FIG. 1, the same code | symbol is attached | subjected and shown, and the overlapping description is abbreviate | omitted.

  As shown in FIG. 6, the signal processing device 6 according to this embodiment includes a reception buffer unit 61, a transmission buffer unit 62, a voice switch unit 63, and a synchronization control unit 64. The reception signal input terminal 2 is a terminal for inputting reception voice data in frame units, and is connected to the reception buffer unit 61.

  The reception buffer unit 61 controls a storage unit 65a including a memory for storing received voice data transmitted from the far end side, an attenuation unit 66a for reducing the level of the received voice data, and reading of the stored data. Control unit 67a. The reception buffer unit 61 temporarily stores the reception voice data, thereby preventing an erroneous processing in units of frames in the echo suppression processing. The reception buffer unit 61 is connected to the speaker 4 via a voice switch unit 63, a synchronization control unit 64, a D / A conversion unit (not shown), and the like.

  The transmission buffer unit 62 includes a storage unit 65b that stores transmission voice data transmitted from the microphone 5 via an A / D conversion unit, an attenuation unit 66b that reduces the level of the transmission voice data, and the storage. And a control unit 67b for controlling reading of the read data. The transmission buffer unit 62 stores transmission voice data, thereby preventing an erroneous process in units of frames in the echo suppression process. The transmission buffer unit 62 is connected to the voice switch unit 63, the synchronization control unit 64, and the transmission signal output terminal 3.

  The voice switch unit 63 compares the reception voice data stored in the reception buffer unit 61 with the transmission voice data stored in the transmission buffer unit 62 to determine which is the echo component EC, and the echo component side It acts to determine the level attenuation for the data. The voice switch unit 63 is connected to the reception buffer unit 61, the transmission buffer unit 62, and the synchronization control unit 64.

  The synchronization control unit 64 is composed of, for example, a reference signal oscillator and controls the reading of received voice data and transmitted voice data to be processed and determines the timing of signal processing. The synchronization control unit 64 is connected to the reception buffer unit 61, the transmission buffer unit 62, and the voice switch unit 63.

  The signal processing device 6 includes a synchronization control unit 64 and supplies a synchronization signal to the reception buffer unit 61, the transmission buffer unit 62, and the voice switch unit 63. Since the received voice data and the transmitted voice data to be compared by the voice switch unit 63 are synchronized, erroneous processing can be reduced even when the echo suppression processing is realized by software.

  Next, the operation of the signal processing apparatus according to this embodiment will be described with reference to FIG. 6 and FIG. FIG. 7 is a flowchart showing the operation of the signal processing apparatus according to this embodiment. In the signal processing apparatus of this embodiment, echo suppression processing is executed for audio data in units of frames. Hereinafter, one cycle of this echo suppression process will be described.

  Also in the signal processing device according to the second embodiment, as in the signal processing device according to the first embodiment, the synchronization control unit 64 first receives the received voice data, the transmitted voice data, the attenuation processing timing, A synchronization signal serving as a reference for processing execution timing of each of the buffer unit 61, the transmission buffer unit 62, and the voice switch unit 63 is generated and supplied to the reception buffer unit 61, the transmission buffer unit 62, and the voice switch unit 63 (S71). ).

  When reception voice data in units of frames is input to the reception voice signal input terminal 2, the reception buffer unit 61 stores the reception voice data in the storage unit 65a. The stored data is read based on the synchronization signal from the synchronization control unit 64 and output to the speaker 4 through the D / A conversion unit or the like (S72).

  On the other hand, a part of the sound output from the speaker 4 is input to the microphone 5 together with the sound to be transmitted as the echo component EC. The audio signal including the echo component input to the microphone 5 is input to the transmission buffer unit 62 as transmission audio data in units of frames via an A / D conversion unit or the like. When the transmission voice data is input, the transmission buffer unit 62 stores the transmission voice data in the storage unit 65b. The stored data is read based on the synchronization signal from the synchronization control unit 64 and output to the transmission signal output terminal 3 (S73).

