JP5169913B2 - Data superimposing apparatus, communication system, and acoustic communication method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication system for improving reliability of data transmission without degrading voice quality of a voice signal by utilizing human auditory characteristics. <P>SOLUTION: A communication system includes a transmission device and a receiving device. The transmission device includes a data superposing device including: a voice signal input section for inputting the voice signal; a gap forming section for forming a gap section which is a time missing section in the voice signal; and a noise combining section for combining a noise signal modulated by a data code in the gap section of the voice signal, and a sounding section for sounding the voice signal in which the noise signal is combined by the data superposing device into a medium. The receiving device includes: a sound collecting section for collecting a voice signal which is transmitted through the medium; and a decoding section for decoding the data signal from the noise signal included in the voice signal collected the sound collecting section. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a data superimposing apparatus, a communication system, and an acoustic communication method capable of realizing highly reliable acoustic communication.

  As an acoustic communication technique for transmitting data using a sound wave propagating in a medium such as air, a technique has been proposed in which a data signal is spectrum-spread and converted into a spread signal to emit sound (see Patent Document 1). Since the spread signal becomes unpleasant noise for human beings, in the technique of Patent Document 1, the spread signal is mixed with an audio signal such as a musical tone signal, and the signal level of the spread signal is set to be equal to or less than the masking threshold value. I have control.

International Publication No. 02/45286 Pamphlet

  However, in the above communication method, in order to prevent the spread signal from being noticed by the audience, it is necessary to control the signal level of the spread signal to a level lower than the masking threshold value by the audio signal. However, there is a problem that the waveform of the spread signal deteriorates and it is difficult to transmit data with high reliability.

  Further, even if the signal level of the spread signal is set to be equal to or less than the masking threshold value, it is undeniable that the sound quality of the sound signal is deteriorated because the audience can hear the sound signal in which the spread signal (noise) is mixed.

  The present invention provides a data superimposing device, a communication system, and an acoustic communication method capable of utilizing the characteristics of human hearing, not deteriorating the sound quality of an audio signal, and increasing the reliability of data transmission. Objective.

The data superimposing device according to the present invention includes an audio signal input unit that inputs an audio signal, a low-pass filter that extracts a low frequency range component of the audio signal, and a gap forming unit that forms a gap section that is a temporal omission in the audio signal A noise synthesizer that synthesizes a noise signal modulated by a data code in the gap interval of the audio signal, and a low frequency range that further synthesizes the low frequency component extracted by the low-pass filter in the gap interval of the audio signal And a combining unit .

  According to a second aspect of the present invention, in the first aspect of the invention, the noise synthesizer controls the signal level of the noise signal according to the signal level of the audio signal before or after the gap section or the gap section. Features.

According to the acoustic communication method of the present invention , a gap section which is a time missing portion is formed in a voice signal, a noise signal modulated by a data code is synthesized in the gap section, and the low frequency range of the voice signal is synthesized in the gap section. The components are further combined, and the synthesized voice signal is emitted.

According to a fourth aspect of the present invention, in the above invention, the signal level of the noise signal is controlled according to the signal level of the audio signal before or after the gap section or the gap section.

The invention according to claim 5 includes the data superimposing device according to claim 1 or 2 , and a sound emitting unit that emits an audio signal synthesized with a noise signal by the data superimposing device into a medium. A receiving device comprising: a transmitting device; a sound collecting unit that collects an audio signal propagating in a medium; and a demodulating unit that demodulates a data signal from a noise signal included in the audio signal collected by the sound collecting unit It is characterized by comprising.

  The above invention uses the psychoacoustic effect of auditory guidance and synthesizes a noise signal modulated with a data code in a gap section formed in a speech signal, thereby transmitting data using sound without the listener's awareness. Is realized. The present invention is not limited to acoustic communication, and can also be applied to communication using wired / wireless transmission of analog audio signals, streaming of digital audio signals, and communication using file transfer.

  As described above, according to the present invention, when data is embedded and transmitted in an audio signal so as not to be noticed by the audience, highly reliable data transmission is possible without deteriorating the sound quality of the audio signal.

