JP2017201351A - Wireless device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce noise components included in a transmitted signal of a wireless device by a simple process.SOLUTION: A transceiver comprises a sound collection part 10 for collecting sounds and generating a sound collection signal, a noise data generation part 22 for generating noise data on the basis of the sound collection signal, and a transmission unit 28 for transmitting the sound collection signal as a wireless signal. The transmission unit 28 is switched on the basis of a user operation between a transmission on state in which a wireless signal is transmitted and a transmission off state in which no wireless signal is transmitted. The sound collection part 10 generates sound frames as sound collection signals successively along with the passage of time. The noise data generation unit 22 generates noise data on the basis of a prescribed number of sound frames dating back to the past from the latest sound frame that were generated when the transmission unit 28 was in the off state. The transmission unit 28 reduces noise components included in the wireless signal on the basis of the noise data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a radio apparatus, and more particularly to processing for noise components included in a transmission signal.

  Transceivers are widely used as means for communicating between remote points. The transceiver collects sound with a microphone and transmits the audio signal to the other transceiver. Further, the transceiver receives a radio signal transmitted from the counterpart transceiver, and outputs a sound included in the received signal from the speaker.

  When the surroundings of the transceiver are noisy, the sound output from the speaker may be buried in noise and difficult to hear. In addition, noise is superimposed on the voice of the sender and collected by the microphone. In the transceiver on the receiver side, noise is output from the speaker in addition to the voice, making it difficult to hear the sound emitted from the speaker of the receiver transceiver. is there.

  As a technique for solving such a problem, Patent Document 1 describes a technique for outputting a sound for canceling noise from a speaker. An apparatus using this technique collects noise by a microphone to generate a noise signal, and outputs a sound based on a signal obtained by combining a cancellation signal in which the noise signal has a reverse polarity and an audio signal, from a speaker. Patent Document 2 describes a technique for reducing a noise component included in a transmission signal when voice is transmitted by a communication device such as a mobile phone. An apparatus using this technology selects a frequency spectrum close to the frequency spectrum in the silent section of the audio signal as a noise frequency spectrum from a plurality of types of frequency spectra prepared in advance. And the noise component of an audio | voice signal is reduced using the frequency spectrum of noise.

JP 2008-244852 A JP-A-6-282297

  As described in Patent Document 2, when noise components included in a transmission signal are reduced using a noise frequency spectrum, it is necessary to prepare a plurality of frequency spectra in advance, and the process for obtaining noise is complicated. It may become.

  An object of the present invention is to reduce noise components included in a transmission signal of a wireless device by simple processing.

  The present invention includes a sound collection unit that collects sound and generates a sound collection signal, a noise data generation unit that generates noise data based on the sound collection signal, and a transmission unit that transmits the sound collection signal as a radio signal; The transmission unit is switched to a transmission-on state in which the radio signal is transmitted or a transmission-off state in which the radio signal is not transmitted based on a user operation. Sequentially, an audio frame is generated as the collected sound signal, and the noise data generation unit is a predetermined number of the audio frames retroactively from the latest audio frame, and the transmission unit is in the transmission off state. The noise data is generated based on the generated voice frame, and the transmission unit generates a noise component included in the radio signal based on the noise data. Characterized in that it reduced.

  In addition, the present invention provides a sound collection unit that collects sound and generates a sound collection signal, a noise data generation unit that generates noise data based on the sound collection signal, and a transmission that transmits the sound collection signal as a radio signal The transmission unit is switched to a transmission on state for transmitting the radio signal or a transmission off state for not transmitting the radio signal based on a user operation, and the noise data generation unit is The noise data is generated based on the sound collection signal when the transmission unit is first in the transmission off state and the sound collection signal when the transmission unit is next in the transmission off state. The transmission unit reduces a noise component included in the radio signal based on the noise data.

  Preferably, the noise data generation unit generates the noise data based on the collected sound signals in the transmission off state over a plurality of times when the transmission off state and the transmission on state are alternately repeated. .

  Preferably, the sound collection unit sequentially generates sound frames as the sound collection signal as time passes, and the noise data generation unit includes a predetermined number of the sound frames retroactively from the latest sound frame. The noise data is generated based on the audio frame generated when the transmission unit is in the transmission off state.