  Here, the voice switch unit 63 receives the received voice data stored in the storage unit 65a of the reception buffer unit 61 based on the synchronization signal from the synchronization control unit 64, and the transmission buffer unit 62 corresponding to the received voice data. With reference to the data stored in the storage unit 65b, the signal levels of the two are compared (S74). As a result of the comparison, it is determined that the audio data with the lower signal level is the non-priority side.

  When the non-priority side is determined, the voice switch unit 63 determines the attenuation amount of the attenuation unit 66a or 66b of the buffer unit to be attenuated, either the reception buffer unit 61 or the transmission buffer unit 62. When the attenuation amount is determined, the voice switch unit 63 issues a sound data level attenuation instruction to the attenuation unit 66a or 66b (S75). As a result, the signal level attenuation process is performed on either the received voice data or the transmitted voice data determined as the non-priority side.

  As described above, the signal processing device 6 of this embodiment includes the synchronization control unit 64 and synchronizes each of the received voice data and the transmitted voice data to be referred to when determining the attenuation amount in the frame unit. Echo suppression processing can be executed accurately and deterioration of sound quality can be suppressed.

In addition, this invention is not limited only to the said embodiment.
In the above embodiment, the constituent parts such as the reception buffer unit, the transmission buffer unit, the pseudo echo generation unit, the subtraction unit, and the voice switch unit are described as being implemented by software. However, the present invention is not limited to this. That is, the same effect can be obtained even if each component is configured by hardware parts or configured in a form in which hardware and software are combined using a programmable IC or the like.
In FIG. 6, it has been described that the echo suppression process is performed by attenuating the non-priority side by the attenuation unit 66a or 66b. However, an ON / OFF switch may be used instead of the attenuator.

  In addition, each process such as reading received voice data, reading transmitted voice data, and generating a pseudo echo signal in the above embodiment is not limited to the above-described flow order. That is, the reception voice data, the transmission voice data, the pseudo echo signal, the attenuation instruction synchronization signal, and the like may be executed in another order as long as they are processed in synchronization with each other.

  In the above-described embodiment, the signal processing apparatus has been described as removing or suppressing the echo component mixed into the microphone from the near-end speaker, but the present invention is not limited to this. That is, the present invention can also be applied to the removal or suppression of echoes generated in a 2-wire to 4-wire converter or the like. In this case, the side on which the echo is generated is the near end side, and the side on which the signal that is the source of the echo is transmitted is the far end side.

  Here, the software in the above embodiment may be stored in a computer-readable storage medium such as a flexible disk, or may be transmitted as software (program) alone. In this case, the software (program) stored in the storage medium is read by the computer, or downloaded and installed from a site (server) on the LAN or the Internet, thereby enabling processing in each embodiment.

  That is, the software (program) in the present invention is not limited to that stored in a storage medium independent of the computer, but includes software distributed via a transmission medium such as a LAN or the Internet.

  Further, based on an instruction of a program installed in a computer from a storage medium, an OS (operating system) running on the computer, database management software, MW (middleware) such as network software, and the like realize this embodiment. A part of each process may be executed.

  The computer executes each process in the present embodiment based on a program stored in a storage medium, and includes any one device such as a personal computer or a system in which a plurality of devices are connected to a network. It may be configured as follows.

  The computer is not limited to a personal computer, but includes an arithmetic processing device, a microcomputer, and the like included in an information processing device, and is a generic term for devices and devices that can realize the functions of the present invention by a program.

  The present invention can be applied to uses such as an electronic device manufacturing industry, a communication device manufacturing industry, and a telecommunications industry.