The figure which shows the example of the audio | voice signal and pseudo noise signal which are used for the acoustic communication system which is embodiment of this invention The block diagram which shows the basic composition of the synthetic | combination part which is embodiment of this invention The figure which shows the signal waveform of each part of the composition part The block diagram of the acoustic communication system which is embodiment of this invention The block diagram of the synthetic | combination part with which the transmitter of the acoustic communication system is provided The figure which shows the waveform of the synthetic signal which the synthetic | combination part outputs The block diagram of the demodulation part with which the receiver of the said acoustic communication system is provided Configuration diagram of matched filter of demodulator The figure which shows the example of the output waveform of the peak detection part of the demodulation part

≪Description of data communication system≫
First, the acoustic communication system of the present invention will be described with reference to the drawings. FIG. 1A is a diagram illustrating an example of an audio signal used in this acoustic communication method. FIG. 5B is a diagram showing an example of a pseudo noise signal (pseudo noise signal) used in the present acoustic communication system. In the present embodiment, the pseudo noise signal of FIG. 5B is modulated with the data code, and the modulated pseudo noise signal is embedded in the audio signal of FIG. To embed in the audio signal is to form a short gap (temporal omission) interval in the audio signal, and to synthesize a pseudo noise signal modulated with a data code in this gap interval.

  The audio signal in FIG. 3A is a signal including a component of a human audible frequency band. The pseudo noise signal in FIG. 5B is a 1-bit sequence signal generated based on an M-sequence (Maximal length sequence) polynomial. Note that a 1-bit sequence signal usually takes a value of 1/0, but here it is converted to a value of 1 / -1 in order to facilitate modulation and propagation by a data code.

  In the present embodiment, a direct spread spectrum modulation method in which a pseudo noise signal is phase-modulated with a data code is used as a data modulation method, but the modulation method is not limited to this. That is, the signal embedded in the gap section of the audio signal may be a noise signal including a data code. Further, the noise signal does not need to have a completely flat frequency characteristic, and may be anything that can be heard as noise (a sound whose frequency cannot be specified) by the listener.

  The embedding of the pseudo noise signal of the present invention utilizes human auditory characteristics. Here, the psychoacoustic effect which is the characteristic of human hearing utilized by the present invention will be described. When there is a gap (temporal omission) in the middle of normal sound (for example, a sine wave whose frequency changes continuously), the sound can be heard by the listener. However, if white noise is inserted into the gap section, the listener can hear the missing voice as a continuous voice with the missing voice being supplemented. This phenomenon is an auditory psychological effect called “auditoryinduction”. It has also been reported that even if some notes of a continuous melody are replaced with noise, the melody is connected and heard. Furthermore, even in the case of verbal speech, even if a part of the speech is missing, a phenomenon called “phonological restoration” is known in which the listener can listen to the missing part by complementing it from the context before and after the missing part. . In this way, even when a short interval (gap interval) of about several tens of ms is replaced with noise, human hearing perceives the audio in the gap interval by reconstructing the audio in the gap interval from that before and after that, that is, The illusion of listening to continuous speech does not perceive the existence of a gap interval in the speech signal.

  The acoustic communication system of the present invention uses this psychoacoustic effect to form a temporal gap section in the audio signal, and a signal having white noise frequency characteristics modulated by a data code in this gap section ( By embedding a pseudo-noise signal), data is transmitted using voice, and the listener is made aware that continuous voice is heard and data is transmitted.

≪Basic configuration of data superimposition part≫
A basic configuration of a data superimposing apparatus according to an embodiment of the present invention will be described with reference to the drawings. The data superimposing device is a device that forms a gap section in an audio signal and embeds a noise signal modulated by a data code in the gap section. FIG. 2 is a block diagram showing a basic configuration of the data superimposing unit, and FIG. 3 is a diagram showing signal waveforms of respective units of the data superimposing device.

  In FIG. 2, the audio signal S input from the audio signal input unit 13 is input to the window function multiplier 32. The window function multiplier 32 receives the window function W1 generated by the window function generator 31, and the window function W1 is multiplied by the audio signal S, thereby forming a temporal gap section in the audio signal S. Is done. Here, the audio signal S is a signal having a waveform as shown in FIG. 1 (A) and FIG. 3 (A) described above. The window function W1 is a signal having a waveform as shown in FIG.

