Transmission device, reception device, communication system, transmission method, and reception method

Abstract

A transmission device (2) shifts the frequency of a voice signal from a first band to a second band higher than the first band by using a lapped orthogonal transform using a segmenting window function that satisfies a Princen-Bradley condition and an inverse transform of the lapped orthogonal transform, and transmits a transmission signal obtained by shifting the frequency of the voice signal to the second band. A reception device (3) receives the transmission signal transmitted by the transmission device (2), shifts the frequency of the transmission signal from the second band to the first band by using the lapped orthogonal transform using the segmenting window function and the inverse transform of the lapped orthogonal transform, and outputs an output signal obtained by shifting the frequency of the transmission signal to the first band.

Description

Transmitting device, receiving device, communication system, transmitting method, and receiving method

[0001] The present invention relates to a transmitting device, a receiving device, a communication system, a transmitting method, and a receiving method.

[0002] In recent years, technologies have been developed to selectively provide sound information to places and people for the purposes of effective guidance and direction in public places, audio guidance for visually impaired people, etc. For example, a method has been proposed in which audible sound is converted into ultrasonic waves, transmitted, picked up by a microphone, and then converted back into audible sound by a computer (see, for example, Non-Patent Document 1).

[0003] In the communication system described in Non-Patent Document 1, a transmitting device performs an FFT (Fast Fourier Transform) on an audio signal to obtain frequency data, passes the resultant frequency data through a low-pass filter with a cutoff frequency of 20 kHz, and shifts the frequency by 20 kHz to the high-frequency side, then performs an IFFT (Inverse Fast Fourier Transform) to obtain ultrasound, and transmits the obtained ultrasound. A receiving device performs an FFT on ultrasound received by a microphone to obtain frequency data, passes the resultant frequency data through a high-pass filter with a cutoff frequency of 20 kHz, and shifts the frequency by 20 kHz to the low-frequency side, then performs an IFFT to obtain an output signal in the audible band, and outputs the obtained output signal to a user.

[0004] Yoshino, Yoshiyuki et al., "Basic study on the expansion of real sound waves using ultrasound," FIT2006 (5th Forum on Information Science and Technology), pp. 287-288, 2006

[0005] In the conventional communication system, the waveform of the output signal output to the user by the receiving device may be significantly distorted compared to the waveform of the original audio signal. When such a large waveform distortion occurs, the information contained in the original audio signal cannot be properly conveyed to the user of the receiving device.

[0006] An object of the present invention is to provide a communication system in which the waveform of an output signal output from a receiving device is less likely to be distorted compared to the waveform of the original audio signal, a transmitting device and transmitting method that can be used in the communication system, and a receiving device and receiving method.

[0007] A transmission device according to a first aspect of the present invention includes: a frequency upconversion unit that shifts the frequency of an audio signal from a first band to a second band that is higher than the first band, using a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; and a transmission unit that transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band using the frequency upconversion unit.

[0008] In the transmitting device according to the first aspect, the frequency up-conversion unit may convert a signal extracted from the audio signal using the extraction window function into a first coefficient sequence in the frequency domain, obtain a second coefficient sequence from the first coefficient sequence by performing a frequency up-conversion in which a coefficient sequence subset for the first band of the first coefficient sequence is transferred to a coefficient sequence subset for the second band, and convert the obtained second coefficient sequence back into a time domain signal by the inverse conversion, thereby shifting the frequency of the audio signal from the first band to the second band.

[0009] In the transmitting device according to the first aspect, the frequency upconversion unit may: transform the signal extracted from the audio signal by the extraction window function into the first coefficient sequence using a discrete Fourier transform or a discrete cosine transform; obtain the second coefficient sequence from the first coefficient sequence using a transform equation that represents a relationship between coefficients constituting the first coefficient sequence and coefficients constituting the second coefficient sequence, the transform equation being selected depending on the number of shifts on the frequency axis of the coefficients constituting the first coefficient sequence and the coefficient numbers in the first coefficient sequence for achieving the frequency upconversion; and restore the obtained second coefficient sequence to the time domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform.

[0010] In the transmitting device according to the first aspect, the frequency upconversion unit may perform phase inversion on the extracted signals before and after the shift number when the shift number represents an odd shift. In this case, it is preferable that the frequency upconversion unit determines whether the shift number represents an even shift or an odd shift, and performs phase inversion on the extracted signals before and after the shift number when the shift number represents an odd shift, and does not perform the phase inversion when the shift number represents an even shift.

[0011] a transmitting device according to a second aspect of the present invention, comprising: a frequency upconversion unit that uses a lapped orthogonal transform using an extraction window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform to shift the frequency of an audio signal from a first band to a second band that is higher than the first band; and a transmitting unit that transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band by the frequency upconversion unit, wherein the frequency upconversion unit: transforms the signal extracted from the audio signal by the extraction window function into a first coefficient sequence in the frequency domain using a discrete Fourier transform or a discrete cosine transform; and obtains the second coefficient sequence from the first coefficient sequence using a transform equation that represents a relationship between coefficients constituting the first coefficient sequence and coefficients constituting a second coefficient sequence, the transform equation being used depending on whether an even or odd shift number of the coefficients constituting the first coefficient sequence on the frequency axis and the number of the coefficients in the first coefficient sequence, in order to realize a frequency upconversion that transfers a coefficient sequence subset of the first coefficient sequence for the first band to a coefficient sequence subset for the second band; The obtained second coefficient sequence is converted back into a time domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform, it is determined whether the number of shifts represents an even shift or an odd shift, and if the number of shifts represents an even shift, no phase inversion is performed, and if the number of shifts represents an odd shift, phase inversion is performed on the extracted previous and next time domain signals, thereby shifting the frequency of the audio signal from the first band to the second band.

[0012] A transmitting device according to a third aspect of the present invention comprises: a frequency up-conversion unit that shifts the frequency of an audio signal from a first band to a second band higher than the first band by using a lapped orthogonal transform using an extraction window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform, the frequency up-conversion unit converting a signal extracted from the audio signal by the extraction window function into a first coefficient sequence in the frequency domain, obtaining a second coefficient sequence from the first coefficient sequence by frequency up-conversion that transfers a coefficient sequence subset of the first coefficient sequence for the first band to a coefficient sequence subset of the second band, and converting the obtained second coefficient sequence back into a time domain signal by the inverse transform, thereby shifting the frequency of the audio signal from the first band to the second band; and a transmitting unit that transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band by the frequency up-conversion unit, wherein the frequency up-conversion unit determines whether the number of shifts on the frequency axis of coefficients constituting the first coefficient sequence to achieve the frequency up-conversion represents an even shift or an odd shift, When obtaining the transmission signal, if the number of shifts represents an odd number of shifts, phase inversion is performed on the extracted signals before and after the shift, and if the number of shifts represents an even number of shifts, phase inversion is not performed.

[0013] In the transmitting device according to any one of the first to third aspects, the frequency upconversion unit may obtain the transmission signal by multiplying the time-domain signal restored from the second coefficient sequence by a reconstruction window function that satisfies the Princen-Bradley condition, and connecting the signals multiplied by the reconstruction window function with overlapping.

[0014] In the transmitting device according to any one of the first to third aspects, the frequency up conversion unit may perform a permutation operation on the first coefficient sequence, using the first coefficient sequence or a coefficient sequence obtained by performing the frequency up conversion on the first coefficient sequence as a permutation target coefficient sequence, to permutate the order of coefficients constituting the permutation target coefficient sequence on the frequency axis, and the result of performing the permutation operation and the frequency up conversion on the first coefficient sequence may be the second coefficient sequence.

[0015] In the transmitting device according to any one of the first to third aspects, the frequency upconversion unit may perform a high resolution process during the frequency upconversion to increase the number of data points of the second coefficient sequence to be greater than the number of data points of the first coefficient sequence, thereby obtaining the transmission signal having a higher temporal resolution than the audio signal.

[0016] In the transmitting device according to any one of the first to third aspects, the transmitting device may further include an acquisition unit that acquires the audio signal, and the transmission signal formed by the frequency upconversion unit may be transmitted by the transmitting unit while the audio signal is acquired by the acquisition unit.

[0017] A receiving device according to a fourth aspect of the present invention includes: a receiving unit that receives a transmission signal whose frequency belongs to a second band; a frequency down-conversion unit that shifts the frequency of the transmission signal from the second band to a first band that is lower than the second band by using a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; and an output unit that outputs an output signal obtained by shifting the frequency of the transmission signal to the first band by the frequency down-conversion unit.

[0018] In the receiving device according to the fourth aspect, the frequency down-converter may convert a signal extracted from the transmission signal by the extraction window function into a third coefficient sequence in the frequency domain, obtain a fourth coefficient sequence from the third coefficient sequence by frequency down-conversion that transfers a coefficient sequence subset of the third coefficient sequence for the second band to a coefficient sequence subset of the first band, and convert the obtained fourth coefficient sequence back into a time domain signal by the inverse transform, thereby shifting the frequency of the transmission signal from the second band to the first band.

[0019] In the receiving device according to the fourth aspect, the frequency down-conversion unit may: transform the signal extracted from the transmission signal by the extraction window function into the third coefficient sequence using a discrete Fourier transform or a discrete cosine transform; obtain the fourth coefficient sequence from the third coefficient sequence using a transform equation that represents a relationship between coefficients constituting the third coefficient sequence and coefficients constituting the fourth coefficient sequence, the transform equation being selected depending on the number of shifts of the coefficients constituting the third coefficient sequence on the frequency axis for achieving the frequency down-conversion and the index of the coefficient in the third coefficient sequence; and restore the obtained fourth coefficient sequence to the time domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform.

[0020] In the receiving device according to the fourth aspect, the frequency down-converter may perform phase inversion on the extracted signals before and after the shift number when the shift number represents an odd shift. In this case, it is preferable that the frequency down-converter determines whether the shift number represents an even shift or an odd shift, and performs phase inversion on the extracted signals before and after the shift number when the shift number represents an odd shift, and does not perform the phase inversion when the shift number represents an even shift.

[0021] A receiving device according to a fifth aspect of the present invention comprises: a receiving unit that receives a transmission signal whose frequency belongs to a second band; a frequency down-conversion unit that shifts the frequency of the transmission signal from the second band to a first band that is lower than the second band by using a lapped orthogonal transform that uses an extraction window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; and an output unit that outputs an output signal obtained by shifting the frequency of the transmission signal to the first band by the frequency down-conversion unit, wherein the frequency down-conversion unit transforms the signal extracted from the transmission signal by the extraction window function into a third coefficient sequence in the frequency domain by using a discrete Fourier transform or a discrete cosine transform, the fourth coefficient sequence is obtained from the third coefficient sequence using a transformation equation that represents a relationship between the coefficients constituting the third coefficient sequence and the coefficients constituting the fourth coefficient sequence, the transformation equation being used depending on an even or odd number of shifts of the coefficients constituting the third coefficient sequence on the frequency axis and the number of the coefficient in the third coefficient sequence, in order to realize a frequency down-conversion that transfers a coefficient sequence subset of the third coefficient sequence for the second band to a coefficient sequence subset for the first band; the fourth coefficient sequence obtained is converted back into a time-domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform; it is determined whether the number of shifts represents an even shift or an odd shift; and if the number of shifts represents an even shift, no phase inversion is performed, and if the number of shifts represents an odd shift, phase inversion is performed on the extracted previous and next time-domain signals, thereby shifting the frequency of the transmission signal from the second band to the first band.

[0022] A receiving device according to a sixth aspect of the present invention comprises: a receiving unit that receives a transmission signal whose frequency belongs to a second band; and a frequency down-conversion unit that shifts the frequency of the transmission signal from the second band to a first band lower than the second band by using a lapped orthogonal transform that uses an extraction window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform, the frequency down-conversion unit converting a signal extracted from the transmission signal by the extraction window function into a third coefficient sequence in the frequency domain, obtaining a fourth coefficient sequence from the third coefficient sequence by frequency down-conversion that transfers a coefficient sequence subset of the third coefficient sequence for the second band to a coefficient sequence subset of the first band, and converting the obtained fourth coefficient sequence back into a time domain signal by the inverse transform, thereby shifting the frequency of the transmission signal from the second band to the first band; and an output unit that outputs an output signal obtained by shifting the frequency of the transmission signal to the first band by the frequency down-conversion unit, It is determined whether the number of shifts on the frequency axis of the coefficients constituting the third coefficient sequence for realizing the frequency down-conversion represents an even number shift or an odd number shift, and when obtaining the output signal, if the number of shifts represents an odd number shift, phase inversion is performed on the extracted signals before and after the signal, and if the number of shifts represents an even number shift, phase inversion is not performed.

[0023] In the receiving device according to any one of the fourth to sixth aspects, the frequency down-conversion unit may multiply the time-domain signal returned from the fourth coefficient sequence by a reconstruction window function that satisfies the Princen-Bradley condition, and combine the signals multiplied by the reconstruction window function with overlapping to obtain the output signal.

[0024] In the receiving device according to any one of the fourth to sixth aspects, the transmission signal is obtained by performing a process of performing a permutation operation on a permutation target coefficient sequence in the frequency domain, in which the order of coefficients constituting the permutation target coefficient sequence on the frequency axis is permuted; the frequency down-conversion unit may perform an inverse permutation operation on the inverse permutation target coefficient sequence, which represents an inverse mapping of the permutation operation, using a coefficient sequence obtained by performing the frequency down-conversion on the third coefficient sequence or the third coefficient sequence as an inverse permutation target coefficient sequence; and the result of performing the inverse permutation operation and the frequency down-conversion on the third coefficient sequence may be the fourth coefficient sequence.

[0025] In the receiving device according to any one of the fourth to sixth aspects, the output signal formed by the frequency down-conversion unit may be output by the output unit while the transmission signal is received by the receiving unit.

[0026] In the receiving device according to any one of the fourth to sixth aspects, the power of the output signal output by the output unit may be increased as the power of the transmission signal received by the receiving unit increases.

[0027] In the receiving device according to any one of the fourth to sixth aspects, the receiving unit may have a left half body partial receiving unit worn on the left half of the body of the user, and a right half body partial receiving unit worn on the right half of the body of the user; the frequency down conversion unit forms a left ear output signal, which is the output signal obtained by shifting the frequency of the transmission signal received by the left half body partial receiving unit to the first band, and a right ear output signal, which is the output signal obtained by shifting the frequency of the transmission signal received by the right half body partial receiving unit to the first band; and the output unit may have a left ear partial output unit that outputs the left ear output signal to the left ear of the user, and a right ear partial output unit that outputs the right ear output signal to the right ear of the user.