  Preferably, a reception unit that receives a radio signal and reproduces audio based on the radio signal is provided, and the noise data generation unit is based on the sound collection signal when the magnitude of the audio signal does not exceed a predetermined value. The noise data is generated, and the receiving unit executes a process for improving a voice-to-noise ratio for the reproduced voice based on the noise data.

  According to the present invention, the noise component included in the transmission signal of the wireless device can be reduced by simple processing.

It is a figure which shows the structure of the transceiver which concerns on embodiment of this invention. It is a figure explaining the timing which a control part controls a selection part. It is a figure which shows notionally the process which calculates | requires noise spectrum data. It is a figure which shows an initial stage spectrum, a noise spectrum, and an emphasis spectrum. It is a figure which shows the frequency spectrum of a sound collection signal, a noise spectrum, and the frequency spectrum of a transmission signal. It is a figure which shows the time waveform of a sound-collection signal and a received audio | voice level value, and the state of a transmission switch and a selection part.

  FIG. 1 shows the configuration of a transceiver according to an embodiment of the present invention. This transceiver is a simplex transceiver. A push button type transmission switch 58 is provided, and while the transmission switch 58 is pressed, the transceiver converts the voice collected from the microphone 12 into a radio signal and transmits it to the counterpart transceiver. While the transmission switch 58 is released, the transceiver receives the radio signal transmitted from the counterpart transceiver, and outputs the sound included in the received signal from the speaker 52.

  The user speaks toward the microphone 12 while pressing the transmission switch 58. Further, while releasing the transmission switch 58, the user listens to the voice of the other party emitted from the speaker 52. The user who continues the conversation repeats the operation of pressing the transmission switch 58 to speak, releasing the transmission switch 58 and listening to the other party's voice. In the transceiver, a transmission state (transmission on state) in which the transmission switch 58 is pressed and a radio signal is transmitted and a reception state (transmission off state) in which the transmission switch 58 is released and a radio signal is received are alternately repeated. It is.

  The configuration of the transceiver will be described. The transceiver includes a sound collection unit 10, a selection unit 20, a noise data generation unit 22, a transmission unit 28, a reception unit 38, a reception detection unit 60, a transmission switch 58, and a control unit 56. The sound collection unit 10 includes a microphone 12, an amplification unit 14, a sound encoder 16, and a sound collection FFT unit 18. FFT is an abbreviation for Fast Fourier Transform.

  The microphone 12 outputs a sound collection signal corresponding to the voice of the user and the noise around the transceiver to the amplification unit 14. The amplifying unit 14 amplifies the collected sound signal and outputs it to the audio encoder 16. The voice encoder 16 converts the collected sound signal into a digital signal and outputs it to the collected sound FFT unit 18. The digital signal output from the audio encoder 16 is a signal in which frames having a predetermined time length are continuous on the time axis. Each frame includes a plurality of audio samples that are continuous on the time axis. Each audio sample represents a signal level by a digital value. Some standards for digitally encoding speech include a frame time length of 20 msec. A sound collection FFT unit 18 as a frequency analysis unit performs fast Fourier transform processing on the sound collection signal to generate frequency spectrum data. This fast Fourier transform process converts one frame into a set of frequency spectrum data. One set of frequency spectrum data represents one frequency spectrum. The frequency spectrum is obtained by associating a complex frequency component value with each frequency on the frequency axis. The sound collection FFT unit 18 sequentially outputs the frequency spectrum data of the sound collection signal to the selection unit 20 for each set as time passes.

  Note that the sound collection FFT unit 18 may execute a process of converting a predetermined plurality of frames into a set of frequency spectrum data. In this case, each time a predetermined plurality of frames are input to the sound collection FFT unit 18, the sound collection FFT unit 18 obtains one set of frequency spectrum data.

  The selection unit 20 includes a basic terminal B, a transmission terminal T, an open terminal O, and a noise collection terminal N. The selection unit 20 selects any one of the transmission terminal T, the open terminal O, and the noise collection terminal N according to the control of the control unit 56, and connects the selected terminal and the backbone terminal B.