  That is, the window function W1 is a signal that takes a value of “1” during a period in which the audio signal is output as usual, and “0” during a period in which a gap is formed in the audio signal. Further, the window function W1 uses a half-wave waveform of the Hanning window in the transitional section between “1” and “0”, thereby preventing noise generation due to distortion of the audio signal. In FIG. 3B, the waveform of one gap section of the window function W1 is shown. The window function W1 is a signal in which the gap section is repeated at a predetermined interval. The gap width and gap interval are arbitrary, but are set to, for example, a gap width (width) of 17 ms and a gap interval (interval) of about 1 second.

  By multiplying the audio signal S shown in FIG. 3A by the window function W1 shown in FIG. 3B, the window function multiplier 32 outputs an audio signal having a gap section as shown in FIG.

  On the other hand, the pseudo noise signal generation unit 30 generates a pseudo noise signal PN as shown in FIG. 1B and inputs the generated pseudo noise signal PN to the modulation unit 34. A data code D obtained by converting data to be transmitted into a serial bit string is input to the modulator 34 together with the pseudo noise signal PN. The modulation unit 34 modulates the pseudo noise signal PN by controlling the phase of the pseudo noise signal PN with the bit string of the data code D.

  For example, the modulation unit 34 converts both the pseudo noise signal PN and the data code D generated by the pseudo noise signal generation unit 30 into binary values of −1/1, and multiplies them. Since the pseudo noise signal PN and the data code D are both binary data of −1/1, if the data code D is “1”, the pseudo noise signal PN is output with the same phase, and the data code D is If it is “−1” (“0” as bit data), the pseudo noise signal PN is output in the opposite phase. As described above, the pseudo noise signal PN is phase-modulated at 0 ° or 180 ° according to the data code D to be superimposed.

  The modulated pseudo noise signal PNM is input to the window function multiplier 35. The window function multiplier 35 receives the window function W2 obtained by subtracting the window function W1 from 1 by subtracting the window function W1 of FIG. In other words, the window function W2 is a signal that takes a value of “1” only in the gap section when the section in which the audio signal is output as usual is “0”. By multiplying the modulated pseudo noise signal PNM by the window function W2, the window function multiplier 35 outputs a pseudo noise signal PNM having a waveform as shown in FIG.

  Both the audio signal S having the gap interval in FIG. 3C and the pseudo noise signal PNM limited by the window function W2 in FIG. 3D are input to the adder 33. The adder 33 adds and synthesizes the audio signal S (FIG. 3C) and the pseudo noise signal PNM (FIG. 3D) to form a synthesized signal as shown in FIG.

  In this way, by forming a short gap section in the audio signal S and forming a synthesized signal in which the pseudo noise signal PNM modulated with the data code is embedded in the gap section, even if this synthesized signal is emitted, The listener is unaware that due to the psychoacoustic effect, it is heard as a normal musical tone and contains noise (modulated with a data code). In addition, since the sound signal S and the pseudo noise signal PNM modulated by the data code D are divided in time, and the sound signal S is not emitted in a state where both are mixed at the same timing, the sound quality of the sound signal S is good. In addition, the waveform of the pseudo noise signal PNM does not collapse, and high-quality data transmission can be realized.

≪Description of communication system≫
Next, an acoustic communication system to which the data communication method is applied will be described. FIG. 4 is a diagram showing the configuration of the acoustic communication system. This acoustic communication system includes a transmission device 1 and a reception device 2.

  The transmission device 1 includes a data superimposing unit 10, an analog circuit unit 11, and a speaker 12. The data superimposing unit 10 is provided with a function such as gain control based on the data superimposing apparatus shown in FIG. The data superimposing unit 10 includes a digital signal processing device such as a DSP.

  The analog circuit unit 11 includes a D / A converter and an audio amplifier. The analog circuit unit 11 converts the digital composite signal output from the data superimposing unit 10 into an analog signal, amplifies the signal, and supplies the analog signal to the speaker 12. The speaker 12 emits the synthesized signal input from the analog circuit unit 11 as sound. The emitted synthesized signal sound propagates through the space and reaches the microphone 22 of the receiving device 2.

  The receiving device 2 includes a microphone 22, an analog circuit unit 23, and a demodulation unit 21. The analog circuit unit 23 includes an amplifier that amplifies the audio signal picked up by the microphone 22 and an A / D converter that converts the audio signal into a digital signal. The demodulator 21 is a circuit unit that detects the pseudo noise signal PNM included in the collected audio signal and demodulates the data code D superimposed on the pseudo noise signal PNM. Details of the configuration and operation of the demodulator 21 will be described later.