[0028] In the receiving device according to any one of the fourth to sixth aspects, the receiving unit may have a plurality of partial receiving units, each having the function of receiving the transmission signal and being located at different parts of the user's body, and the frequency down-conversion unit may have the function of switching which of the plurality of partial receiving units uses the transmission signal received by to form the output signal to be output to the output unit.

[0029] The plurality of partial receiving units may include a partial receiving unit for a proximity area that is attached to the distal end of the user's body, and a partial receiving unit for an approach area that is attached to a position on the user's body closer to the head than the distal end.

[0030] A communication system according to a seventh aspect of the present invention comprises: a transmitting device that uses a lapped orthogonal transform using a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform to shift the frequency of an audio signal from a first band to a second band different from the first band, and transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band; and a receiving device that receives the transmission signal transmitted by the transmitting device, and outputs an output signal obtained by using the lapped orthogonal transform using the clipping window function and the inverse transform of the lapped orthogonal transform to return the frequency of the transmission signal from the second band to the first band, and returning the frequency of the transmission signal to the first band. The transmitting device of the communication system according to the seventh aspect may be the transmitting device according to the second or third aspect. Furthermore, the receiving device of the communication system according to the seventh aspect may be the receiving device according to the fifth or sixth aspect.

[0031] In a transmission method according to an eighth aspect of the present invention, a lapped orthogonal transform using a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform are used to shift the frequency of an audio signal from a first band to a second band that is higher than the first band, and a transmission signal obtained by shifting the frequency of the audio signal to the second band is transmitted.

[0032] In the transmission method according to the eighth aspect, it is preferable that the signal extracted from the audio signal by the extraction window function is transformed into a first coefficient sequence in the frequency domain using a discrete Fourier transform or a discrete cosine transform, the second coefficient sequence is obtained from the first coefficient sequence using a transform equation that represents a relationship between coefficients constituting the first coefficient sequence and coefficients constituting a second coefficient sequence, the transform equation being used depending on an even or odd number of shifts of the coefficients constituting the first coefficient sequence on the frequency axis and the coefficient numbers in the first coefficient sequence, in order to realize a frequency up-transformation that transfers a coefficient sequence subset for the first band in the first coefficient sequence to a coefficient sequence subset for the second band, the second coefficient sequence is converted back into a time-domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform, and it is determined whether the number of shifts represents an even shift or an odd shift, and if the number of shifts represents an even shift, no phase inversion is performed, and if the shift number represents an odd shift, phase inversion is performed on the extracted time-domain signals before and after the extraction, thereby shifting the frequency of the audio signal from the first band to the second band.

[0033] A receiving method according to a ninth aspect of the present invention includes receiving a transmission signal whose frequency belongs to a second band, shifting the frequency of the transmission signal from the second band to a first band lower than the second band using a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform, and outputting an output signal obtained by shifting the frequency of the transmission signal to the first band.

[0034] In the receiving method according to the ninth aspect, it is preferable that the signal extracted from the transmission signal by the extraction window function is transformed into a third coefficient sequence in the frequency domain using a discrete Fourier transform or a discrete cosine transform, the fourth coefficient sequence is obtained from the third coefficient sequence using a transform equation that represents a relationship between coefficients constituting the third coefficient sequence and coefficients constituting a fourth coefficient sequence, the transform equation being used depending on an even or odd number of shifts of the coefficients constituting the third coefficient sequence on the frequency axis and the number of the coefficient in the third coefficient sequence, in order to realize a frequency down-transform that transfers a coefficient sequence subset for the second band of the third coefficient sequence to a coefficient sequence subset for the first band, the fourth coefficient sequence is obtained from the third coefficient sequence, the fourth coefficient sequence is converted back into a time-domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform, it is determined whether the number of shifts represents an even shift or an odd shift, and the frequency of the transmission signal is shifted from the second band to the first band by not performing phase inversion when the number of shifts represents an even shift, and by performing phase inversion on the time-domain signals before and after the extracted signal when the number of shifts represents an odd shift.

[0035] According to the communication system of the seventh aspect of the present invention, the waveform of the output signal output from the receiving device is less likely to be distorted compared to the waveform of the original audio signal.

[0036] 4A and 4B. A diagram illustrating the configuration of a communication system according to an embodiment. A flowchart illustrating an example of the operation of a transmitting device according to an embodiment. A flowchart illustrating another example of the operation of a transmitting device according to an embodiment. A flowchart specifically illustrating the content of the processing in step S2e of FIGS. 2A and 2B. A flowchart illustrating an example of the operation of a receiving device according to an embodiment. A flowchart illustrating another example of the operation of a receiving device according to an embodiment. A flowchart specifically illustrating the content of the processing in step S12e of FIGS. 4A and 4B. A diagram illustrating an example of a Kaiser-Bessel derived window. A diagram illustrating an example of frequency shift when DFT is adopted as the orthogonal transform (when N = 4 and M = +1, +2). A diagram illustrating another example of frequency shift when DFT is adopted as the orthogonal transform (when N = 4 and M = -1, -2). A diagram illustrating an example of frequency shift when DCT is adopted as the orthogonal transform (when N = 4 and M = +1, +2). A diagram illustrating another example of frequency shift when DCT is adopted as the orthogonal transform (when N = 4 and M = -1, -2). A diagram illustrating an example of a waveform extracted using an extraction window function when an even shift is performed. 1 is a diagram showing an example of a waveform cut out using a cut-out window function when an odd shift is performed. 2 is a diagram showing another example of a waveform cut out using a cut-out window function when an even shift is performed. 3 is a diagram showing another example of a waveform cut out using a cut-out window function when an odd shift is performed. 4 is a diagram showing an example of a reconstructed waveform when an even shift is performed. 5 is a diagram showing an example of a reconstructed waveform when an odd shift is performed. 6 is a diagram showing example spectrograms of an original audio signal and a transmission signal. 7 is a diagram showing an example of a frequency shift when performing multi-channel communication. 8 is a conceptual diagram showing a pattern of high resolution processing. 9 is a diagram showing an example of a configuration of a receiving device according to a modified example. 10 is a diagram showing an example of a configuration of a receiving device according to another modified example.

[0037] Hereinafter, embodiments will be described with reference to the drawings.

[0038] 1, a communication system 1 according to this embodiment includes a transmitting device 2 and a receiving device 3. The transmitting device 2 transmits a transmission signal to the outside. The transmission signal is transmitted in the form of a wave that propagates through air. The receiving device 3 receives the transmission signal.

[0039] 1 illustrates only one receiving device 3, the communication system 1 may include multiple receiving devices 3 that receive a common transmission signal. That is, the transmitting device 2 can function as a broadcasting station that broadcasts a transmission signal to the multiple receiving devices 3.

[0040] The transmitting device 2 includes an acquiring unit 21 , a frequency upconverting unit 22 , and a transmitting unit 23 .

[0041] The acquisition unit 21 acquires an audio signal representing information to be transmitted to a user of the receiving device 3. The audio signal may be an analog signal or a digital signal. The acquisition unit 21 may be a microphone that generates an audio signal, or may acquire a recorded audio signal.

[0042] The frequency upconversion unit 22 performs frequency upconversion to increase the frequency of the audio signal acquired by the acquisition unit 21, thereby forming the transmission signal described above. Specifically, the frequency upconversion unit 22 shifts the frequency of the audio signal from a first band to a second band that is higher than the first band. In this embodiment, the first band refers to the audible band of 20 kHz or less, and the second band refers to the ultrasonic band above 20 kHz. In other words, the transmission signal obtained by performing frequency upconversion on the audio signal is transmitted as ultrasonic waves.

[0043] The transmitting unit 23 outputs a transmission signal, which is an audio signal converted into ultrasonic waves. The transmitting unit 23 is, for example, a speaker capable of reproducing ultrasonic waves.

[0044] The transmitter 2 also includes a storage unit 25 that stores a frequency up conversion program 26 that defines the procedure for the above-described frequency up conversion, and a processor 24 that executes the frequency up conversion program 26. The processor 24 executes the frequency up conversion program 26 to realize the function of the above-described frequency up conversion unit 22. The frequency up conversion program 26 also includes mathematical formulas, predetermined values, etc. used in the frequency up conversion.

[0045] The receiving device 3 includes a receiving unit 31 , a frequency down-conversion unit 32 , and an output unit 33 .

[0046] The receiving unit 31 receives the transmission signal as an ultrasonic wave transmitted by the transmitting unit 23 and propagating through the air. The receiving unit 31 is configured, for example, by a microphone capable of collecting ultrasonic waves. As described above, the transmission signal is obtained by shifting the frequency of the audio signal from a first band, which is the audible band, to a second band, which is the ultrasonic band.

[0047] The frequency down converter 32 performs frequency down conversion to lower the frequency of the transmission signal received by the receiver 31. Specifically, the frequency down converter 32 converts the frequency of the transmission signal from the second band, which is the ultrasonic band, back to the first band, which is the audible band.

[0048] The output unit 33 outputs the transmission signal, which has been restored to the audible band, as an output signal. The output unit 33 is, for example, a speaker, earphones, or headphones. Bone conduction earphones may be used as the earphones. Bone conduction headphones may be used as the headphones. A bone conduction speaker may be used as the speaker.

[0049] The receiving device 3 also includes a storage unit 35 that stores a frequency down conversion program 36 that defines the procedure for the above-mentioned frequency down conversion, and a processor 34 that executes the frequency down conversion program 36. The processor 34 executes the frequency down conversion program 36 to realize the function of the above-mentioned frequency down conversion unit 32. The frequency down conversion program 36 also includes mathematical formulas, predetermined values, etc. used in the frequency down conversion.

[0050] An example of the operation of the transmitting device 2 will be specifically described with reference to FIG. 2A.

[0051] (Step S1) The acquisition unit 21 acquires an audio signal. The audio signal carries information to be transmitted to the user of the receiving device 3.

[0052] (Step S2) The frequency upconversion unit 22 performs a frequency upconversion to increase the frequency of the audio signal acquired by the acquisition unit 21. Specifically, the frequency upconversion unit 22 shifts the frequency of the audio signal from the first band to the second band using a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform.

[0053] An extraction window function is used to extract a signal of a predetermined time interval from an audio signal. By discretely shifting the extraction window function on the time axis, signals can be extracted sequentially. In a lapped orthogonal transform, the extraction window function is shifted on the time axis under the condition that the extraction window function used for the previous extraction and the extraction window function used for the current extraction overlap by half the time width of the extraction window functions. In other words, a signal with a time length equal to the time width of the extraction window function is extracted every half the time width of the extraction window function. Hereinafter, the lapped orthogonal transform will also be simply referred to as the orthogonal transform.

[0054] Each of the overlapping signals extracted in this manner is subjected to an orthogonal transform and an inverse transform. Examples of the orthogonal transform include a discrete Fourier transform (DFT) and a discrete cosine transform (DCT). Examples of the inverse transform include an inverse discrete Fourier transform (IDFT) and an inverse discrete cosine transform (IDCT).

[0055] After the orthogonal transform is performed, but before the inverse transform is performed, a process is performed to shift the frequency of the frequency domain data to a higher frequency side. Through this process, the audio signal in the first band is transformed into an audio signal in the second band (band transformation). Note that, with the lapped orthogonal transform, the signal can be reproduced using N frequency components, which is half the number of data points 2N that represent the window width.

[0056] Step S2 will now be described in detail.

[0057] (Step S2a) The frequency upconversion unit 22 converts the frequency of the acquired audio signal into a frequency of the time interval [t ０ +mW / 2,t ０ +mW / 2+W] is extracted using the extraction window function. ０ represents the start time of the audio signal. W is the number of data points representing the window width. m is a parameter representing a positive integer with 0 as its initial value, and is incremented by 1 each time the determination in step S2f described below is "NO."

[0058] (Step S2b) The frequency upconversion unit 22 converts the time domain signal S ｍ A coefficient sequence in the frequency domain (hereinafter referred to as a first coefficient sequence) is obtained by performing an orthogonal transform on

[0059] In this specification, the term "coefficient" refers to an expansion coefficient obtained when a function representing a signal to be orthogonally transformed is expanded in an orthogonal function system. Each coefficient has a one-to-one correspondence with a frequency. Each coefficient is a numerical value representing the intensity of the corresponding frequency component. In particular, when DFT is used as the orthogonal transform, each coefficient represents not only the intensity but also the phase of the corresponding frequency component. Furthermore, the term "coefficient sequence" refers to data consisting of multiple coefficients arranged on the frequency axis. A coefficient sequence can be represented by a one-dimensional array in which coefficients as elements are identified by an index k representing a frequency. Furthermore, a subset of a coefficient sequence, which is a set of coefficients, is referred to as a "coefficient sequence subset."

[0060] (Step S2c) The frequency upconversion unit 22 obtains a new coefficient sequence (hereinafter referred to as a second coefficient sequence) from the first coefficient sequence by frequency shifting the coefficient sequence subset for the audible band to the coefficient sequence subset for the ultrasonic band. For example, when DFT is used as the orthogonal transform, the frequency upconversion unit 22 obtains X' ｍ (second coefficient sequence), X ｍ A part (coefficient sequence subset) of the (first coefficient sequence) is used, and the other coefficients are set to zero.

[0061] (Step S2d) The frequency upconversion unit 22 performs an inverse transformation of the above-mentioned orthogonal transformation on the second coefficient sequence, and multiplies the signal obtained by the inverse transformation by a reconstruction window function that satisfies the Princen-Bradley condition described above, thereby generating a time domain signal S' having a time width W. ｍ The frequency upconversion unit 22 performs, for example, IDFT as the inverse transformation. It is preferable that the reconstruction window function is the same as the extraction window function described above.

[0062] (Step S2e) The frequency upconversion unit 22 converts the time domain signal S' obtained in step S2d into ｍ is the time domain signal S' obtained in the previous step S2d. ｍ－１ The splicing is performed with a time overlap. The width of the overlapping section is equal to half the window width. Specifically, the time domain signal S' ｍ－１ and S' ｍ From these time domain signals S' ｍ－１ and S' ｍ The time interval [t ０ +mW / 2,t ０ +(m+1)W / 2] is obtained. This process of joining together is called reconstruction.

[0063] (Step S2f) The frequency upconversion unit 22 determines whether the processing up to step S2e has been completed for all time periods of the audio signal to be transmitted. If the processing up to step S2e has not been completed for all time periods of the audio signal (step S2f; NO), the frequency upconversion unit 22 returns to step S2a. In this way, the processing from step S2a to S2e is repeated multiple times.

[0064] On the other hand, if the processing from steps S2a to S2e has been completed for all time periods of the audio signal (step S2f; YES), the frequency upconversion unit 22 proceeds to step S3, since this means that the transmission signal has been completed by repeating step S2e multiple times.

[0065] (Step S3) The transmitter 23 transmits the transmission signal described above as an ultrasonic wave.