  Here, the configuration and operation of the transmission unit 28 will be described assuming that the basic terminal B is connected to the transmission terminal T. The transmission unit 28 includes a subtracter 30, a transmission side IFFT unit 32, and a transmission unit 34. IFFT is an abbreviation for Inverse Fast Fourier Transform. The frequency spectrum data of the sound collection signal is output from the sound collection unit 10 to the subtracter 30 via the selection unit 20. In addition, the storage unit 26 included in the noise data generation unit 22 stores noise spectrum data indicating the frequency spectrum of noise around the transceiver by processing described later. This noise spectrum data is read from the storage unit 26 to the subtractor 30.

  The subtractor 30 and the transmission side IFFT unit 32 constitute a noise reduction unit 36, which reduces a noise component included in the frequency spectrum data of the collected sound signal based on the noise spectrum data and generates a transmission signal. That is, the subtractor 30 generates noise reduction spectrum data by subtracting each frequency component value indicated by the noise spectrum data from each frequency component value indicated by the frequency spectrum data of the collected sound signal, and outputs the noise reduction spectrum data to the transmission side IFFT unit 32. To do. The process of reducing the noise component is performed for each set of frequency spectrum data. The transmission-side IFFT unit 32 performs fast Fourier inverse transform processing on one set of noise reduction spectrum data to generate one frame. That is, the transmission-side IFFT unit 32 performs fast Fourier inverse transform processing on the noise reduction spectrum data for each set to generate a time-domain transmission signal composed of frames that are continuous on the time axis. Output to. The transmission unit 34 converts the transmission signal into a radio signal and transmits it.

  When the sound collection FFT unit 18 executes a process of converting a plurality of predetermined plurality of frames into a set of frequency spectrum data, the transmission-side IFFT unit 32 performs the processing for the noise reduction spectrum data of one set. A predetermined plurality of frames are generated.

  Next, the receiving unit 38 will be described. The reception unit 38 includes a reception unit 40, a reception-side FFT unit 42, a spectrum enhancement unit 44, a reception-side IFFT unit 46, an audio decoder 48, an amplification unit 50, and a speaker 52. The receiving unit 40 receives a radio signal transmitted from the counterpart transceiver. The receiving unit 40 extracts a digital audio signal included in the radio signal, and outputs the received audio signal to the reception detecting unit 60 and the reception-side FFT unit 42. The received audio signal is composed of frames that are continuous on the time axis, like the transmission signal.

  The reception-side FFT unit 42, the spectrum enhancement unit 44, and the reception-side IFFT unit 46 perform the following enhancement processing on the received audio signal based on the noise spectrum data stored in the storage unit 26.

  The reception-side FFT unit 42 performs a fast Fourier transform process on the received audio signal to generate frequency spectrum data. This fast Fourier transform process is to transform one frame into a set of frequency spectrum data, similar to the fast Fourier transform process executed by the sound collection FFT unit 18. The reception-side FFT unit 42 sequentially outputs the frequency spectrum data to the spectrum enhancement unit 44 for each set as time elapses. Noise spectrum data is read from the storage unit 26 to the spectrum emphasizing unit 44.

  Note that the reception-side FFT unit 42 may execute processing for converting a predetermined plurality of frames into a set of frequency spectrum data. In this case, every time a plurality of predetermined frames are input to the reception-side FFT unit 42, the reception-side FFT unit 42 obtains one set of frequency spectrum data.

  The spectrum enhancement unit 44 obtains a frequency weighting function based on the noise spectrum data. The frequency weighting function is a function that associates a weighting coefficient with a frequency. For example, the greater the absolute value of the frequency component value indicated by the noise spectrum data, the greater the weighting coefficient. The weighting coefficient for each frequency may be binarized. For example, the weighting coefficient is set to a value greater than 1 for frequencies where the absolute value of the frequency component value of the noise spectrum data exceeds a predetermined value, and the weighting coefficient is set to 1 for frequencies where the absolute value of the frequency component value does not exceed the predetermined value. Or a value of 1 or less exceeding 0.

  The spectrum emphasizing unit 44 generates emphasized spectrum data by multiplying each frequency component value (complex number) indicated by the frequency spectrum data of the received audio signal by a weighting coefficient obtained from the frequency weighting function, and outputs the emphasized spectrum data to the receiving-side IFFT unit 46. .