≪Description of transmitting device≫
FIG. 5 is a block diagram illustrating a configuration of the data superimposing unit 10 of the transmission device 1. The basic configuration of the data superimposing unit 10 is the same as that of the data superimposing apparatus shown in FIG. 2, and the following functions are further added to the data superimposing apparatus of FIG.
(1) Of the audio signal S, the signal component in the low range where the pseudo noise signal PN is not distributed is output even in the gap section.
(2) The level of the high frequency signal component to be suppressed in the gap section is detected, and the level of the pseudo noise signal PNM is controlled based on the detected level.
In the configuration of the data superimposing unit 10 in this figure, the same components as those in the data superimposing apparatus in FIG.

  The audio signal S input from the audio signal input unit 13 is input to the window function multiplication unit 32 and also input to the low pass filter (LPF) 42 and the high pass filter (HPF) 43. The LPF 42 is a filter that allows the low sound region component SL, in which the components of the pseudo noise signal PN generated by the pseudo noise signal generator 30 to be distributed, of the audio signal S to pass through. On the other hand, the HPF 43 is a filter that passes the high-frequency component SH in which the components of the pseudo noise signal generated by the pseudo noise signal generator 30 are distributed in the audio signal S.

  Here, the cut-off frequency of the LPF 42 and the HPF 43 may be determined by the chip rate of the spreading code to be used and the cyclic period. For example, in the case of an M-sequence spreading code having a cyclic period of 511 bits, the same bit does not continue beyond nine. Therefore, the longest period of the pseudo noise signal formed by the M-sequence spread code string having a cyclic period of 511 bits is 9 bits (upper half wave) +9 bits (lower half wave) = 18 bits. Therefore, when a pseudo-noise signal having a chip rate of 44.1 kHz is formed with this M-sequence spreading code string, the fundamental frequency becomes 2.45 kHz (= 44.1 kHz ÷ 18), and the pseudo-noise signal is higher than the fundamental frequency. Distributed. This fundamental tone frequency is adopted as the cutoff frequency of the LPF 42 and HPF 43.

  Of the audio signal S that has passed through the LPF 42, the low-frequency signal component SL is multiplied by the same window function W2 that is multiplied by the pseudo noise signal PN by the window function multiplier 45. The signal component SL in the low frequency range that is limited in time only to the gap section by the window function W2 is input to the gain multiplier 47. The gain multiplication unit 47 multiplies the low-frequency signal component SL by the gain GainL and outputs the result to the adder 33. Thereby, even in the gap section in which the audio signal S is cut off, only the low-frequency component SL in the audio signal S is output, and the sense of incongruity in the gap section can be reduced.

  The high frequency range component SH that has passed through the HPF 43 is input to the level detection unit 44. The level detection unit 44 detects the signal level HL of the input high range component SH. The detected signal level HL is multiplied by the gain GainH in the gain multiplier 48 and then input to the multiplier 50.

  As described with reference to FIG. 2, the pseudo noise signal generation unit 30 generates the pseudo noise signal PN and inputs it to the modulation unit 34. A data code D is further input to the modulation unit 34. The data code D is appropriately read by the data code reading unit 40 and input to the modulation unit 34. Reading of the data code D by the data code reading unit 40 is controlled by the gap timing control unit 41.

  The gap timing control unit 41 controls the width and interval of the gap section formed in the audio signal S. The gap timing control unit 41 instructs the window function generating unit 31 and the data code reading unit 40 to start timing and end timing of the gap section. The window function generator 31 generates and outputs window functions W1 and W2 having a window with a width of the start timing and end timing. In addition, the data code reading unit 40 reads the data code D and inputs it to the modulation unit 34 only during the start timing and the end timing.

  The modulation unit 34 modulates the pseudo noise signal PN by controlling the phase of the pseudo noise signal PN with the bit string of the data code D. The modulated pseudo noise signal PNM is multiplied by the window function W2 in the window function multiplier 35. Thereby, the pseudo noise signal PNM is limited only to the gap section of the audio signal S. The pseudo noise signal PNM limited to the gap section only by the window function W2 is multiplied by the gain GainN in the gain multiplier 49, and further multiplied by the level HL of the high-frequency component SH of the audio signal S in the multiplier 50.