[0066] 2A illustrates an example of a configuration in which the entire audio signal is converted into a transmission signal and then the transmission signal is transmitted. For example, this configuration may be employed in cases where time-series data representing the entire audio signal is prepared in advance in the storage unit 25, the time at which the transmission signal is to be broadcast is predetermined, or there is a sufficient time difference between the time at which the audio signal is received and the time at which the transmission signal is transmitted.

[0067] However, the transmitting device 2 can also perform real-time processing (hereinafter referred to as real-time frequency upconversion processing) in which the transmitting device 2 acquires an audio signal and outputs a transmission signal in parallel with the acquisition of the audio signal. A specific example of this will be described below with reference to FIG. 2B.

[0068] As shown in Fig. 2B, the frequency upconversion (step S2) according to this example does not include the determination of step S2f shown in Fig. 2A. Instead, step S4, which determines whether the operation of the transmitter 2 has ended, is placed after the already-described step S3. That is, in this example, the already-described steps S1 to S3 are repeated until the operation of the transmitter 2 has ended. Each time this is repeated, the already-described parameter m is incremented by 1, which is the same as in Fig. 2A.

[0069] In this example, in step S1, the audio signal is acquired every time period equal to half the window width (hereinafter referred to as the unit time period) by the acquisition unit 21. The storage unit 25 also serves as a buffer (hereinafter referred to as the transmission input buffer) for storing the audio signal for that unit time period.

[0070] Then, the audio signal for the current unit time length is acquired from the transmission input buffer by the frequency upconversion unit 22. Then, the audio signal for the window width, which is a combination of the audio signal for the current unit time length and the audio signal for the previously acquired unit time length, is provided to the above-mentioned steps S2a to S2d, and as a result, a time domain signal for the window width is obtained.

[0071] In the reconstruction in step S2e, a transmission signal for a unit time length is formed using the time-domain signal for the currently obtained window width and the time-domain signal for the previous window width obtained in step S2d. The memory unit 25 also serves as a buffer (hereinafter referred to as a transmission output buffer) for storing the transmission signal for the unit time length.

[0072] In step S3, the transmitter 23 acquires a unit time length of transmission signals from the transmission output buffer, and transmits the acquired unit time length of transmission signals.

[0073] As described above, the audio signal for a unit time length is repeatedly converted into a transmission signal for a unit time length by the loop of steps S1 to S3, thereby realizing a real-time frequency upconversion process in which a transmission signal is output while an audio signal is being acquired.

[0074] Next, an example of the operation of the receiving device 3 will be described in detail with reference to FIG. 4A.

[0075] (Step S11) The receiving unit 31 receives a transmission signal as an ultrasonic wave transmitted by the transmitting unit 23 and propagating through the air.

[0076] (Step S12) The frequency down-conversion unit 32 performs a frequency down-conversion to lower the frequency of the transmission signal received by the receiving unit 31. Specifically, the frequency down-conversion unit 32 converts the frequency of the transmission signal from the second band back to the first band using a lapped orthogonal transform that uses an extraction window function that satisfies the Princen-Bradley condition, an inverse transform of the lapped orthogonal transform, and reconstruction by multiplication and addition of a reconstruction window function that satisfies the Princen-Bradley condition.

[0077] The extraction window function and reconstruction window function used in the receiving device 3 may be the same as the extraction window function and reconstruction window function used in the transmitting device 2, or may be a different window function that satisfies the Princen-Bradley condition. Even if the number of points in the DCT or DFT as an orthogonal transform, or the extraction window function and reconstruction window function are not the same between the receiving device 3 and the transmitting device 2, as long as the amount of frequency shift (Hz) is approximately the same between the transmitting device 2 and the receiving device 3, the receiving device 3 can reproduce an output signal that approximates the original audio signal. As a result, according to this embodiment, even in a pair of the transmitting device 2 and the receiving device 3 that do not necessarily have the same sampling frequency (for example, sampling frequencies of 44.1 kHz and 48 kHz), the information contained in the original audio signal can be transmitted to the user of the receiving device 3.

[0078] The procedure for frequency down-transform is the same as that for frequency up-transform described above, except for the direction on the frequency axis of the frequency shift performed after the orthogonal transform and before the inverse transform. Step S12 will now be described in detail.

[0079] (Step S12a) The frequency down converter 32 converts the frequency of the transmission signal received in step S11 into a frequency of the time interval [t ０ +mW / 2,t ０ +mW / 2+W] is extracted using the extraction window function. ０ represents the start time of the transmission signal. W is the number of data points representing the window width. m is a positive integer with 0 as the initial value, and is incremented by 1 each time the determination in step S12f described below is "NO."

[0080] (Step S12b) The frequency down-converter 32 converts the time domain signal U ｍ is transformed into a coefficient sequence in the frequency domain (hereinafter referred to as a third coefficient sequence) by orthogonal transformation.

[0081] (Step S12c) The frequency down-conversion unit 32 obtains a new coefficient sequence (hereinafter referred to as a fourth coefficient sequence) from the third coefficient sequence by frequency shifting the coefficient sequence subset for the ultrasonic band to the coefficient sequence subset for the audible band. For example, when DFT is used as the orthogonal transform, the frequency down-conversion unit 32 uses a part of the third coefficient sequence (coefficient sequence subset) and sets the other coefficients to zero when obtaining the fourth coefficient sequence.

[0082] (Step S12d) The frequency down-conversion unit 32 performs an inverse transformation of the above-mentioned orthogonal transformation on the fourth coefficient sequence, and multiplies the signal obtained by the inverse transformation by a reconstruction window function that satisfies the Princen-Bradley condition described above, thereby generating a time domain signal U′ of time width W. ｍ The frequency down-conversion unit 32 performs, for example, IDFT as the inverse transformation. It is preferable that the reconstruction window function is the same as the extraction window function described above.

[0083] (Step S12e) The frequency downconverter 32 converts the time domain signal U′ obtained in step S12d into ｍ is the time domain signal U' obtained in the previous step S12d. ｍ－１ The splicing is performed with a time overlap. The width of the overlapping section is equal to half the window width. Specifically, the time domain signal U' ｍ－１ and U' ｍ From these time domain signals U' ｍ－１ and U' ｍ The time interval [t ０ +mW / 2,t ０ +(m+1)W / 2] is obtained. This process of joining together is called reconstruction.

[0084] (Step S12f) The frequency down converter 32 determines whether or not the processing up to step S12e has been completed for all time periods of the transmission signal, which is the collected ultrasound. If the processing up to step S12e has not been completed for all time periods of the transmission signal (step S12f; NO), the frequency down converter 32 returns to step S12a. In this way, the processing from step S12a to S12e is repeated multiple times.

[0085] On the other hand, if the processing from steps S12a to S12e has been completed for all time intervals of the transmission signal (step S12f; YES), the frequency downconversion unit 32 proceeds to step S13, since this means that the output signal has been completed by repeating step S12e multiple times.

[0086] (Step S13) The output unit 33 outputs an output signal in the audible band. As described above, this output signal is a reproduction of the original audio signal.

[0087] 4A shows an example of a configuration in which the entire transmission signal as an ultrasonic wave is converted into an output signal and then the output signal is output. For example, this configuration may be adopted when there is a sufficient time difference between the time when the transmission signal is received and the time when the output signal is output.

[0088] However, the receiving device 3 can also perform real-time processing (hereinafter referred to as real-time frequency down-conversion processing) in which an output signal is output in parallel with receiving a transmission signal as an ultrasonic wave. A specific example of this will be described below with reference to FIG. 4B.

[0089] As shown in Fig. 4B, the frequency down-conversion (step S12) according to this example does not include the determination of step S12f shown in Fig. 4A. Instead, step S14, which determines whether the operation of the receiving device 3 has ended, is placed after the already-described step S13. That is, in this example, the already-described steps S11 to S13 are repeated until the operation of the receiving device 3 has ended. Each time this is repeated, the already-described parameter m is incremented by 1, which is the same as in Fig. 4A.

[0090] In this example, in step S11, a transmission signal as an ultrasonic wave is acquired for each unit time length by the receiving unit 31. The storage unit 35 also serves as a buffer (hereinafter referred to as a receiving input buffer) that stores the transmission signal for that unit time length.

[0091] Then, the transmission signal for the current unit time length is acquired from the receiving input buffer by the frequency down converter 32. Then, the transmission signal for the current unit time length and the transmission signal for the previously acquired unit time length, which is the sum of the transmission signal for the window width, are provided to the above-mentioned steps S12a to S12d, and as a result, a time domain signal for the window width is obtained.

[0092] In the reconstruction in step S12e, an output signal for a unit time length is formed using the time-domain signal for the currently obtained window width and the time-domain signal for the previous window width obtained in step S12d. The storage unit 35 also serves as a buffer (hereinafter referred to as a receiving output buffer) for storing the output signal for that unit time length.

[0093] In step S13, the output unit 33 acquires the output signal for the unit time length from the receiving output buffer, and outputs the acquired output signal for the unit time length to the user.

[0094] As described above, the loop of steps S11 to S13 repeatedly converts a transmission signal for a unit time length into an output signal for a unit time length, thereby realizing real-time frequency down-conversion processing that outputs an output signal while receiving a transmission signal.

[0095] The above provides an overview of the operation of the communication system 1. As described above, in the communication system 1, the frequency up conversion performed by the transmitting device 2 and the frequency down conversion performed by the receiving device 3 each use a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition, and an inverse transform of the lapped orthogonal transform. Therefore, according to this embodiment, the waveform of the output signal output by the receiving device 3 is less likely to be distorted compared to the waveform of the original audio signal.

[0096] The frequency up-conversion and frequency down-conversion techniques will be described in more detail below.

[0097] [Regarding Window Functions] First, the extraction window function to be used will be described. As described above, in this embodiment, an extraction window function that satisfies the Princen-Bradley condition is used for the frequency up-conversion and frequency down-conversion. Here, the "extraction window function that satisfies the Princen-Bradley condition" specifically means a window function that satisfies the Princen-Bradley condition when the number of data points representing the window width of the extraction window function is 2N, where j is an integer equal to or greater than 0 and satisfies j<N), and w[j] ２ +w[j+N] ２ = 1, where n is a parameter that represents the time position in the length direction of the window, and takes an integer value from 0 to 2N-1.

[0098] It is preferable that the extraction window function satisfy not only the Princen-Bradley condition but also line symmetry (w[j] = w[2N-1-j]). FIG. 6 shows a Kaiser-Bessel derived window as an example of a window function that satisfies the Princen-Bradley condition and line symmetry. By performing a lapped orthogonal transform using an extraction window function that satisfies the Princen-Bradley condition and line symmetry, and an inverse transform of the lapped orthogonal transform, the waveform of the output signal output from the receiving device 3 can be made even closer to the waveform of the original audio signal.

[0099] [Regarding Signal Extraction] First, the amplitude time-series data of the signal s to be frequency converted is extracted by the window width 2N using the extraction window function w[n] with the window width 2N described above. Extraction is repeated periodically at intervals of N samples. Here, "signal s to be frequency converted" refers to the audio signal described above if the frequency conversion is an up-conversion, and refers to the transmission signal described above if the frequency conversion is a down-conversion.

[0100] Array x representing the m-th signal extracted from the signal s to be frequency converted using the extraction window function w[n] ｍ is expressed by the following equation (1): ｍ Since the number of data points is 2N, n in the following equation (1) takes an integer value from 0 to 2N-1, as described above.ｍ If the frequency conversion is a frequency upconversion, the signal S ｍ and if the frequency conversion is a frequency down conversion, the signal U ｍ is equivalent to

[0101]

[0102] [Regarding Orthogonal Transform] Next, the extracted signal x ｍ By performing an orthogonal transformation on ｍ The coefficient sequence X representing the intensity distribution of each frequency component ｍ The coefficient sequence X ｍ The number of data points of the signal x ｍ The number of data points is equal to 2N. ｍ corresponds to the first coefficient sequence described above when the frequency transform is a frequency up transform, and corresponds to the third coefficient sequence described above when the frequency transform is a frequency down transform.

[0103] DFT as an orthogonal transform is expressed by the following equation (2): In the following equation (2), k is an index that identifies the position on the frequency axis, i.e., the above-mentioned frequency component, and takes an integer value from 0 to 2N-1. ｍ Each coefficient constituting the matrix A is generally a complex number, and A is a predetermined real number (for example, A=1 or A=1 / 2N).

[0104]

[0105] Alternatively, DCT may be used as the orthogonal transform. There are various subtypes of DCT, but it is preferable to use, for example, DCT (the following formula (3)) which returns to the original signal by a pair of transform and inverse transform. In the following formula (3), k is an index that identifies the above-mentioned frequency component and takes an integer value from 0 to 2N-1. The coefficient sequence X shown in the following formula (3) ｍ Each coefficient constituting the matrix A is a real number.

[0106]

[0107] [Regarding Frequency Transformation] Next, the coefficient sequence X obtained by the above equation (2) or (3) is ｍ, the coefficient sequence X' after frequency up-conversion (or frequency down-conversion) ｍ In this embodiment, the coefficient sequence X' is obtained. ｍ The number of data points is also ｍ and the number of data points in the signal x ｍ The number of data points is equal to 2N. ｍ corresponds to the second coefficient sequence described above when the frequency transformation is a frequency up transformation, and corresponds to the fourth coefficient sequence described above when the frequency transformation is a frequency down transformation.

[0108] Coefficient sequence X ｍ to the coefficient sequence X' ｍ The coefficient sequence X ｍ The number of shifts on the frequency axis of the coefficients constituting x, i.e., the number of shifts related to the parameter k shown in equation (2) or (3) (hereinafter referred to as the number of frequency shifts or simply the number of shifts) is set to M (..., -2, -1, +1, +2,...). ｍ From X ｍ When the orthogonal transformation to obtain X is DFT or DCT, ｍ From X' ｍ The conversion formula to obtain the above will be explained.

[0109] [DFT Case] ​​First, a frequency transform technique when the orthogonal transform is DFT will be described. I. When M>0, that is, in the case of frequency up-transformation, the coefficient sequence X' ｍ are written as the following equations (4a) to (4c) (excluding the cases of k=N, k=M, and k=2N-M).

[0110]

[0111] In this way, when DFT is used as the orthogonal transform, the frequency upconversion unit 22 converts the first coefficient sequence X ｍ is the coefficient number in the first coefficient sequence X, the number of frequency shifts for that number is M, and the window width of the window function for extraction is 2N. Except for the cases of k=N, k=M, and k=2N-M, the first coefficient sequence X is calculated using the above formulas (4a)-(4c) which are used depending on the values ​​of k, M, and N. ｍA part of the second coefficient sequence X' ｍ It is mapped to.