  The receiving-side IFFT unit 46 performs fast Fourier inverse transform processing on one set of emphasized spectrum data to generate one frame. In other words, the receiving IFFT unit 46 performs fast Fourier inverse transform processing on the emphasized spectrum data, generates a reception audio signal in the time domain constituted by frames continuous on the time axis, and outputs the received audio signal to the audio decoder 48. This received audio signal is a digital signal obtained by performing enhancement processing on the received audio signal extracted from the radio signal.

  When the reception-side FFT unit 42 executes processing for converting a predetermined plurality of frames into a set of frequency spectrum data, the reception-side IFFT unit 46 applies the predetermined plurality of frames to the set of emphasized spectrum data. Generate a frame.

  The audio decoder 48, the amplification unit 50, and the speaker 52 constitute a reproduction unit 54, which reproduces the received audio signal as audio. That is, the audio decoder 48 converts the received audio signal output from the reception-side IFFT unit 46 into an analog signal and outputs the analog signal to the amplifying unit 50. The amplifying unit 50 amplifies the received audio signal converted into an analog signal and outputs the amplified signal to the speaker 52. The speaker 52 outputs a sound corresponding to the received sound signal. The above-described enhancement processing improves the ratio of voice to noise heard by the user (speech-to-noise ratio).

  On the other hand, the reception detection unit 60 outputs a reception sound level value indicating a sound level indicated by the reception sound signal output from the reception unit 40 to the control unit 56. The control unit 56 controls the selection unit 20 based on the state of the transmission switch 58 and the received voice level value as follows. When the control unit 56 detects that the transmission switch 58 has been pressed, the control unit 56 controls the selection unit 20 to connect the basic terminal B to the transmission terminal T. On the other hand, when the control unit 56 detects that the transmission switch 58 has been released, the control unit 56 determines whether or not the reception voice level value output from the reception detection unit 60 exceeds a predetermined threshold value. When the control unit 56 determines that the received voice level value does not exceed the predetermined threshold value, the control unit 56 controls the selection unit 20 to connect the basic terminal B to the noise collection terminal N. In addition, when the control unit 56 detects that the transmission switch 58 is released and further determines that the received voice level value exceeds a predetermined threshold value, the control unit 56 controls the selection unit 20 to B is connected to the open terminal O.

  According to such control, when the transmission switch 58 is pressed, the trunk terminal B is connected to the transmission terminal T, and the transceiver is in a transmission-on state. When the transmission switch 58 is released, the main terminal B is connected to the noise collection terminal N or the open terminal O according to the received voice level value. That is, when the transmission switch 58 is released and the reception voice level value does not exceed the predetermined threshold value, the main terminal B is connected to the noise collection terminal N, and the transmission switch 58 is released and the reception voice level value becomes the predetermined threshold value. Is exceeded, the basic terminal B is connected to the open terminal O.

  FIG. 2 is a diagram illustrating the timing at which the control unit 56 controls the selection unit 20. FIG. 2A shows whether the transmission switch is pressed or released in association with the time axis. FIG. 2B conceptually shows the time waveform of the received voice level value. FIG. 2C shows which one of the transmission terminal T, the open terminal O, and the noise collection terminal N is selected in the selection unit 20. The period during which the transmission switch is pressed is the transmission period Tt in which the transceiver is in the transmission on state. At this time, the trunk terminal B is connected to the transmission terminal T, and the collected sound signal is transmitted by a radio signal. During the period when the transmission switch is released, the transceiver is in a reception state (transmission off state). In the reception state, the reception voice level value does not exceed the predetermined threshold A, the reception quiet state in which the basic terminal B is connected to the noise collection terminal N, and the reception voice level value exceeds the predetermined threshold A. There is a reception listening state in which the terminal B is connected to the open terminal O. The reception period includes a reception quiet period Tn in which the transceiver is in a reception quiet state and a reception listening period To in which the transceiver is in a reception listening state. In the example shown in FIG. 2C, each reception period in which the transmission switch is released includes these periods in the order of reception quiet period Tn, reception listening period To, reception quiet period Tn. ing.

  Next, the configuration and operation of the noise data generation unit 22 will be described with reference to FIG. The noise data generation unit 22 includes a moving average calculation unit 24 and a storage unit 26. While the main terminal B is connected to the noise collection terminal N, the frequency spectrum data of the sound collection signal is sequentially output from the noise collection terminal N to the moving average calculation unit 24 for each set as time passes.