  By multiplying the pseudo noise signal PNM by the level HL of the high-frequency component SH of the audio signal S, the signal level of the pseudo-noise signal PNM is controlled according to the signal level of the high-frequency component SH of the audio signal S. Thus, the volume change between the section in which the audio signal S is output and the gap section is suppressed, and the sense of discomfort in hearing can be reduced.

  The control of the signal level of the pseudo noise signal PNM is not limited to the high-frequency component SH, but may be based on the volume level of the entire band of the audio signal S. Further, the section that refers to the signal level is not limited to the gap section, and may be a section immediately before or after the gap section.

  The pseudo noise signal PNM multiplied by the gain GainN and the level HL is input to the adder 33.

  The adder 33 embeds the modulated pseudo noise signal PNM and the low-frequency component SL of the audio signal in the gap interval of the audio signal S (see FIG. 3C) having the gap interval, as shown in FIG. A simple composite signal. In the synthesized signal shown in FIG. 6, the level of the pseudo noise signal PNM is controlled by multiplying the pseudo noise signal PNM by the level HL of the high frequency range component SH of the audio signal, so that the level of the pseudo noise signal PNM becomes the level of the high frequency range component SH of the audio signal. It fluctuates together, and the sense of incongruity on hearing is reduced. In addition, since the low-frequency component SL of the audio signal is output even in the gap section, the pseudo-noise signal PNM is biased with the low-frequency component SL, which reduces the sense of discomfort in hearing. Note that this low-frequency component SL can be easily removed by using HPF in the receiving device.

  For the cap width, the gap interval, and Gain L, Gain H, and Gain N multiplied by the gain multipliers 47, 48, and 49, appropriate values in a range that does not make the listener feel uncomfortable may be determined by auditory evaluation.

  Further, when a human voice is used as the audio signal S, the frequency component is concentrated in the low frequency range, and the output level of the high frequency range component SH of the HPF 43 may become almost zero. In such a case, the level detection unit 44 may fix the output value to a predetermined value (for example, “1”) and not perform the control based on the level of the high-frequency component SH.

<< Description of receiver >>
FIG. 7 is a diagram illustrating a detailed configuration of the demodulation unit 21 provided in the receiving device 2. The demodulating unit 21 receives a composite signal collected by the microphone 22 and converted into a digital signal by the analog circuit unit 23. The synthesized signal is a signal in which the pseudo noise signal PNM modulated in the audio signal S is embedded as described above. The demodulator 21 extracts the pseudo noise signal PNM from the combined signal, obtains a correlation value (peak value) with the original pseudo noise signal (code string) PN as a reference signal, and codes the peak value of the modulation signal PNM. The data code D is demodulated based on whether (positive / negative) matches or reverses the sign (positive / negative) of the reference signal PN.

  The demodulator 21 includes a high-pass filter (HPF) 61, a waveform shaping unit 62, a matched filter 63, a peak detection unit 64, and a code determination unit 65. Hereinafter, the configuration and function of each functional unit will be described.

  The high-pass filter (HPF) 61 is a functional unit that extracts a high-frequency component including the pseudo noise signal PNM from the received synthesized signal. The cut-off frequency of the HPF 61 may be set to be the same as the cut-off frequency (2.45 kHz) of the HPF 43 used in the modulation unit 10 of the transmission device 1.

  The sound signal of the high frequency range component extracted by the HPF 61 is converted into a binary signal by the waveform shaping unit 62. That is, the digitized audio signal is, for example, a signal quantized to 8 bits (256 steps) and 16 bits (65536 steps), but the pseudo noise signal PNM to be demodulated is 1 / -1. This is a binary signal. In order to approximate the waveform to the pseudo noise signal PNM, the high frequency component signal extracted by the HPF 61 is converted into a binary signal using a comparator. This binarized digital audio signal is input to the matched filter 63.

  Note that, as indicated by the dotted line in FIG. 7, the waveform shaping unit 62 may be omitted, and the high-frequency component signal that has passed through the HPF 61 may be directly input to the matched filter 63.