[0112] When k=N, the coefficient sequence X' ｍ is written as the following equation (5), (6), or (7): Any of equations (5) to (7) may be used.

[0113]

[0114]

[0115]

[0116] Furthermore, when k=M and k=2N-M, the coefficient sequence X' ｍ is written as the following equation (8).

[0117]

[0118] However, if there is no need to maximize the amount of information conveyed to the user of the receiving device 3, then for simplicity of implementation, X' ｍ [N] may be set to 0, or X' ｍ [M]=X′ ｍ It is also possible to set [2N-M]=0.

[0119] 7 illustrates an example of frequency shift when N=4 and M=+1, +2. Symbol g11 represents the case where N=4 and M=+1, and symbol g12 represents the case where N=4 and M=+2. As described above, X ｍ is the coefficient sequence before frequency shift, and X' ｍ is the coefficient sequence after frequency shift.

[0120] In the case of code g11, the shift number M is "+1", so the shift is performed by one using the above formula. As mentioned above, the formula used varies depending on the value of k. In the case of code g12, the shift number M is "+2", so the shift is performed by two using the above formula. In FIG. 7, the chain lines, dotted lines, solid lines, etc. indicate the correspondence between before and after the shift.

[0121] II. When M<0, i.e., in the case of frequency down-conversion, the coefficient sequence X' ｍare written as the following equations (9a) to (9c) (excluding the cases of k=0, k=N+M, and k=N−M).

[0122]

[0123] In this way, when DFT is employed as the orthogonal transform, the frequency down-conversion unit 32 calculates the third coefficient sequence Xm by using the above equations (9a) to (9c), which are used depending on the values ​​of k, M, and N, except for the cases where k=0, k=N+M, and k=N−M, where k is the coefficient number in the third coefficient sequence Xm, M is the number of frequency shifts associated with that number, and 2N is the window width of the extraction window function. ｍ A part of the fourth coefficient sequence X' ｍ It is mapped to.

[0124] When k=0, the coefficient sequence X' ｍ is written as the following equation (10), (11), or (12): Any of equations (10) to (12) may be used.

[0125]

[0126]

[0127]

[0128] Furthermore, when k=N+M or k=N-M, the coefficient sequence X' ｍ is written as the following equation (13).

[0129]

[0130] If it is not necessary to maximize the amount of information transmitted to the user of the receiving device 3, then for simplicity of implementation, X' ｍ [0]=0, or X' ｍ [N+M]=X' ｍ [N−M] may be set to 0.

[0131] 8 illustrates an example of frequency shift when N=4 and M=-1, -2. Reference symbol g21 represents the case where N=4 and M=-1, and reference symbol g22 represents the case where N=4 and M=-2.

[0132] In the case of code g21, the shift number M is "-1", so the shift is made by -1 using the above formula. In the case of code g22, the shift number M is "-2", so the shift is made by -2 using the above formula. In FIG. 8, the chain lines, dotted lines, solid lines, etc. indicate the correspondence between before and after the shift.

[0133] [In the Case of DCT] Next, a frequency transform technique when the orthogonal transform is DCT will be described. I. In the case of M>0, that is, in the case of frequency up-transformation, the coefficient sequence X' ｍ are written as the following equations (14a) and (14b).

[0134]

[0135] In this way, when DCT is used as the orthogonal transform, the frequency upconversion unit 22 converts the first coefficient sequence X ｍ is the coefficient number in the above equation (14a) and (14b) that can be used depending on the values ​​of k and M. ｍ A part of the second coefficient sequence X' ｍ It is mapped to.

[0136] 9 illustrates an example of frequency shift when N=4 and M=+1, +2. Reference symbol g31 represents the case where N=4 and M=+1, and reference symbol g32 represents the case where N=4 and M=+2.

[0137] In the case of code g31, the shift number M is "+1", so the shift is made by 2M = 2 using the above formula. In the case of code g32, the shift number M is "+2", so the shift is made by 2M = 4 using the above formula. In Fig. 9, the chain lines, dotted lines, solid lines, etc. indicate the correspondence between before and after the shift.

[0138] II. When M<0, i.e., in the case of frequency down-conversion, the coefficient sequence X' ｍ are written as the following equations (15a) and (15b).

[0139]

[0140] In this way, when DCT is used as the orthogonal transform, the frequency down-conversion unit 32 converts the third coefficient sequence X ｍ is the coefficient number in the above equation (15a) and (15b) that can be used depending on the values ​​of k and M. ｍ A part of the fourth coefficient sequence X' ｍ It is mapped to.

[0141] 10 illustrates an example of frequency shift when N=4 and M=-1, -2. Reference symbol g41 represents the case where N=4 and M=-1, and reference symbol g42 represents the case where N=4 and M=-2.

[0142] In the case of code g41, the shift number M is "-1", so the shift is made by 2M = -2 using the above formula. In the case of code g42, the shift number M is "-2", so the shift is made by 2M = -4 using the above formula. In Fig. 10, the chain lines, dotted lines, solid lines, etc. indicate the correspondence between before and after the shift.

[0143] To facilitate understanding of the manner in which a coefficient sequence is shifted in the frequency domain, the above description exemplifies the case in which the absolute value |M| of the number of frequency shifts M is 1 or 2. In practice, the number of frequency shifts M depends on the frequency resolution (frequency difference between k and k+1) ε [Hz] and the amount of frequency shift Δf [Hz] required to move from one of the first band and the second band to the other.

[0144] Specifically, the frequency shift number M is a value obtained by rounding up or down the decimal part of Δf / ε. As a specific example, when 2N=256, the frequency resolution ε is 187.5 [Hz], and the frequency shift amount Δf in the frequency up conversion is 20 [kHz], the frequency shift number M is 106 or 107. Note that in the case of frequency down conversion, the frequency shift amount Δf is negative.

[0145] [Regarding the inverse transformation of the orthogonal transformation] Next, the coefficient sequence X' whose frequency has been shifted in the above-described manner is ｍ Perform an inverse transform on x' in the time domain ｍ To get this inverse transformation, we need toｍ From X ｍ In this embodiment, the signal x' is transformed by the same orthogonal transform (for example, DFT or DCT) as that used to obtain the signal x'. ｍ The number of data points is also ｍ The number of data points, the coefficient sequence X ｍ and the number of data points of the signal x ｍ The number of data points is equal to 2N.

[0146] The IDFT, which is the inverse transform of the DFT, is expressed by the following equation (16). In the following equation (16), Real is a function that returns the real part of a complex number. k is an index that identifies a position on the time axis and takes an integer value from 0 to 2N-1. A' is a predetermined real number (e.g., A'=1).

[0147]

[0148] The IDCT, which is the inverse transform of the DCT, is expressed by the following equation (17): In the following equation (17), k is an index that identifies a position on the time axis and takes an integer value from 0 to 2N-1, and A' is a predetermined real number (e.g., A'=1).

[0149]

[0150] [Regarding reconstruction] x' obtained by the above equation (16) or (17) ｍ In this embodiment, in order to obtain a continuous signal in the time domain using the above formula, the reconstruction window function w[n] that satisfies the Princen-Bradley condition is set to x' ｍ Signal y ｍ It is preferable that the reconstruction window function w[n] is the same as the extraction window function used to extract s[n]. ｍ If the frequency conversion is a frequency upconversion, the signal S' ｍ and if the frequency conversion is a frequency down conversion, the signal U' already mentioned corresponds to ｍ is equivalent to

[0151]

[0152] Then, for all m, ..., yｍ－１ , y ｍ , y ｍ＋１ , ... are connected together. As a result, the signal s', which is the transmission signal or output signal, is obtained. The connection is performed with overlapping, as in the case of lapped orthogonal transform. That is, the set y ｍ－１ Set a part of y ｍ Overlap with part of.

[0153] Specifically, as shown in the following equation (19), a set y ｍ－１ and the next set y to be spliced ​​into that set ｍ and are overlapped by the number N of data points corresponding to a time width of half the window width, ..., y ｍ－１ , y ｍ , y ｍ＋１ , ... are joined together. This joining process is called reconstruction, as described in step S2e of Figures 2A and 2B and step S12e of Figures 4A and 4B.

[0154]

[0155] In equation (19), n is an index for identifying a position on the time axis and takes an integer value of 0, 1, 2, ..., N-1. m is an index for identifying the extracted signal and takes an integer value of 0, 1, 2, .... If side lobes caused by the reconstruction window function cannot be ignored, s' may be passed through a band limiting filter to remove signals outside the band desired to be used.

[0156] In this manner, the signal s' as the transmission signal and output signal described above are obtained. The transmitter 23 outputs the signal s' as the transmission signal in the ultrasonic band, and the output unit 33 outputs the signal s' as the output signal in the audible band.

[0157] [Regarding Even Shift and Odd Shift] A preferred embodiment of this embodiment will be described below. As described above, the frequency shift number M in the frequency transform after the orthogonal transform can be any integer value (excluding 0). Hereinafter, a shift number M of even numbers will be referred to as an even shift, and an odd shift will be referred to as an odd shift. In this embodiment, whether DFT or DCT is used, in the case of an odd shift, phase inversion is performed on the signals before and after the extraction. In other words, each time a signal is extracted using an extraction window function, a process of alternately repeating a "phase inversion operation" and a "phase preservation operation" (hereinafter referred to as an alternating phase inversion process) is performed. On the other hand, in the case of an even shift, the alternating phase inversion process is not performed, and the phases of all signals extracted using the extraction window function are preserved as they are. As a result, according to this embodiment, the phase information of the original audio signal is preserved regardless of whether the shift is an even shift or an odd shift.

[0158] Here, "phase inversion operation" refers to the operation of performing phase inversion on a signal extracted using an extraction window function, and "phase preservation operation" refers to not performing phase inversion on a signal extracted using an extraction window function, i.e., preserving the phase of the signal as is.

[0159] Here, the "signal extracted by the extraction window function" means a time domain signal obtained after the extraction operation using the extraction window function. Specifically, the "signal extracted by the extraction window function" is the time domain signal before orthogonal transformation, i.e., the aforementioned x ｍ or x', which is a time domain signal after inverse transformation and before multiplication by a reconstruction window function, as described later. ｍ or the previously mentioned y ｍ The "phase inversion operation" is equivalent to an operation of inverting the positive or negative sign of each instantaneous value constituting the time-series data representing the time-domain signal.

[0160] That is, x shown in the above formula (1) ｍ , x' shown in the above equation (18) ｍ , or y shown in the above formula (19) ｍFor odd shifts, preferably, represents the result of an alternating phase reversal process.

[0161] In the case of odd shifts, by performing alternating phase inversion processing, it is possible to prevent the loss of original amplitude information from the frequency-converted signal even if the shift number M is odd. As a result, even if the same transmission band selectability is realized, it is possible to reduce the window width of the extraction window function compared to the case where alternating phase inversion processing is not performed with odd shifts.

[0162] Generally, the time width of unwanted signals that appear in the output signal due to the effects of pre-echo and post-echo is approximately the window width of the window function used for extraction. For example, if the window width of the window function used for extraction is 682.67 ms, unwanted signals will appear approximately 300 ms before and after a sudden change in sound pressure.

[0163] Therefore, even if the transmission signal may be disturbed by sudden noise in the real environment, sudden audio occlusion, the Doppler effect due to sudden movement, etc., by keeping the window width of the extraction window function small, it is possible to reduce the time during which the influence of the disturbance appears in the output signal to a level that does not interfere with the recognition of the original amplitude information. For example, by keeping the window width of the extraction window function to 10.67 ms, the time during which an unwanted signal appears can be reduced to about 5 ms before and after, and humans can hardly recognize an unwanted signal of about 5 ms. Below, a specific example of the timing for performing the alternating phase inversion process will be described.

[0164] 3 illustrates the timing of the alternating phase inversion process in the frequency upconversion. In this example, the alternating phase inversion process is performed in step S2e of FIG. 2A or 2B, where the above-mentioned reconstruction is performed. That is, in this example, the time domain signal y ｍ corresponds to the above-mentioned "signal extracted by the extraction window function." FIG. 3 will now be described in detail.

[0165] First, the frequency upconversion unit 22 determines whether the frequency shift number in step S2c of FIG. 2A or FIG. 2B represents an even number shift or an odd number shift (step S31).

[0166] If the frequency shift number is an even number (step S31; NO), the frequency upconversion unit 22 converts the time domain signal y ｍ Without performing phase inversion of the time domain signal y ｍ are joined together (step S33).

[0167] On the other hand, if the frequency shift number indicates an odd number (step S31; YES), the frequency upconversion unit 22 performs the above-mentioned alternating phase inversion process (step S32) and then proceeds to step S33. ｍ and y ｍ－１ One of the two is subjected to the phase inversion operation, and the other is phase-preserved.

[0168] Note that FIG. 3 illustrates a configuration in which the alternating phase inversion process is performed in step S2e of FIG. 2A or FIG. 2B. However, in the frequency upconversion, the determination in step S31 and the alternating phase inversion process in step S32 may be performed in step S2a of FIG. 2A or FIG. 2B, or in step S2d of FIG. 2A or FIG. 2B.

[0169] 5 illustrates the timing of the alternating phase inversion process in the frequency down-conversion. In this example, the alternating phase inversion process is performed in step S12e of FIG. 4A or 4B, where the above-mentioned reconstruction is performed. That is, in this example, the time-domain signal y after the inverse transformation and multiplication by the reconstruction window function is ｍ corresponds to the above-mentioned "signal extracted by the extraction window function." Fig. 5 will now be described in detail.

[0170] First, the frequency downconversion unit 32 determines whether the frequency shift number in step S12c of FIG. 4A or FIG. 4B represents an even shift or an odd shift (step S41).

[0171] If the frequency shift number is an even number (step S41; NO), the frequency down-converter 32 converts the time domain signal y ｍ Without performing phase inversion of the time domain signal y ｍare joined together (step S43).

[0172] On the other hand, if the frequency shift number indicates an odd number (step S41; YES), the frequency down-conversion unit 32 performs the above-mentioned alternating phase inversion process (step S42) and then proceeds to step S43. ｍ and y ｍ－１ One of the two is subjected to the phase inversion operation, and the other is phase-preserved.

[0173] Note that while FIG. 5 illustrates a configuration in which the alternating phase inversion process is performed in step S12e of FIG. 4A or 4B, in the frequency down-conversion, the determination in step S41 and the alternating phase inversion process in step S42 may be performed in step S12a of FIG. 4A or 4B, or in step S12d of FIG. 4A or 4B.

[0174] 3 and 5, the time domain signal y after inverse transformation and multiplication by the reconstruction window function is ｍ is the "signal extracted by the extraction window function" mentioned above. In the following, the time domain signal before orthogonal transformation, ｍ Here, we will discuss the case where the signal corresponds to the "signal extracted by the extraction window function" described above.