  The moving average calculation unit 24 stores a predetermined number K sets (K frames) of frequency spectrum data retroactively including the latest set of frequency spectrum data. The moving average calculation unit 24 may delete older frequency spectrum data, or may not use it for processing. In this case, the spectrum set number K, which is a positive integer, may be a value such that frequency spectrum data can be obtained from the reception quiet period that is at least one before the latest reception quiet period.

  The transmission quiet period changes according to the operation of the transmission switch 58 by the user and the speaking time of the other party. For example, the longer the time for the user to release the transmission switch 58 and the shorter the time the other party is speaking, the longer the transmission quiet period. Therefore, the number of spectrum sets K may be determined based on an average transmission quiet period when the user has a conversation.

  The moving average calculation unit 24 obtains a noise spectrum by averaging the K frequency spectra indicated by the K sets of frequency spectrum data stored in this manner. That is, the moving average calculation unit 24 obtains an average value of K frequency component values indicated by K set frequency spectrum data for each frequency on the frequency axis, and obtains the average value obtained for each frequency as noise spectrum data. Each frequency component indicated by The moving average calculation unit 24 stores the noise spectrum data in the storage unit 26. The moving average calculation unit 24 newly obtains noise spectrum data each time a set of frequency spectrum data is output from the selection unit 20 and stores the noise spectrum data in the storage unit 26. The storage unit 26 may store the latest noise spectrum data and delete the past noise spectrum data.

  FIG. 3 conceptually shows a process for obtaining the noise spectrum data S2 based on the collected sound signals output from the speech encoder 16 in the two reception quiet periods Tn-1 and Tn-2. Each of the reception quiet periods Tn-1 and Tn-2 includes a plurality of frames F consecutive on the time axis. The sound collection FFT unit 18 generates frequency spectrum data S1 for each frame included in the reception quiet periods Tn-1 and Tn-2, and generates K sets of frequency spectrum data S1. The moving average calculator 24 averages the K frequency spectra indicated by the K sets of frequency spectrum data S1 to obtain noise spectrum data S2.

  Note that the moving average calculation unit 24 may store a plurality of L sets of frequency spectrum data obtained during the M reception quiet periods retroactive to the past including the latest reception quiet period. Here, the reception quiet period number M is an arbitrary integer of 2 or more. This storage processing is performed as follows, for example. The control unit 56 outputs timing information indicating the timing of the reception quiet period to the noise data generation unit 22. The timing information is information indicating, for example, 1 in the reception quiet period and 0 in a period other than the reception quiet period. The moving average calculation unit 24 acquires the frequency spectrum data output from the selection unit 20 and recognizes the timing of each reception quiet period based on the timing information output from the control unit 56. As a result, for frequency spectrum data, L sets obtained during the M reception quiet periods retroactive to the past including the latest reception quiet period are stored, and older frequency spectrum data is not deleted or used for processing. To.

  The moving average calculator 24 averages the L sets of frequency spectrum data to obtain a noise spectrum. Here, the positive integer L changes according to the user's operation of the transmission switch 58 and the utterance time of the other party. For example, the positive integer L increases as the time for the user to release the transmission switch 58 is longer and the time for which the other party is speaking is shorter. When the positive integer L reaches the predetermined upper limit value P, the moving average calculating unit 24 averages the frequency spectrum data of the upper limit value P set retroactively including the latest set of frequency spectrum data. To obtain the noise spectrum.

  The number K of spectrum sets or the number M of reception quiet periods may be determined according to the nature of noise. That is, the time length of the collected signal used for obtaining the noise spectrum data may be determined according to the nature of the noise. In general, the frequency spectrum of noise varies with time. For example, in a noise in which a high sound and a low sound appear alternately such as an ambulance or a police car siren, the frequency spectrum changes in a cycle in which the sound level changes. For noise having such a dynamic frequency spectrum, the noise reduction unit 36 and the spectrum enhancement unit 44 are made by making the time length of the collected signal used for obtaining the noise spectrum data longer than the period in which the frequency spectrum changes. The effect of is often increased. That is, the effect of reducing noise by the noise reduction unit 36 and the effect of improving the voice-to-noise ratio by the spectrum enhancement unit 44 are often increased. In the example of the siren described above, the number of spectrum sets K or the number of reception quiet periods M is determined so that the length of the collected sound signal used for obtaining the noise spectrum data is longer than the period in which the siren sound level changes. As a result, the noise reduction effect and the improvement effect of the voice-to-noise ratio are often increased.