It is a figure which shows the structure of the signal processing apparatus of the 1st Embodiment of this invention. It is a flowchart which shows operation | movement of the signal processing apparatus of this embodiment. It is a figure which shows the flow of a signal of the signal processing apparatus of this embodiment. It is a figure explaining the initial operation of pseudo echo generation in the signal processor of this embodiment. It is a figure explaining the initial operation of pseudo echo generation in the signal processor of this embodiment. It is a figure which shows the structure of the signal processing apparatus of the 2nd Embodiment of this invention. It is a flowchart which shows operation | movement of the signal processing apparatus of this embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Signal processing apparatus, 2 ... Reception signal input terminal, 3 ... Transmission signal output terminal, 4 ... Speaker, 5 ... Microphone, 6 ... Signal processing apparatus, 11 ... Reception buffer part, 12 ... Pseudo echo generation part, 13 ... Transmission buffer unit, 14 ... Subtraction unit, 15 ... Synchronization control unit, 16 ... Storage unit, 17 ... Control unit, 18 ... Storage unit, 19 ... Control unit, 61 ... Reception buffer unit, 62 ... Transmission buffer unit, 63 ... voice switch part, 64 ... synchronization control part, 65a · b ... storage part, 66a · b ... attenuation part, 67a · b ... control part.

Claims (4)

A signal processing device in a communication device using a transmitter and a receiver independent of each other,
A synchronization control unit for supplying a synchronization signal;
First buffer means for storing a received signal from the other party in units of frames and outputting the received signal based on the synchronization signal;
Second buffer means for storing transmission signals from the transmitter including echoes output from the receiver in units of frames and outputting the transmission signals based on the synchronization signals;
A pseudo echo signal is generated based on the transmission signal transmitted to the other party and the reception signal stored in the first buffer means, and is synchronized with the transmission signal output from the second buffer means based on the synchronization signal. Pseudo echo signal generation means for outputting the pseudo echo signal;
A signal processing apparatus comprising: subtracting means for subtracting the pseudo echo signal from a transmission signal output from the second buffer means and outputting a transmission signal to be transmitted to the other party. A signal processing device in a communication device using a transmitter and a receiver independent of each other,
A synchronization control unit for supplying a synchronization signal;
A first buffer means for outputting the received signal to the receiver based on the synchronization signal, comprising: a storage unit for storing the received signal from the other party in units of frames; and an attenuation unit for attenuating the received signal;
A storage unit for storing a transmission signal from the transmitter including an echo output from the receiver in units of a frame; and an attenuation unit for attenuating the transmission signal. The transmission signal is determined based on the synchronization signal. Second buffer means for outputting;
The amount of attenuation in the first buffer means and the amount of attenuation in the second buffer means by comparing the received signal stored in the first buffer means with the transmission signal stored in the second buffer means And a voice switch unit for setting an attenuation amount of each attenuation unit of the first buffer unit and the second buffer unit based on the synchronization signal. A signal processing device in a communication device using a transmitter and a receiver independent of each other,
A synchronization control unit for supplying a synchronization signal;
First buffer means for storing a received signal from the other party in units of frames and outputting the received signal based on the synchronization signal;
Second buffer means for storing transmission signals from the transmitter including echoes output from the receiver in units of frames and outputting the transmission signals based on the synchronization signals;
A pseudo echo signal is generated based on the transmission signal transmitted to the other party and the reception signal stored in the first buffer means, and is synchronized with the transmission signal output from the second buffer means based on the synchronization signal. Pseudo echo signal generation means for outputting the pseudo echo signal;
A computer program for causing a computer to function as a signal processing apparatus comprising subtracting means for subtracting the pseudo echo signal from a transmission signal output from the second buffer means and outputting a transmission signal to be transmitted to the other party. A signal processing device in a communication device using a transmitter and a receiver independent of each other,
A synchronization control unit for supplying a synchronization signal;
A first buffer means for outputting the received signal to the receiver based on the synchronization signal, comprising: a storage unit for storing the received signal from the other party in units of frames; and an attenuation unit for attenuating the received signal;
A storage unit for storing a transmission signal from the transmitter including an echo output from the receiver in units of a frame; and an attenuation unit for attenuating the transmission signal. The transmission signal is determined based on the synchronization signal. Second buffer means for outputting;
The amount of attenuation in the first buffer means and the amount of attenuation in the second buffer means by comparing the received signal stored in the first buffer means with the transmission signal stored in the second buffer means And a computer that causes the computer to function as a signal processing device including a voice switch unit that sets an attenuation amount of each attenuation unit of the first buffer unit and the second buffer unit based on the synchronization signal program.

Images