[0175] FIG. 11 shows an example of a waveform extracted by an extraction window function when performing an even shift. Symbols g51 to g53 are examples of signals extracted sequentially with a time width that is half the time width of the window function. The horizontal axis represents time (ms), and the vertical axis represents the normalized magnitude of the signal. Waveforms g54 to g56 are the waveforms of the extracted signal x １ [n] to x ３ [n] is indicated.

[0176] In FIG. 11, when an even number of shifts is performed, any of the extracted signals x １ [n] to x ３ To clarify that the phase of [n] is also preserved, the extracted signal x １ [n] to x ３ [n] is displayed superimposed on the original signal s[n].

[0177] Figure 12 shows an example of a waveform extracted by an extraction window function when an odd shift is performed. Symbols g61 to g63 are examples of signals repeatedly extracted at intervals equal to half the window width. The horizontal axis represents time (ms), and the vertical axis represents the normalized signal magnitude. Waveforms g64 to g66 show the extracted signals.

[0178] In Figure 12, when an odd shift is performed, the extracted signal is displayed superimposed on the original signal s[n] to clearly show that the aforementioned ``phase inversion operation'' and ``phase preservation operation'' are alternately repeated each time a signal is extracted using the extraction window function.

[0179] Also, in FIG. 12, the signal x before frequency shifting is ｍ However, as described above, the alternating phase inversion process only needs to be performed once before the reconstruction described later, and the phase before the frequency shift (x ｍ ), after frequency shift (x' ｍ or y ｍ ) may be performed at any time.

[0180] FIG. 13 is a diagram showing an example of an extracted waveform when an even number of shifts is performed. Symbols g71 to g73 are examples in which a signal is repeatedly extracted at intervals equal to half the window width. The horizontal axis represents time (ms), and the vertical axis represents normalized signal magnitude. In FIG. 13, a waveform g74 before orthogonal transform for frequency upconversion and a waveform g75 after multiplying the inverse-transformed signal by a reconstruction window function are superimposed. Note that FIG. 13 shows an example in which the number of shifts is +2.

[0181] As shown in Fig. 13, in this embodiment, for example, frequency up-conversion is performed around 128+2 within a sample point period of 0 to 256. Note that Fig. 13 illustrates a waveform extracted from an audio signal by the frequency up-conversion unit 22, but the waveform extracted from a transmission signal by the frequency down-conversion unit 32 is similar to Fig. 13. However, the number of shifts during frequency down-conversion is a negative number.

[0182] Fig. 14 shows an example of an extracted waveform when an odd shift is performed. Symbols g81 to g83 are examples of signals repeatedly extracted at intervals equal to half the window width. The horizontal axis represents time (ms), and the vertical axis represents the normalized signal magnitude. In Fig. 14, a waveform g84 before frequency up conversion and a waveform g85 after frequency up conversion are shown superimposed.

[0183] 14 illustrates an example of a waveform extracted from an audio signal by the frequency up-conversion unit 22, but the waveform extracted from a transmission signal by the frequency down-conversion unit 32 is similar to that shown in Fig. 14. However, the number of shifts during frequency down-conversion is a negative number.

[0184] 15 is a diagram showing an example of reconstruction when an even shift is performed. Waveform g91 is the input waveform, and waveform g92 is the waveform after frequency upconversion. In this manner, in this embodiment, the "signal obtained by multiplying the inversely transformed signal by the reconstruction window function" shown in FIG. 13, which is the "signal extracted using the extraction window function" described above, is connected together, i.e., reconstructed, to obtain the ultrasound wave to be transmitted.

[0185] 15 illustrates the reconstruction performed in the frequency up conversion (step S2e in FIG. 2A or 2B), but the reconstruction performed in the frequency down conversion (step S12e in FIG. 4A or 4B) is also similar to that shown in FIG. 15. That is, the reconstruction is performed by each of the frequency up conversion unit 22 and the frequency down conversion unit 32.

[0186] 16 is a diagram showing an example of reconstruction when an odd-numbered shift is performed. Waveform g96 is the input waveform, and waveform g97 is the waveform after frequency upconversion. In this manner, in this embodiment, the "signal obtained by multiplying the inversely transformed signal by the reconstruction window function" shown in FIG. 14, which is the "signal extracted using the extraction window function" described above, is connected together, i.e., reconstructed, to obtain the ultrasound wave to be transmitted.

[0187] 16 illustrates the reconstruction performed in the frequency up conversion (step S2e in FIG. 2A or 2B), but the reconstruction performed in the frequency down conversion (step S12e in FIG. 4A or 4B) is also similar to that shown in FIG. 16. That is, the reconstruction is performed by each of the frequency up conversion unit 22 and the frequency down conversion unit 32.

[0188] The above-mentioned formulas are merely examples, and the formulas used for frequency transformation are not limited to these. Furthermore, the extraction window function and the reconstruction window function may be any window function that satisfies the Princen-Bradley condition. Furthermore, the orthogonal transformation method is not limited to the above-mentioned DFT or DCT, and other methods may also be used.

[0189] 7 to 16 are merely examples, and the number of shifts may be any value that corresponds to the frequency to be shifted. The frequency band to be shifted is not limited to the ultrasonic band, but may be within the audible band, for example, 16 kHz or higher.

[0190] FIG. 17 illustrates spectrograms of an original audio signal and a transmission signal. The horizontal axis of the spectrogram represents time, and the vertical axis represents frequency. Symbol g101 is a spectrogram of the original audio signal. Symbols g102 and g103 are spectrograms of the transmission signals. The frequency upconversion unit 22 upconverts the audio signal, for example, to the ultrasonic band, as shown by symbol g101. The transmission signal obtained by the frequency upconversion is then output from a speaker serving as the transmission unit 23, thereby performing ultrasonic transmission. The receiving device 3 then receives the transmission signal. The frequency downconversion unit 32 shifts the frequency band of the transmission signal to the audible band (frequency downconversion). Note that the frequency conversion processes on the transmitting and receiving sides can also be performed by a general-purpose processor or the like.

[0191] The audio signals that the transmitting device 2 targets for frequency upconversion are not limited to those representing human voices or other sounds. The audio signal may represent, for example, a warning sound, or may include a warning sound and an announcement. The audio signal may also represent a Morse code signal, or may be music that indicates that walking is permitted when the traffic light is green. In other words, in this specification, "audio" is not limited to the voices of living creatures, but is a concept that includes all sounds that propagate as sound waves. The transmitting device 2 can convert the various audio signals described above into ultrasonic band signals in real time.

[0192] Furthermore, the receiving device 3 may be used independently of the transmitting device 2. In other words, the transmission signal that the receiving device 3 performs frequency down-conversion on is not limited to that output by the transmitting device 2. For example, if the audio signal represents a warning sound, the frequency up-conversion in the transmitting device 2 can be omitted, and a signal similar to the above-described processing can be transmitted, received, and reproduced by transmitting a pre-synthesized ultrasonic sweep sound or the like and performing frequency down-conversion in the receiving device 3. Furthermore, the transmission signal that the receiving device 3 performs frequency down-conversion on may be a call emitted by a living creature, including animals such as bats and dolphins. The receiving device 3 can convert the transmission signal, such as the call, as an ultrasonic wave into an audible band sound in real time.

[0193] As described above, in this embodiment, the transmitting device 2 converts an audio signal of one band into an audio signal of another band (band conversion) using a lapped orthogonal transform and an inverse transform of the lapped orthogonal transform. For example, the transmitting device 2 converts an audio signal of an audible band into a transmission signal of an ultrasonic band. The transmission signal is transmitted through the air to the receiving device 3 carried by the user.

[0194] The receiving device 3 picks up the transmission signal as an ultrasonic wave with a microphone and converts the band of the signal into an output signal that reproduces the original audio signal in the audible band. The output signal is then heard by the user. This embodiment thus realizes audio transmission that is inaudible to anyone other than the user, so that audio information can be delivered only to those who need it.

[0195] Furthermore, the configuration of this embodiment does not require a tuning circuit or a modulation / demodulation circuit, which are essential for transmitting information via radio waves. To achieve a wavelength equivalent to that of an ultrasonic wave for spatially limited information transmission, it is necessary to use radio waves with a frequency of, for example, 20 GHz or higher. However, this radio wave band is already used for communications from space to Earth, making it impractical for everyday use.

[0196] Another prior art technique proposed is to use regularly arranged ultrasonic transducers to perform beamforming, generating spatially localized distorted sound for audio transmission (referred to as ultrasonic distorted sound audio transmission). This prior art technique poses a potential health risk because the ultrasonic waves used for transmission have extremely high sound pressure (110 dBSPL or higher). Furthermore, audio transmission using a typical speaker or ultrasonic distorted sound audio transmission can transmit audio to people who do not need it.

[0197] In contrast, in this embodiment, the sound pressure of the ultrasonic waves that are frequency-upconverted and transmitted through the air is at most about 80 dBSPL. Also, since the sound transmitted in this embodiment is converted into ultrasonic waves, it cannot be heard by people who do not have the receiving device 3. This feature makes it possible to prevent sound from being transmitted to people who do not need it.

[0198] By applying the method of this embodiment, real-time audio transmission is possible without relying on an external information line such as the Internet, and there is no need to obtain audio data in advance by downloading, etc. In contrast, when RFID (radio frequency identification) tags are used to transmit spatially limited information, real-time audio transmission without relying on an external information line is impossible, and if tags and audio are to be linked without using an external information line, audio data must be obtained in advance.

[0199] When prior art technologies using radio waves such as Bluetooth (registered trademark) are used, real-time voice transmission is possible without relying on external information lines, but the extent to which information transmission is localized in space is, in principle, highly dependent on the gain and sensitivity of the receiver and transmitter, which are not strictly defined.

[0200] In contrast, in this embodiment, the directivity of ultrasonic waves is utilized, making it possible to transmit information within the spatial range intended by the designer, and it is not greatly dependent on the gain or sensitivity of the receiver or transmitter. Furthermore, unlike information transmission using Bluetooth or the like, voice transmission using the method of this embodiment is not subject to the restrictions of the Radio Law as long as it does not emit unnecessary radio waves.

[0201] Prior art that achieves highly spatially selective audio transmission using ultrasound includes technology for transmitting and receiving audio using FM modulation (see, for example, Reference 1 below). However, like typical FM receivers used in radios, prior art uses a limiter circuit, which prevents the receiver user from directly utilizing the changes in received sound pressure due to the distance and direction between the ultrasonic transmitter and receiver. Reference 1: Kazuhiro Goshi and Yoshiaki Tadokoro, "Small-area communication using airborne ultrasound and its application to walking assistance for the visually impaired," Transactions of the Institute of Systems, Control and Information Engineers, Vol. 15, No. 11, 2002, pp. 569-576

[0202] In contrast, in this embodiment, the user hears the change in sound pressure due to distance and direction as it is, and can therefore know the distance from the transmitter 23 of the transmitter 2 to the receiver 3 and the direction of the receiver 3 relative to the transmitter 23b of the transmitter 2 from the change in sound pressure heard from the output unit 33 of the receiver 3.

[0203] [Modification 1] In the above embodiment, the first coefficient sequence X ｍ to X' as the second coefficient sequence ｍThe procedure for shifting on the frequency axis to M = 50 as a third coefficient sequence is illustrated. The shifting on the frequency axis may be performed in multiple steps. As an example, if the number of shifts required to achieve frequency up conversion is 106, the shift of 106 may be performed in two steps: a first shift operation with M = 50 and a second shift operation with M = 56. ｍ to X' as the fourth coefficient sequence ｍ The same applies to the shift on the frequency axis to

[0204] [Modification 2] In the above embodiment, in the transmitting device 2, the first coefficient sequence X ｍ The result of frequency upconversion of ｍ However, the second coefficient sequence is a second coefficient sequence that is a target of the inverse orthogonal transform to X ｍ In addition to the frequency upconversion, other conversion may be performed on the signal, specifically, an operation of changing the order of coefficient sequences on the frequency axis (hereinafter, a "changing operation").

[0205] Similarly, in the receiver 3, the time domain signal x' ｍ The fourth coefficient sequence, which is the target of the inverse orthogonal transform to ｍ In addition to the frequency down-transform, other transformations, specifically, an inverse permutation operation representing the inverse mapping of the permutation operation, may be performed on the data. Specific examples of such transformations are described below.

[0206] First, the processing in the transmitting device 2 will be described. In this modification, X ｍ is the target of the above-mentioned replacement operation (hereinafter referred to as the replacement target coefficient sequence). ｍ The swapping operation for is performed between steps S2b and S2c in FIG. 2A or 2B.

[0207] The frequency upconversion unit 22 then performs frequency upconversion on the signal that has undergone the replacement operation (hereinafter referred to as a replacement coefficient sequence). ｍAs with [k], Z is a one-dimensional array ｍ It is written as [k].

[0208] That is, in this modification, the first coefficient sequence X in the above-described equations (4a) to (8) and equations (14a) and (14b) used for frequency upconversion is ｍ [k] is the coefficient sequence Z after replacement ｍ It is replaced by [k].

[0209] The above replacement operation is performed on a set X with 2N elements. ｍ [i] is a set Z with 2N elements ｍ It can be expressed as a mapping F that has a one-to-one correspondence with [k]. Here, i is an index that identifies the position on the frequency axis, similar to the previously mentioned k, and takes integer values ​​of 0, 1, 2, ..., 2N-1. "Swapping the order of the coefficient sequence with respect to the position on the frequency axis" means, for example, shuffling the sequence of i=0, 1, 2... randomly or according to a predetermined rule, and defining it as the sequence of k again. In other words, as a result of the shuffling operation, for example, X ｍ [0] = Z ｍ [5], X ｍ [1] = Z ｍ [2], X ｍ [3] = Z ｍ [1]...the result is obtained.

[0210] In this modification, the coefficient sequence Z ｍ The result of frequency upconversion of [k] is X' ｍ [k] is the time domain signal x' ｍ The operation of the transmitting device 2 after this inverse orthogonal transform is the same as in the above embodiment. The transmission signal according to this modification is the first coefficient sequence X ｍ The result of performing a swap operation and frequency up-conversion on [i] is converted into a signal in the time domain.

[0211] Next, the processing in the receiving device 3 will be described. In this modification, X ｍ The coefficient sequence X' obtained by performing a frequency down-conversion on ｍis the target of the inverse permutation operation (hereinafter referred to as the inverse permutation target coefficient sequence). ｍ The reverse swap operation on is performed between steps S12c and S12d of FIG. 4A or 4B.