  A specific operation at the time of transmitting and receiving by the transceiver will be described. The user presses the transmission switch 58 when transmitting, and the user releases the transmission switch 58 when receiving a call. When the control unit 56 detects that the transmission switch 58 has been released, the control unit 56 determines whether or not the reception voice level value output from the reception detection unit 60 exceeds a predetermined threshold value.

  When the control unit 56 determines that the received voice level value does not exceed the predetermined threshold value, the control unit 56 controls the selection unit 20 to connect the basic terminal B to the noise collection terminal N. If the user is not speaking at this time, the microphone 12 outputs a sound collection signal corresponding to the noise around the transceiver. As a result, the sound collection unit 10 sequentially outputs frequency spectrum data corresponding to the noise to the moving average calculation unit 24 for each set. The moving average calculation unit 24 obtains noise spectrum data and stores it in the storage unit 26.

  On the other hand, when the control unit 56 detects that the transmission switch 58 has been released, and further determines that the received voice level value exceeds the predetermined threshold, the control unit 56 controls the selection unit 20 to B is connected to the open terminal O. Further, the moving average calculation unit 24 is controlled to stop the update of the noise spectrum data, and the noise spectrum data stored in the storage unit 26 at that time is held as the latest.

  The receiving unit 38 extracts a received voice signal from the received radio signal, performs enhancement processing on the received voice signal based on the noise spectrum data stored in the storage unit 26, and further performs enhancement processing. The voice based on the received voice signal is output from the speaker 52.

  As described above, when the transmission switch 58 is released and the received voice level value does not exceed the predetermined threshold value, the noise spectrum data is sequentially updated. When the transmission switch 58 is released and the received voice level value exceeds a predetermined threshold value, a voice based on the received voice signal that has been enhanced by the noise spectrum data is output from the speaker 52.

  FIG. 4A conceptually shows the frequency spectrum (initial spectrum 68) of the received audio signal and the noise spectrum 70 before the enhancement processing is performed. FIG. 4B conceptually shows the frequency spectrum (enhanced spectrum 72) of the received audio signal after the enhancement process and the noise spectrum 70. The horizontal axis indicates the frequency, and the vertical axis indicates the absolute value of the frequency component value. The emphasized spectrum 72 is obtained by multiplying the initial spectrum 68 by a weighting coefficient based on the noise spectrum 70. The weighting coefficient has a larger value as the noise spectrum 70 is larger. Therefore, as shown in FIG. 4B, the emphasized spectrum 72 increases as the noise increases, and the sound reaching the user from the speaker 52 increases in accordance with the noise spectrum. This improves the noise-to-noise ratio and makes it easier to hear the sound output from the speaker 52 even when the surroundings of the transceiver are noisy.

  The control unit 56 turns off the power supplied to some or all of the electronic devices constituting the reception unit 38 while the transmission switch 58 is pressed, and the transmission switch 58 is released. At the same time, a process of turning on the power supply supplied to the electronic device whose power supply has been turned off may be executed. As a result, the power consumed by the receiving unit 38 during transmission is reduced.

  When the transmission switch 58 is released and the reception voice level value does not exceed the predetermined threshold value, no sound is output from the speaker 52, or even if the sound is output from the speaker 52, the size is very small. While the transmission switch 58 is released, the user usually listens to the sound output from the speaker 52 and therefore does not speak. Therefore, while the transmission switch 58 is released and the reception voice level value does not exceed the predetermined threshold, the microphone 12 is highly likely to output a sound collection signal corresponding to the noise around the transceiver. Therefore, there is a high possibility that noise spectrum data corresponding to noise around the transceiver is stored in the storage unit 26.

  On the other hand, when the transmission switch 58 is released and the received voice level value exceeds a predetermined threshold, the speaker 52 often outputs a voice that can be heard by the user. Therefore, there is a high possibility that the microphone 12 collects the sound output from the speaker 52. Therefore, the basic terminal B of the selection unit 20 is connected to the open terminal O, and the frequency spectrum data output from the sound collection unit 10 is not used. And the emphasis process with respect to a received audio | voice signal is performed using the noise spectrum data already memorize | stored in the memory | storage part 26. FIG.