  The matched filter 63 is a filter that detects a correlation value between the input digital audio signal and the reference signal, and is configured by an FIR filter. FIG. 8 shows a configuration example of the matched filter 63. The matched filter 63 is composed of 511 stages of FIR filters, and the same code string as the pseudo noise signal PN generated by the pseudo noise signal generator 30 on the transmission side is set as the filter coefficient (reference signal) of each stage. ing. The pseudo-noise code string is a 1/0 bit string, but the filter coefficient of the matched filter 63 is set to 1 / -1 as with the pseudo-noise signal PN. Thereby, the component of the pseudo noise signal PNM included in the input digital audio signal is detected.

  The matched filter 63 outputs a correlation value between the pseudo noise signal PNM and the reference signal PN included in the input digital audio signal, and has a large correlation value (peak) when the phase of the pseudo noise signal PNM is synchronized with the reference signal. Value). Since the modulation signal PNM is phase-modulated by the data code, a positive peak value is output when the data code D is “1”, and a negative peak value is output when the data code D is “0”.

  FIG. 9A is a diagram illustrating an example of an output waveform of the matched filter 63, and is a diagram illustrating an example of a waveform having a positive peak value. When the output signal of the HPF 61 is directly input to the matched filter 63 without passing through the waveform shaping unit 62, the output waveform of the matched filter 63 has a shape as shown in FIG.

  The peak detector 64 detects a peak from the correlation value output from the matched filter 63. The code determination unit 47 determines the data code based on the peak value detected by the peak detection unit 64, demodulates and outputs the data code D. Thereby, it is possible to demodulate the data code D included in the combined signal emitted from the transmission device 1.

≪Read more≫
Although the said embodiment demonstrated the system which emits sound in the air and performs acoustic communication, the medium which propagates a sound is not limited to air. For example, the present invention can be applied to acoustic communication that propagates a solid or liquid. The present invention is not limited to acoustic communication, and can also be applied to wired communication / wireless communication in which an audio signal is propagated electrically or electromagnetically as an electric signal. Further, the present invention can be applied to the case where an audio signal is converted into a digital audio signal for streaming and file transfer.

  In the above embodiment, a pseudo noise signal in the audible frequency band (sampling rate 44.1 kHz) is used, but a pseudo noise signal in a higher frequency band (ultrasonic region) may be used.

  In the above-described embodiment, the direct spread spectrum method for phase-modulating the M-sequence pseudo noise signal is used as the data code modulation method, but the modulation method is not limited to this. For example, the OFDM method may be used.

  The width and interval of the gap section formed in the audio signal are arbitrary. The gap width and gap interval may be appropriately determined based on the content of the audio signal, or a gap section may be inserted by searching for an appropriate location in the audio signal.

DESCRIPTION OF SYMBOLS 1 Transmission apparatus 2 Reception apparatus 10 Data superimposition part 21 Demodulation part

Claims (5)

An audio signal input unit for inputting an audio signal;
A low pass filter for extracting a low frequency component of the audio signal;
A gap forming part for forming a gap section which is a time missing part in the audio signal;
A noise synthesizing unit that synthesizes a noise signal modulated by a data code in a gap section of the audio signal;
A low frequency range synthesizer that further synthesizes a low frequency range component extracted by the low-pass filter in the gap section of the audio signal;
A data superimposing apparatus.   The data superimposing apparatus according to claim 1, wherein the noise synthesis unit controls a signal level of the noise signal in accordance with a signal level of the audio signal before or after the gap section. A gap section, which is a time missing portion, is formed in the audio signal, and a noise signal modulated by the data code is synthesized in the gap section. Further, a low frequency component of the audio signal is further synthesized in the gap section, and the synthesized signal is synthesized. An acoustic communication method characterized by emitting a sound signal .   The acoustic communication method according to claim 3, wherein the signal level of the noise signal is controlled according to the signal level of the audio signal before or after the gap section or the gap section. A transmission device comprising: the data superimposing device according to claim 1 or claim 2 ; and a sound emitting unit that emits an audio signal synthesized with a noise signal by the data superimposing device into a medium;
A receiving device comprising: a sound collection unit that collects an audio signal propagating in the medium; and a demodulation unit that demodulates a data signal from a noise signal included in the audio signal collected by the sound collection unit;
A communication system comprising:

Images