[0212] In this modification, the coefficient sequence X' ｍ The coefficient sequence obtained by performing the inverse permutation operation on the time domain signal x' (hereinafter referred to as the inverse permutation coefficient sequence) is ｍ is the fourth coefficient sequence to be subjected to inverse orthogonal transform.

[0213] In the following, the fourth coefficient sequence, which is the coefficient sequence after the inverse permutation, is represented as Z' ｍ That is, in this modification, the coefficient sequence X′ in equation (16) or (17) used for the inverse orthogonal transform is expressed as [k]. ｍ [k] is the fourth coefficient sequence Z′, which is the coefficient sequence after inverse permutation. ｍ [k]. Note that the inverse mapping F －１ is a set X' with 2N elements ｍ [k] is a set Z′ with 2N elements ｍ It corresponds one-to-one to [k].

[0214] The fourth coefficient sequence Z', which is the coefficient sequence after the inverse permutation, ｍ The operation of the receiving device 3 after the inverse orthogonal transform for [k] is the same as in the above embodiment. ｍ The result of performing an inverse permutation operation and frequency down-conversion on [i] is converted into a time domain signal.

[0215] According to the above-described modified example, the map F representing the permutation operation plays the role of the encryption key, and the inverse map F representing the inverse permutation operation －１ Since the inverse mapping F plays the role of a decryption key, the secrecy of communication can be improved. In other words, even if a transmission signal according to this modification is intercepted, it is difficult to obtain meaningful information from the transmission signal simply by lowering the frequency of the transmission signal to the audible band. －１ Information can be selectively communicated only to those who know the rules.

[0216] In the above modification, the first coefficient sequence X ｍ is the coefficient sequence to be replaced, but the first coefficient sequence X ｍ The coefficient sequence X' obtained by performing frequency upconversion on ｍ In this case, the coefficient sequence X' to be replaced may be ｍ The replacement operation for the replacement target coefficient sequence X' is performed between steps S2c and S2d in FIG. 2A or 2B. ｍ The coefficient sequence W' obtained by performing a replacement operation on ｍ is the time domain signal x' ｍ That is, the coefficient sequence X' in equation (16) or (17) used for the inverse orthogonal transform is the second coefficient sequence used for the inverse orthogonal transform. ｍ [k] is the coefficient sequence W' ｍ Replace with [k].

[0217] In the above modification, the receiving device 3 receives the third coefficient sequence X ｍ The coefficient sequence X' obtained by performing a frequency down-conversion on ｍ is the coefficient sequence to be reverse-interchanged, but the third coefficient sequence X ｍ In this case, the coefficient sequence X ｍ The reverse permutation operation for the reverse permutation target coefficient sequence X is performed between steps S12b and S12c in FIG. 4A or 4B. ｍ The coefficient sequence W obtained by performing the reverse permutation operation on ｍ is the target of frequency down-conversion. That is, the coefficient sequence X ｍ [k] is the coefficient sequence W ｍ In this case, the coefficient sequence W ｍ The result of frequency down-conversion is ｍ is the time domain signal x' ｍ is the fourth coefficient sequence used in the inverse orthogonal transform to

[0218] [Variation 3] By limiting the bandwidth of a signal for one channel and setting a different frequency shift number for each channel, multi-channel communication may be realized in which signals from multiple channels are superimposed on one band.

[0219] 18 shows an example of a frequency shift mode when performing multi-channel communication using DCT. Symbol g201 indicates a shift mode in the case of multi-channel frequency up-conversion in which the number of shifts used to raise the frequency differs for each channel, and symbol g202 indicates a shift mode in the case of multi-channel frequency down-conversion in which the number of shifts used to lower the frequency differs for each channel.

[0220] The absolute values ​​of the frequency shift numbers for the same channel are made equal in the multi-channel frequency up-conversion and the multi-channel frequency down-conversion. Fig. 18 illustrates an example in which, when the orthogonal transform is DCT, the absolute value |MCH1| of the frequency shift number for the first channel (CH1) is made equal to 3, and the absolute value |MCH2| of the frequency shift number for the second channel (CH2) is made equal to 6.

[0221] By applying this multi-channel communication, the following can be achieved: I. Multiple channel signals can be stacked in one band. For example, two channel signals, each with a band of 0-2 kHz, can be stacked in a band of 20-24 kHz. II. A wideband signal transmission path can be realized by bundling signal transmission paths with limited bandwidth. For example, by bundling two transmission paths, each with a band of 0-48 kHz, a single channel transmission path with a band of 0-96 kHz can be realized.

[0222] Note that the above-mentioned frequencies and the number of channels that can be stacked are merely examples and are not limited to these. Furthermore, the bandwidths of all channels do not need to be equal. The bandwidth of each channel can be freely set. For example, when performing multi-channel communication using channels 1 to 3, if the bandwidth of the first channel is 1, the bandwidth of the second channel may be 3 / 4 and the bandwidth of the third channel may be 1 / 2. Note that the bandwidth of each channel can be freely set at the interval of the frequency resolution determined by the width of the window function.

[0223] As described above, according to this modification, information for multiple channels can be included in the transmission signal output by the transmitter 2. In other words, the transmitter 2 can simultaneously broadcast information for multiple channels.

[0224] For example, each channel can carry audio information in a different language, such as channel 1 carrying Japanese audio information, channel 2 carrying English audio information, and channel 3 carrying German audio information. In other words, simultaneous multi-language broadcasting is possible, in which audio information in multiple languages ​​is broadcast simultaneously.

[0225] 18 illustrates a configuration in which the receiving device 3 performs frequency down-conversion for all channel bands in a transmission signal containing information for multiple channels. However, the receiving device 3 may have a channel selection function that performs frequency down-conversion only for a predetermined channel band and outputs an output signal obtained by the frequency down-conversion.

[0226] To achieve the channel selection function, the frequency down-converter 32 down-converts only a subset of the third coefficient sequence that corresponds to a predetermined channel band to obtain a fourth coefficient sequence corresponding to that channel. The frequency down-converter 32 then inversely converts the fourth coefficient sequence to obtain a time-domain signal, and reconstructs the obtained time-domain signal to obtain an output signal. The output signal thus obtained contains only the information to be transmitted through the predetermined channel.

[0227] As an example, even if the transmitting device 2 broadcasts simultaneously in multiple languages, if the receiving device 3 has a channel selection function, it can extract only the information of the channel corresponding to the language used by the user from the transmission signal broadcast simultaneously in multiple languages ​​and allow the user to listen to it.

[0228] [Modification 4] In the above-described embodiment and modifications 1 to 3, the value of the maximum frequency of the second band is not particularly limited, but it is preferable that the value of the maximum frequency of the second band be less than the Nyquist frequency when the audio signal is formed. This is because, even if the formation of the audio signal satisfies the sampling theorem, if part of the information contained in the audio signal (hereinafter referred to as audio information) is lost during frequency upconversion, the lost audio information cannot be reproduced by the receiving device 3.

[0229] An audio signal is a sampled audio or other analog signal (hereinafter referred to as the original signal). The Nyquist frequency in this sampling is the "Nyquist frequency when forming the audio signal" mentioned above. Furthermore, "the formation of the audio signal satisfies the sampling theorem" means that the maximum frequency included in the original signal is less than the Nyquist frequency.

[0230] In light of the above, a preferred embodiment in which audio information is less likely to be lost in frequency upconversion will be described below. This modification is applicable to any of the above-described embodiment and modifications 1 to 3.

[0231] That is, in this modification, the frequency upconversion unit 22 performs upsampling on the audio signal. The upsampling is a process of converting a signal sampled at a first sampling frequency into a signal equivalent to a signal sampled at a second sampling frequency higher than the first sampling frequency. Note that a known method can be used for the upsampling.

[0232] That is, upsampling increases the temporal resolution of the audio signal. Here, "temporal resolution" refers to the number of points per unit time of the parameter n, which identifies the position on the time axis, or the shortness of the time width represented by the interval between parameters n and n+1.

[0233] The target of the upsampling process is an audio signal acquired by the acquisition unit 21. Note that "acquire" here does not necessarily mean the above-mentioned sampling. The acquisition unit 21 may externally receive an audio signal obtained by sampling an original signal, or the acquisition unit 21 may sample the original signal to form an audio signal.

[0234] The upsampling process is performed before the extraction using the extraction window function w[n]. Specifically, the frequency upconversion unit 22 performs the upsampling process on the audio signal between steps S1 and S2a in FIG. 2A or 2B.

[0235] The audio signal that has been subjected to the upsampling process is then subjected to the processes from step S2b onward in Fig. 2A or 2B as the audio signal described in the embodiment and Modifications 1-3. As a result, in this modification, the frequency upconversion unit 22 forms a transmission signal that has a higher temporal resolution than the audio signal before the upsampling process.

[0236] According to this modification, even if the maximum frequency of the second band is set to a high value, audio information is less likely to be lost in the frequency upconversion compared to when upsampling processing is not performed.

[0237] A specific example will be described below in which the sampling frequency for sampling the original signal is 48 kHz, that is, the Nyquist frequency is 24 kHz.

[0238] In this case, if no upsampling process is performed, the maximum frequency of the second band is preferably limited to less than the Nyquist frequency of 24 kHz in order to minimize the loss of audio information in the frequency upconversion.

[0239] In contrast, in this modification, the temporal resolution of the audio signal is increased by a factor of two through upsampling processing. In other words, the audio signal after upsampling processing corresponds to a signal sampled at a sampling frequency of 96 kHz, i.e., the Nyquist frequency of 48 kHz.

[0240] In this case, even if the maximum frequency of the second band is 24 kHz or higher, loss of audio information in the frequency upconversion can be avoided as long as the Nyquist frequency after upsampling is less than 48 kHz. In other words, it is possible to transmit a transmission signal using a frequency band with a Nyquist frequency of 24 kHz or higher before upsampling.

[0241] [Variation 5] In the above-described variation 4, the loss of audio information in the frequency upconversion is suppressed by increasing the temporal resolution of the audio signal in advance before the frequency upconversion. Below, another method for suppressing the loss of audio information in the frequency upconversion will be described.

[0242] This modification can also be applied to any of the above-described embodiment and modifications 1 to 3. Furthermore, this modification can also be applied to the above-described modification 4.

[0243] During frequency upconversion, the frequency upconversion unit 22 according to this modification converts the second coefficient sequence X' obtained by the frequency upconversion. ｍ The number of data points in [k] is the first coefficient sequence X ｍ A high-resolution process is performed to increase the number of data points beyond [k]. Specifically, the high-resolution process is performed in step S2c of FIG. 2A or FIG. 2B.

[0244] By the high-resolution processing, the second coefficient sequence X' is obtained in the frequency domain. ｍ The number of data points in [k] is calculated as the first coefficient X ｍ Increasing the number of data points in [k] beyond that of the window function w[n] in the time domain corresponds to increasing the number of data points per time length represented by the window width of the window function w[n] for extraction. As a result, in this modification, as in the previous modification 4, a transmission signal having higher temporal resolution than the audio signal is formed.

[0245] The high resolution processing will be specifically described below.

[0246] 19 conceptually shows the process of increasing the resolution. ｍ The number of data points in [k] is 2N, which is equal to the number of data points in the extraction window function w[n], as in the above embodiment.

[0247] In the above embodiment, the second coefficient sequence X' ｍ The number of data points in [k] is also 2N. On the other hand, in this modification, the second coefficient sequence X' ｍ The number of data points in [k] is increased by 2 Ｐ (2N) (where P is an integer equal to or greater than 1). In other words, according to the high resolution processing, the number of data points in the frequency domain is increased to 2 Ｐ will be doubled.

[0248] Second coefficient sequence X' ｍ The number of data points in [k] is the first coefficient sequence X ｍ 2 of the data points [k] Ｐ Therefore, even if the number of shifts M is large, the first coefficient sequence X ｍ [k] is the second coefficient sequence X' ｍ [k], the first coefficient sequence X ｍ The coefficients constituting [k] are less likely to be lost. In other words, audio information is less likely to be lost. ｍ The newly added region EX in [k], that is, (2 Ｐ -1) For example, zero may be substituted for the coefficients for 2N points.

[0249] As described above, in this modification, the second coefficient sequence X' ｍThe number of data points in [k] is 2 Ｐ Since (2N) can be obtained, in the above-described reconstruction (step S2e in FIG. 2A or FIG. 2B), the number of data points is set to 2 as the reconstruction window function. Ｐ The formula representing the reconstruction window function may be the same as the formula representing the extraction window function.

[0250] The number of data points of the reconstruction window function according to this modification is 2. Ｐ Although (2N) is greater than the number of data points of the extraction window function, the time length represented by the reconstruction window function according to this modification is the same as the time length represented by the extraction window function. As a result, a transmission signal with higher temporal resolution than the audio signal is obtained.

[0251] Specifically, according to this modification, the temporal resolution of the transmission signal is increased to 2P times the temporal resolution of the audio signal. ｓ Then, the transmission signal according to this modification is Ｐ ・f ｓ This corresponds to a signal sampled at a sampling frequency of .

[0252] According to this modification, even if the maximum frequency of the second band is set to a high value, audio information is less likely to be lost in the frequency upconversion than when the resolution enhancement process is not performed. A specific explanation will be given below using an example in which the sampling frequency for sampling the original signal is 12 kHz, i.e., the Nyquist frequency is 6 kHz.

[0253] In this case, if high-resolution processing is not performed, the maximum frequency of the second band is preferably limited to less than the Nyquist frequency of 6 kHz in order to minimize the loss of audio information in the frequency upconversion.

[0254] In contrast to this, in this modification, the second coefficient sequence X' is obtained by the high-resolution processing. ｍ The number of data points in [k] is the first coefficient sequence X ｍAssume that the number of data points [k] is increased to eight times (P=3). In this case, it is possible to transmit the transmission signal using a frequency band higher than 6 kHz, for example, the ultrasonic band of 30 kHz±3 kHz, while avoiding the loss of audio information.

[0255] Furthermore, the resolution enhancement process according to this modification is a process performed between the extraction using the extraction window function and the splicing using the reconstruction window function, i.e., a process that is repeated each time the extraction using the extraction window function is performed. Therefore, compared to the fourth modification in which the above-mentioned upsampling process is performed all at once prior to the extraction using the extraction window function, the real-time performance from the acquisition of the audio signal to the output of the transmission signal is less likely to be impaired.

[0256] In the high-resolution processing according to this modification, the number of data points in the frequency domain is increased, but the first coefficient sequence X ｍ This does not involve quantization of the coefficients that make up [k], so audio information is less likely to be lost compared to Modification 4, which employs a method involving quantization of audio signals as upsampling processing.

[0257] [Variation 6] In the following description, the "power" of a signal refers to the temporal average value of the squared amplitude of the wave represented by the signal. If the wave is a sound wave, the higher the sound pressure, the greater the power.