  Here, the reception detection unit 60 is assumed to detect a reception voice signal, but the reception detection unit 60 may be configured to detect a radio signal received by the reception unit 40. In this case, a radio signal level value indicating the level of the radio signal is output from the reception detection unit 60, and the radio signal level value is used for control instead of the received audio level value.

  Next, transmission by the transceiver will be described. When the control unit 56 detects that the transmission switch 58 has been pressed, the control unit 56 controls the selection unit 20 to connect the basic terminal B to the transmission terminal T.

  The microphone 12 collects the user's voice and noise around the transceiver, and outputs a sound collection signal corresponding to the user's voice and noise. The sound collection unit 10 sequentially outputs the frequency spectrum data of the sound collection signal to the transmission unit 28 via the selection unit 20 for each set. The transmission unit 28 generates and wirelessly transmits a transmission signal with reduced noise components based on the frequency spectrum data corresponding to the collected sound signal.

  Thus, when the transmission switch 58 is pressed, the frequency spectrum data corresponding to the sound collection signal is generated by the sound collection unit 10. Further, a transmission signal obtained by reducing noise components from the frequency spectrum data of the collected sound signal is wirelessly transmitted from the transmission unit 28.

  FIG. 5A shows a frequency spectrum 62 and a noise spectrum 64 of the sound collection signal. The horizontal axis indicates the frequency, and the vertical axis indicates the absolute value of the frequency component value. FIG. 5B shows the frequency spectrum 66 of the transmission signal. The noise spectrum 64 is subtracted from the frequency spectrum 66 of the transmission signal, and the noise component is suppressed.

  The wirelessly transmitted transmission signal is received by the other party's transceiver, which is the receiving side, and audio based on the audio signal included in the transmission signal is output from the speaker. The transmitting transceiver transmits a transmission signal with reduced noise components, so even if the surroundings of the transmitting transceiver are noisy, you can listen to the audio output from the speaker of the receiving transceiver. Easy to take.

  Note that while the transmission switch 58 is released, the control unit 56 turns off the power supplied to some or all of the electronic devices constituting the transmission unit 28, and the transmission switch 58 is pressed. At the same time, a process of turning on the power of the electronic device that has been turned off may be executed. Thereby, the power consumed by the transmission unit 28 during reception is reduced.

  An example of the operation of the transceiver will be described with reference to FIGS. FIG. 6A conceptually shows a time waveform of a sound collection signal output from the microphone 12 of the sound collection unit 10. FIG. 6B conceptually shows the time waveform of the received voice level value. FIG. 6C shows a timing chart indicating whether the transmission switch 58 is pressed or released, and FIG. 6D shows the transmission terminal T and the release in the selection unit 20. A timing chart showing which one of the terminal O and the noise collecting terminal N is selected is shown.

  From time t0 to time t1, from time t2 to time t3, and after time t4, the transmission switch 58 is released, and the received voice level value does not exceed the threshold A. This period is a reception quiet period. Therefore, the selection unit 20 selects the noise collection terminal N, and the noise spectrum data is sequentially updated and stored in the storage unit 26. During this time, if the user is not speaking, the noise spectrum data is likely to indicate noise.

  The transmission switch 58 is pressed from time t1 to time t2. A sound collection signal corresponding to the user's voice is output from the microphone 12. This period is a transmission period. Therefore, the selection unit 20 selects the transmission terminal T, and the transmission unit 28 wirelessly transmits a transmission signal based on the collected sound signal.

  From time t3 to time t4, the transmission switch 58 is released, and the received voice level value exceeds the threshold A. This period is a reception listening period. Therefore, the selection unit 20 selects the open terminal O, and a sound based on the received sound signal subjected to the enhancement process is output from the speaker 52.

  According to the transceiver according to the present embodiment, the timing of generating noise spectrum data is controlled according to the level of the voice signal transmitted from the counterpart transceiver and the operation of the transmission switch by the user. Control for generating data is simplified.

  In addition, the noise reduction unit 36 can reduce noise by appropriately setting the number K of spectrum sets or the number M of reception quiet periods and adjusting the time length of the collected sound signal used for obtaining noise spectrum data. And the effect of improving the noise-to-noise ratio by the spectrum emphasizing unit 44 is often increased. That is, in many cases, these effects are enhanced by setting the time length of the collected signal used for obtaining the noise spectrum data to be longer than the period in which the noise frequency spectrum changes.