[0258] In the above-described embodiment and modifications 1-5, the receiving device 3 preferably has a configuration in which the power of the output signal output by the output unit 33 increases as the power of the transmission signal received by the receiving unit 31 increases. Here, the power of the output signal output by the output unit 33 refers to the volume of sound heard by the user of the receiving device 3.

[0259] In general, the power of the transmission signal at the position of the receiving unit 31 changes depending on the direction and distance of the receiving unit 31 relative to the position of the transmitting unit 23 as the source of the transmission signal. Therefore, when the receiving device 3 has the above configuration, the power of the output signal output from the output unit 33 changes in accordance with the change in the power of the transmission signal at the position of the receiving unit 31.

[0260] Specifically, the power of the output signal increases as the direction and distance of the position of the receiving unit 31 are closer to the position of the transmitting unit 23. Also, the power of the output signal decreases as the direction and distance of the position of the receiving unit 31 are farther from the position of the transmitting unit 23.

[0261] According to this modification, for example, the following effects (A) and (B) can be obtained.

[0262] (A) When multiple transmitting units 23 that emit transmission signals with different contents are distributed in a certain area, a user can selectively listen to only the information contained in the transmission signal emitted by a specific transmitting unit 23.

[0263] That is, when the receiver 31 is brought closer to a desired transmitter 23 among the plurality of transmitters 23, the power of the transmission signal emitted by the desired transmitter 23 becomes relatively much larger among the plurality of types of transmission signals that can be received by the receiver 31. On the other hand, the power of the transmission signals emitted by transmitters 23 other than the desired transmitter 23 among the plurality of transmitters 23 becomes negligibly small, so that unnecessary information is essentially blocked. Therefore, the user can selectively listen to information contained in the transmission signal emitted by the desired transmitter 23.

[0264] (B) The user of the receiving device 3 can grasp the approximate position of the transmitting unit 23 without relying on vision by hearing how the power of the output signal changes when the position or posture of the receiving unit 31 held by the user is changed.

[0265] In particular, when the transmission signal output from the transmitter 23 is an ultrasonic wave, ultrasonic waves have higher directionality than sound waves, so there is a noticeable change in the power of the received transmission signal when the position or posture of the receiver 31 is changed. Therefore, there is also a noticeable change in the power of the output signal heard by the user, making it easier for the user to grasp the position of the transmitter 23.

[0266] [Modification 7] A preferred embodiment for achieving the effect (A) according to the above-described modification 6 will now be described.

[0267] Consider a case where work is performed according to location at a work site such as a factory. A transmitter 23 is installed at each location, and the transmitter 23 emits a transmission signal indicating the work to be performed at the location where the transmitter 23 is installed.

[0268] In this case, each time a user wearing the receiving unit 31 moves to a new location, the user can hear instructions on the work content appropriate to that location through the transmission signal emitted by the transmitting unit 23 at the new location.

[0269] When the user moves to a certain location, the receiver 31 worn by the user is located within a range of, for example, several tens of centimeters from the transmitter 23 at that location. When the receiver 31 is brought this close to the transmitter 23, it is possible to selectively receive instructions about work content through that transmitter 23, and it is possible to substantially block instructions about work content issued by other transmitters 23.

[0270] Furthermore, in the configuration described above, the transmitter 23 installed at each location may perform simultaneous broadcasting in multiple languages ​​as described in Modification Example 3. This allows the transmitter 23 installed at each location to be shared by multiple users even when multiple users who speak different languages ​​work at the site. In other words, it is not necessary to install a transmitter 23 for each language used by a user at each location.

[0271] In this case, it is preferable that the receiving device 3 used by each user has the channel selection function described in Modification 3. When the receiving device 3 has the channel selection function, each user can hear instructions about the work content in the language that he or she uses at each location.

[0272] [Variation 8] The effect (B) according to Variation 6 described above can be obtained even when the receiving device 3 includes only one microphone constituting the receiving unit 31. In that case, however, the user needs to perform an action of turning the body (hereinafter referred to as a rotation action) in order to change the orientation of the receiving unit 31 attached to the body with respect to the transmitting unit 23.

[0273] Therefore, below, as a preferred embodiment for achieving the above effect (B), we will describe a configuration that allows the user of the receiving device 3 to grasp the position of the transmitting unit 23 without performing any rotational movement or even with minimal rotational movement.

[0274] 20, in the receiving device 3 according to this modification, the receiving unit 31 has a left-body partial receiving unit 31a and a right-body partial receiving unit 31b. Each of the left-body partial receiving unit 31a and the right-body partial receiving unit 31b is configured with a microphone capable of collecting sounds including ultrasonic waves.

[0275] The left-body partial receiving unit 31a is worn on the left half of the user's body, such as the left side of the head, left shoulder, left hip, left foot, etc. On the other hand, the right-body partial receiving unit 31b is worn on the right half of the user's body, such as the right side of the head, right shoulder, right hip, right foot, etc.

[0276] The frequency down converter 32 is shared by both the left body receiving section 31 a and the right body receiving section 31 b, and converts the transmission signal received by the left body receiving section 31 a and the transmission signal received by the right body receiving section 31 b into individual output signals.

[0277] Specifically, the frequency downconversion unit 32 forms an output signal (hereinafter referred to as the output signal for the left ear) obtained by shifting the frequency of the transmission signal received by the left half body partial receiving unit 31a to the first band, and an output signal (hereinafter referred to as the output signal for the right ear) obtained by shifting the frequency of the transmission signal received by the right half body partial receiving unit 31b to the first band.

[0278] Furthermore, in the receiving device 3 according to this modification, the output unit 33 has a left-ear partial output unit 33a and a right-ear partial output unit 33b. The left-ear partial output unit 33a outputs the above-described left-ear output signal to the user's left ear. The right-ear partial output unit 33b outputs the above-described right-ear output signal to the user's right ear. It is preferable that the pair of the left-ear partial output unit 33a and the right-ear partial output unit 33b constitute earphones or headphones.

[0279] The effects of the above configuration will be described below. First, consider a case where there is only one transmitter 23 as a source of a transmission signal around the user of the receiving device 3.

[0280] When the transmitter 23 is located to the left of the user, the left-side body receiving unit 31a is closer to the transmitter 23 than the right-side body receiving unit 31b, and therefore the power of the left-ear output signal output to the user's left ear is greater than the power of the right-ear output signal output to the user's right ear. Therefore, the user can sense that the transmitter 23 is located to the left of them solely through hearing, without having to rotate the device.

[0281] Similarly, when the transmitter 23 is located to the right of the user, the power of the right-ear output signal output to the user's right ear is greater than the power of the left-ear output signal output to the user's left ear, so the user can sense that the transmitter 23 is located to the right of them through hearing alone, without having to perform a rotational movement.

[0282] Next, consider a case where there are a plurality of transmitters 23 serving as sources of transmission signals around the user of the receiving device 3.

[0283] Specifically, consider a situation in which a men's restroom with a first transmitting unit 23 installed is located to the left of the user, and a women's restroom with a second transmitting unit 23 installed is located to the right of the user. The transmission signal transmitted by the first transmitting unit 23 is a frequency-up converted audio signal saying "This is the men's restroom," and the transmission signal transmitted by the second transmitting unit 23 is a frequency-up converted audio signal saying "This is the women's restroom."

[0284] In the above-described situation, the first transmitting unit 23 is closer to the left half body receiving unit 31a than the second transmitting unit 23. Therefore, in the left ear output signal output to the left ear of the user, the power of the voice saying "This is the men's restroom" is greater than the power of the voice saying "This is the women's restroom."

[0285] Furthermore, in the above-described situation, the second transmitting unit 23 is closer to the right-side body partial receiving unit 31b than the first transmitting unit 23. Therefore, in the right-ear output signal output to the user's right ear, the power of the voice saying "This is the ladies' restroom" is greater than the power of the voice saying "This is the men's restroom."

[0286] Therefore, the user can know by hearing alone that the men's restroom is on the left side of the user and the women's restroom is on the right side of the user without performing a rotational movement. For example, if the user is male, the user can know by hearing alone that the desired first transmitting unit 23 and the men's restroom are on the left side of the user without performing a rotational movement.

[0287] [Modification 9] When the receiving unit 31 is configured with a plurality of partial receiving units each consisting of a microphone, it may be possible to select and use a desired one of the plurality of microphones as appropriate depending on the situation. A specific example of this is described below.

[0288] 21, in the receiving device 3 according to this modification, the receiving unit 31 has a left-half body partial receiving unit 31a, a right-half body partial receiving unit 31b, and a fingertip partial receiving unit 31c. Each of the left-half body partial receiving unit 31a, the right-half body partial receiving unit 31b, and the fingertip partial receiving unit 31c is made up of a microphone as a partial receiving unit having the function of receiving a transmission signal.

[0289] The fingertip receiving unit 31c is attached to the fingertip of the user's dominant hand. Here, the fingertip of the dominant hand is an example of an extremity of the user's body, and the fingertip receiving unit 31c is an example of a proximity area receiving unit attached to an extremity of the user's body.

[0290] In this modification, the left body partial receiving unit 31a is worn on the left side of the user's head, and the right body partial receiving unit 31b is worn on the right side of the user's head. Each of the left body partial receiving unit 31a and the right body partial receiving unit 31b is an example of an approach area partial receiving unit that is worn at a position closer to the head than to the extremities of the user's body.

[0291] The frequency downconversion unit 32 in this modified example has the function of switching between the transmission signal received by the left half body partial receiving unit 31a, the right half body partial receiving unit 31b, and the fingertip partial receiving unit 31c to be used to form the output signal to be output to the output unit 33.

[0292] As in the sixth and seventh modifications, the power of the output signal increases as the power of the transmission signal used to generate the output signal increases.

[0293] Hereinafter, the state in which the transmission signal received by the left half body receiving section 31a is used to form the output signal to be output by the output section 33 will be referred to as the "left approach area information acquisition state." Also, the output signal output from the output section 33 in the left approach area information acquisition state will be referred to as the "left approach area output signal."

[0294] The state in which the transmission signal received by the right half body receiving section 31b is used to form the output signal to be output from the output section 33 is referred to as the "right approach area information acquisition state." The output signal output from the output section 33 in the right approach area information acquisition state is referred to as the "right approach area output signal."

[0295] Furthermore, a state in which the transmission signal received by the fingertip portion receiving unit 31c is used to form the output signal to be output from the output unit 33 is referred to as a "nearby area information acquisition state." Furthermore, an output signal output from the output unit 33 in the near area information acquisition state is referred to as a "nearby area output signal."

[0296] In this modified example, the frequency downconversion unit 32 receives a user operation and switches to one of the above-mentioned left approach area information acquisition state, right approach area information acquisition state, and proximity area information acquisition state according to the user operation.

[0297] The effects of the above configuration will be described below by taking a specific example of a situation in which a user accesses a desired product in a store.

[0298] A store is equipped with a plurality of product display shelves, each of which is equipped with a transmitter 23. The transmission signal output from the transmitter 23 (hereinafter referred to as the attribute information transmitting unit 23) installed above each product display shelf indicates attributes common to the plurality of types of products stored on that product display shelf.

[0299] Each product display shelf has a plurality of shelves arranged vertically, and each shelf is also equipped with a plurality of transmitters 23. The transmission signal output from the transmitter 23 (hereinafter referred to as the individual information transmitting unit 23) installed on each shelf represents specific explanatory content about the product placed at the location where the individual information transmitting unit 23 is installed.

[0300] In the situation described above, a user entering a store first switches the receiving device 3 to a state for acquiring information for the left approach area or a state for acquiring information for the right approach area. The user then listens to information indicating attributes common to multiple products displayed on the product display shelves from the respective attribute information transmitting units 23 via the left approach area output signal or the right approach area output signal.

[0301] The left approach area output signal and the right approach area output signal may include not only the sound obtained by frequency-down-converting the transmission signal output by the attribute information transmitting unit 23 (hereinafter referred to as the attribute information providing sound), but also the sound obtained by frequency-down-converting the transmission signal output by the individual information providing transmitting unit 23 (hereinafter referred to as the individual information providing sound). However, the left body partial receiving unit 31a and the right body partial receiving unit 31b are located closer to the attribute information providing transmitting unit 23 than the individual information providing transmitting unit 23. For this reason, in the left approach area output signal and the right approach area output signal, the power of the individual information providing sound is negligibly small, and the user can mainly hear the attribute information providing sound.

[0302] Therefore, the user can first identify a desired product display shelf from among multiple product display shelves using the attributes described in the attribute information provision voice as a clue. In the process of identifying a desired product display shelf, the user can know whether the desired product display shelf is on the left or right of the user by switching between the left approach area information acquisition state and the right approach area information acquisition state.

[0303] Next, the user who has identified the desired product display shelf switches the receiving device 3 to a proximity area information acquisition state and then extends his / her hand equipped with the fingertip receiving unit 31c to the shelf of the product display shelf.The user then hears specific explanations about each product placed on the shelf from the individual information transmitting units 23 via the proximity area output signals.

[0304] The near area output signal may include not only the individual information providing audio but also the attribute information providing audio. However, the fingertip portion receiving unit 31c is located closer to the individual information providing transmitter 23 than the attribute information providing transmitter 23. Therefore, in the near area output signal, the power of the attribute information providing audio is negligibly small, and the user can mainly hear the individual information providing audio.

[0305] Therefore, the location of the desired product can be identified using the explanation given by the individual information provision voice as a clue. In the process of identifying the location of the desired product, the user can listen to explanations about various products on the desired product shelf by scanning the hand on which the fingertip receiving unit 31c is attached.

[0306] As described above, with the receiving device 3 according to this modified example, the user can appropriately select whether to receive information provided from the attribute information providing transmitter 23 using the left half body partial receiving unit 31 a or the right half body partial receiving unit 31 b in an approach area, for example, several meters to several tens of centimeters, from the attribute information providing transmitter 23, or (ii) to receive information provided from the individual information providing transmitter 23 using the fingertip partial receiving unit 31 c in a nearby area, for example, several centimeters, from the individual information providing transmitter 23.

[0307] Therefore, for example, the user can first grasp the general location of the desired access destination by receiving information from the attribute information transmitting unit 23. Next, the user can specifically identify the location of the desired access destination by receiving information from the individual information transmitting unit 23. In this way, the user can easily identify the location of the desired access destination without relying on vision.

[0308] 21 may be configured with a plurality of partial output units each capable of outputting an output signal, similar to the output unit 33 shown in Fig. 20. It may also be possible to set which of the plurality of partial output units outputs the output signal.

[0309] In this modification, the access destination is a product placed on a product display shelf, but the access destination is not limited to a product. The access destination may be, for example, the position of a toilet flush button or lever, or an elevator button. The user can also hear a sound indicating the presence of a toilet flush button, lever, elevator button, etc. through the proximity area output signal.

[0310] Furthermore, in this modified example, a configuration has been exemplified in which switching among the left approach area information acquisition state, the right approach area information acquisition state, and the proximity area information acquisition state is triggered by a conscious operation by the user. In this case, the receiving device 3 may be provided with a user interface such as a button for state switching operated by the user. The frequency downconversion unit 32 switches to the state among the left approach area information acquisition state, the right approach area information acquisition state, and the proximity area information acquisition state in accordance with the operation performed on the user interface.

[0311] The trigger for switching the states is not limited to this. The receiving device 3 may be equipped with a motion detection sensor that detects a specific motion of the user, such as an outstretched hand. The frequency downconversion unit 32 may switch between the left approach area information acquisition state, the right approach area information acquisition state, and the proximity area information acquisition state when the specific motion is detected by the motion detection sensor.

[0312] The receiving device 3 may also include a state-switching radio wave receiving unit that receives state-switching radio waves instructing the state to be switched. In this case, upon receiving the state-switching radio wave, the frequency downconverter 32 switches to the state instructed by the state-switching radio wave, among the left approach area information acquisition state, the right approach area information acquisition state, and the proximity area information acquisition state. In this way, the state may be automatically switched using the state-switching radio wave provided from an external source.

[0313] In addition, in this modification, the fingertip partial receiving unit 31c attached to the fingers of the hand is exemplified as the proximity area partial receiving unit attached to the extremity of the user's body, but the proximity area partial receiving unit is not limited to this. The proximity area partial receiving unit may also be attached to the toes of the user's feet, which are extremities of the user's body.

[0314] In addition, in this specification, the concept of an "end part" of the body also includes the end parts of prosthetic limbs such as artificial legs and arms, artificial fingers, etc. For example, for a user who has lost a forearm, the elbow may be considered an end part.

[0315] The above describes the embodiments and variations 1-9, but the present invention is not limited to the embodiments and variations 1-9, and various modifications and substitutions can be made within the scope that does not deviate from the gist of the present invention.

[0316] In addition, a program for realizing all or part of the functions of the transmitting device 2 and receiving device 3 according to the embodiment may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be loaded into a computer system and executed to perform all or part of the processing performed by the transmitting device 2 and receiving device 3. This program includes the frequency up-conversion program 26 and the frequency down-conversion program 36 shown in FIG. 1. Note that the term "computer system" as used herein includes not only hardware such as an operating system (OS) and peripheral devices, but also a World Wide Web (WWW) system equipped with a homepage provision environment (or display environment). Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs (Read Only Memory), and CD-ROMs (Compact Disc Read Only Memory), as well as storage devices such as hard disks built into computer systems. Furthermore, the term "computer-readable recording medium" also refers to storage devices that retain a program for a certain period of time, such as volatile memory (RAM (Random Access Memory)) within a computer system that serves as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. However, the recording medium is preferably a computer-readable non-transitory recording medium.

[0317] The program may also be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by transmission waves in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has the function of transmitting information, such as a network such as the Internet or a communication line such as a telephone line. The program may also be a program that realizes part of the aforementioned functions. Furthermore, the program may be a so-called differential file (differential program) that can realize the aforementioned functions in combination with a program already stored in the computer system.

[0318] The communication system 1 according to the embodiment can be used to provide audio notification information to people in dangerous locations such as station platforms, railroad crossings, and construction sites.

[0319] In this case, it is preferable to install the functions of the receiving device 3 in advance in a mobile terminal, which is a portable computer such as a smartphone or tablet. A speaker constituting the transmitting unit 23 of the transmitting device 2 is installed in a dangerous location, and a transmission signal carrying the above-mentioned notification information is broadcast from the speaker. When a person carrying the above-mentioned mobile terminal approaches a dangerous location, the transmission signal broadcast from the speaker is received by the mobile terminal. The person carrying the above-mentioned mobile terminal hears the above-mentioned notification information as audible sound through the speaker provided in the mobile terminal or earphones or headphones connected to the mobile terminal. Note that the speaker, earphones, headphones, etc. are preferably of a bone conduction type that transmits audible sound to the person's auditory nerves through the person's bones.

[0320] Therefore, a person carrying the mobile device can recognize through their hearing that they are approaching a dangerous location. Therefore, with the above configuration, the notification information can be transmitted to visually impaired people as well. Furthermore, even visually impaired people may be distracted by operating the mobile device and may not be able to notice that a dangerous location is nearby due to their vision. Even in such cases, with the above configuration, the notification information can be transmitted to the person through their hearing.

[0321] The wavelength of the ultrasonic wave used as the transmission signal is shorter than that of the radio waves generally used for communication. For example, the wavelength of a 3 GHz radio wave is approximately 100 mm, while the wavelength of a 30 kHz ultrasonic wave is approximately 11 mm. Therefore, compared to radio waves, ultrasonic waves are less likely to be diffracted and propagate with higher directionality. Therefore, a person carrying the mobile device can easily determine the direction of a dangerous location relative to their own location by observing changes in the intensity of the audible sound that carries the alarm information.

[0322] The above-described configuration can also be used to convey various types of information to people in public facilities such as train stations, public spaces, fast food restaurants, cafes, etc. For example, by installing a speaker as the transmitter 23 of the transmitter 2 according to the embodiment at the entrance of a store, a visually impaired person or the like can find the location of the store entrance without assistance from a store clerk or the like.

[0323] The present invention can be modified in various ways without departing from the broad spirit and scope of the present invention. The above-described embodiments and modifications are intended to explain the present invention and do not limit the scope of the present invention. The scope of the present invention is defined not by the embodiments and modifications but by the claims. Various modifications made within the scope of the claims and the meaning of the invention equivalent thereto are considered to be within the scope of the present invention.

[0324] This application is based on Japanese Patent Application No. 2024-092430, filed on June 6, 2024. The entire specification, claims, and drawings of Japanese Patent Application No. 2024-092430 are incorporated herein by reference.

[0325] 1...communication system, 2...transmitting device, 3...receiving device, 21...acquisition unit, 22...frequency up conversion unit, 23...transmitting unit, 24...processor, 25...storage unit, 26...frequency up conversion program, 31...receiving unit, 31a...partial receiving unit for left half body (partial receiving unit, partial receiving unit for approach area), 31b...partial receiving unit for right half body (partial receiving unit, partial receiving unit for approach area), 31c...partial receiving unit for fingertips (partial receiving unit, partial receiving unit for proximity area), 32...frequency down conversion unit, 33...output unit, 33a...partial output unit for left ear, 33b...partial output unit for right ear, 34...processor, 35...storage unit, 36...frequency down conversion program.

Claims

a frequency upconversion unit that shifts the frequency of the audio signal from a first band to a second band that is higher than the first band by using a lapped orthogonal transform that uses a window function for extraction that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; a transmitter that transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band using the frequency upconversion unit; A transmitting device comprising: 　 The frequency upconversion unit a signal extracted from the audio signal using the extraction window function is transformed into a first coefficient sequence in the frequency domain, a second coefficient sequence is obtained from the first coefficient sequence by a frequency up-transform that transfers a coefficient sequence subset of the first band of the first coefficient sequence to a coefficient sequence subset of the second band, and the obtained second coefficient sequence is converted back into a time domain signal by the inverse transform, thereby shifting the frequency of the audio signal from the first band to the second band; The transmitting device according to claim 1 . 　 The frequency upconversion unit transforming the signal extracted from the audio signal by the extraction window function into the first coefficient sequence using a discrete Fourier transform or a discrete cosine transform; obtaining the second coefficient sequence from the first coefficient sequence using a conversion equation that represents a relationship between coefficients constituting the first coefficient sequence and coefficients constituting the second coefficient sequence, the conversion equation being selected depending on the number of shifts of the coefficients constituting the first coefficient sequence on the frequency axis for realizing the frequency upconversion and the number of the coefficient in the first coefficient sequence; converting the second sequence of coefficients obtained back into the time domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform; The transmitting device according to claim 2 . 　 The frequency upconversion unit When the shift number represents an odd shift, phase inversion is performed on the extracted signals before and after the shift number. The transmitting device according to claim 3 . 　 The frequency upconversion unit multiplying the time-domain signal restored from the second coefficient sequence by a reconstruction window function that satisfies the Princen-Bradley condition, and connecting the signals multiplied by the reconstruction window function with overlapping, thereby obtaining the transmission signal; The transmitting device according to claim 2 . 　 The frequency upconversion unit a coefficient sequence obtained by performing the frequency upconversion on the first coefficient sequence or the first coefficient sequence as a coefficient sequence to be replaced, and performing a replacement operation on the replacement-target coefficient sequence to replace the order of coefficients constituting the replacement-target coefficient sequence on a frequency axis; the result of performing the swapping operation and the frequency upconversion on the first coefficient sequence is the second coefficient sequence; The transmitting device according to any one of claims 2 to 5. 　 The frequency upconversion unit performing a resolution enhancement process in which the number of data points of the second coefficient sequence obtained during the frequency upconversion is increased to be greater than the number of data points of the first coefficient sequence, thereby obtaining the transmission signal having a temporal resolution higher than that of the audio signal; The transmitting device according to any one of claims 2 to 5. 　 an acquisition unit that acquires the audio signal; Furthermore, the transmitting unit transmits the transmission signal formed by the frequency upconversion unit while the acquiring unit acquires the audio signal; The transmitting device according to any one of claims 1 to 5. 　 a receiving unit that receives a transmission signal whose frequency belongs to a second band; a frequency down-conversion unit that shifts the frequency of the transmission signal from the second band to a first band that is lower than the second band by using a lapped orthogonal transform that uses a clipping window function that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; an output unit that outputs an output signal obtained by shifting the frequency of the transmission signal to the first band by the frequency down-conversion unit; A receiving device comprising: 　 The frequency down converter a frequency down-transformation of the third coefficient sequence, which transforms the signal extracted from the transmission signal using the extraction window function into a third coefficient sequence in the frequency domain; a frequency down-transformation of the third coefficient sequence, which transfers a coefficient sequence subset of the second band of the third coefficient sequence to a coefficient sequence subset of the first band, to obtain a fourth coefficient sequence from the third coefficient sequence; and a frequency shift of the fourth coefficient sequence, which is converted back into a time domain signal by the inverse transform, thereby shifting the frequency of the transmission signal from the second band to the first band.

10. The receiving device according to claim 9. 　 The frequency down converter transforming the signal extracted from the transmission signal by the extraction window function into the third coefficient sequence using a discrete Fourier transform or a discrete cosine transform; obtaining the fourth coefficient sequence from the third coefficient sequence using a conversion equation that represents a relationship between the coefficients constituting the third coefficient sequence and the coefficients constituting the fourth coefficient sequence, the conversion equation being selected depending on the number of shifts of the coefficients constituting the third coefficient sequence on the frequency axis for realizing the frequency down-conversion and the number of the coefficient in the third coefficient sequence; converting the obtained fourth coefficient sequence back into the time domain signal using an inverse discrete Fourier transform or an inverse discrete cosine transform; The receiving device according to claim 10. 　 The frequency down converter When the shift number represents an odd shift, phase inversion is performed on the extracted signals before and after the shift number.

12. The receiving device according to claim 11. 　 The frequency down converter multiplying the time-domain signal returned from the fourth coefficient sequence by a reconstruction window function that satisfies the Princen-Bradley condition, and connecting the signals multiplied by the reconstruction window function with overlapping to obtain the output signal; The receiving device according to claim 10. 　 The transmission signal is The replacement target coefficient sequence in the frequency domain is obtained by performing a replacement operation to replace the order of coefficients constituting the replacement target coefficient sequence on the frequency axis, The frequency down converter a coefficient sequence obtained by performing the frequency down-conversion on the third coefficient sequence or the third coefficient sequence as a coefficient sequence to be inversely permuted, and performing an inverse permutation operation on the coefficient sequence to be inversely permuted, the inverse permutation operation representing an inverse mapping of the permutation operation; the result of performing the inverse permutation operation and the frequency down-conversion on the third coefficient sequence is the fourth coefficient sequence; 14. A receiving device according to any one of claims 10 to 13. 　 the output signal formed by the frequency down-conversion unit is output by the output unit while the transmission signal is received by the receiving unit; Receiver according to any one of claims 9 to 13. 　 the greater the power of the transmission signal received by the receiving unit, the greater the power of the output signal output by the output unit; Receiver according to any one of claims 9 to 13. 　 The receiving unit a left-side body partial receiving unit attached to the left side of the user's body; a right-side body partial receiving unit attached to the right side of the user's body; and The frequency down converter a left ear output signal that is the output signal obtained by shifting the frequency of the transmission signal received by the left half body partial receiving unit to the first band; a right-ear output signal that is the output signal obtained by shifting the frequency of the transmission signal received by the right-side body partial receiving unit to the first band; and Forming The output unit a left ear partial output unit that outputs the left ear output signal to the left ear of the user; a right-ear partial output unit that outputs the right-ear output signal to the right ear of the user; 17. The receiving device of claim 16, comprising: 　 The receiving unit a plurality of partial receiving units each having a function of receiving the transmission signal and arranged at different parts of the user's body; and The frequency down converter a function of switching which of the plurality of partial receiving units uses the transmission signal received by the partial receiving unit to form the output signal to be output to the output unit; 17. The receiving device of claim 16, comprising: 　 The plurality of partial receiving units include: a proximity area partial receiving unit attached to a distal end of the user's body; an approach area partial receiving unit attached to a position on the user's body closer to the head than the distal end; 20. The receiving device of claim 18, comprising: a transmitting device that uses a lapped orthogonal transform that uses a window function for extraction that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform to shift the frequency of an audio signal from a first band to a second band different from the first band, and transmits a transmission signal obtained by shifting the frequency of the audio signal to the second band; a receiving device that receives the transmission signal transmitted by the transmitting device, and converts the frequency of the transmission signal from the second band back to the first band using the lapped orthogonal transform that uses the extraction window function and an inverse transform of the lapped orthogonal transform, and outputs an output signal obtained by converting the frequency of the transmission signal back to the first band; A communication system comprising: Shifting the frequency of the audio signal from a first band to a second band higher than the first band using a lapped orthogonal transform using a window function for extraction that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; transmitting a transmission signal obtained by shifting the frequency of the audio signal to the second band; Sending method. 　 receiving a transmission signal whose frequency belongs to a second band; shifting the frequency of the transmission signal from the second band to a first band lower than the second band using a lapped orthogonal transform using a window function for extraction that satisfies the Princen-Bradley condition and an inverse transform of the lapped orthogonal transform; outputting an output signal obtained by shifting the frequency of the transmission signal to the first band; Receiving method.

Images