  Note that the sound collection FFT unit 18, the selection unit 20, the moving average calculation unit 24, the subtractor 30, the transmission side IFFT unit 32, the reception side FFT unit 42, the spectrum enhancement unit 44, and the reception side that configure the transceiver according to the present embodiment. The IFFT unit 46 and the control unit 56 process digital signals. These components may be configured by a processor that executes predetermined arithmetic processing. Also, some or all of these components may be individually configured by digital circuits that are hardware.

  Further, in the transceiver according to the present embodiment, the counterpart transceiver may be a duplex type transceiver. In general, in the duplex system, radio transmission from one transceiver to the other transceiver and radio transmission from the other transceiver to the one transceiver are simultaneously performed in different frequency bands. When the counterpart transceiver is a duplex system, the transmission frequency of the transmitter 34 may be set to the reception frequency of the counterpart transceiver, and the reception frequency of the receiver 40 may be set to the transmission frequency of the counterpart transceiver. Although the counterpart's transceiver is a duplex system, transmission and reception are performed in the same manner as a simplex transceiver.

DESCRIPTION OF SYMBOLS 10 Sound collection part, 12 Microphone, 14, 50 Amplification part, 16 Voice encoder, 18 Sound collection FFT part, 20 Selection part, 30 Subtractor, 32 Transmission side IFFT part, 34 Transmission part, 36 Noise reduction part, 38 Reception unit , 40 receiving unit, 42 receiving side FFT unit, 44 spectrum enhancement unit, 46 receiving side IFFT unit, 48 audio decoder, 52 speaker, 54 playback unit, 56 control unit, 58 transmission switch, 60 reception detecting unit, 62 sound collecting Frequency spectrum of signal, 64 noise spectrum, 66 Frequency spectrum of transmission signal, 68 Initial spectrum, 70 Noise spectrum, 72 Weighted spectrum.

Claims (5)

A sound collection unit that collects sound and generates a sound collection signal;
A noise data generation unit that generates noise data based on the collected sound signal;
A transmission unit that transmits the collected sound signal as a radio signal,
The transmitting unit is
Based on the user's operation, it is switched to a transmission on state for transmitting the radio signal, or a transmission off state for not transmitting the radio signal,
The sound collector is
A sound frame is generated as the sound collection signal sequentially over time,
The noise data generator is
A predetermined number of the audio frames retroactive from the latest audio frame, and generating the noise data based on the audio frames generated when the transmission unit is in the transmission off state;
The transmitting unit is
A radio apparatus that reduces a noise component included in the radio signal based on the noise data. A sound collection unit that collects sound and generates a sound collection signal;
A noise data generation unit that generates noise data based on the collected sound signal;
A transmission unit that transmits the collected sound signal as a radio signal,
The transmitting unit is
Based on the user's operation, it is switched to a transmission on state for transmitting the radio signal, or a transmission off state for not transmitting the radio signal,
The noise data generator is
The noise data is generated based on the sound collection signal when the transmission unit is first in the transmission off state and the sound collection signal when the transmission unit is next in the transmission off state. ,
The transmitting unit is
A radio apparatus that reduces a noise component included in the radio signal based on the noise data. The wireless device according to claim 2, wherein
The noise data generator is
When the transmission off state and the transmission on state are alternately repeated, the radio apparatus generates the noise data based on each sound collection signal in the transmission off state over a plurality of times. The wireless device according to claim 2 or claim 3,
The sound collection unit sequentially generates sound frames as the sound collection signal as time passes,
The noise data generator is
Generating the noise data based on the predetermined number of the audio frames retroactive from the latest audio frame, the audio frames generated when the transmission unit is in the transmission off state. A wireless device characterized. The radio apparatus according to any one of claims 1 to 4,
A receiving unit that receives a radio signal and reproduces sound based on the radio signal;
The noise data generator is
Generating the noise data based on the collected sound signal when the magnitude of the audio signal does not exceed a predetermined value;
The receiving unit is
A wireless device that executes a process for improving a voice-to-noise ratio of a reproduced voice based on the noise data.

Images