JP4346501B2 - Receiver - Google Patents

Description

  The present invention relates to a receiving apparatus, and more particularly to a receiving apparatus suitable for detecting and removing a voice reception signal.

  With recent increase in communication demand, wireless communication devices are rapidly spreading. In order to maintain order and effectively use radio waves in such a situation, it is necessary to use each radio station under certain conditions. However, it is not possible to say that all wireless communication devices are operated in a condition due to a malfunction of the wireless device or illegal operation. If these are left unattended, there is a risk that the operation of a wireless device that is normally operated may be disturbed. Therefore, it is important to monitor the use of radio waves and prevent the occurrence of abnormal radio waves. For this purpose, it is necessary to estimate the specifications of radio waves transmitted from each radio station, and signal processing apparatuses for that purpose have been developed.

  Signals used in wireless communication are classified into analog modulation signals and digital modulation signals. Examples of the analog modulation signal include AM (Amplitude Modulation) and FM (Frequency Modulation), and examples of the digital modulation signal include ASK (Amplitude Shift Keying) and FSK (Frequency Shift Keying). As a specific method of these modulation signals, for example, the received signal is distributed to the I and Q components, the phase difference between each time is detected from the phase at each time from the I and Q components, and the level according to the preset modulation method is used. There has been proposed a method of extracting features depending on whether or not they match the phase difference (see, for example, Patent Document 1).

  As described above, there is a demand for further improvement in accuracy of a signal processing device for radio wave monitoring.

  An increase in the amount of signal processing is inevitable for high-accuracy signal analysis. Therefore, it is desired to remove signals other than the observation target by preprocessing of signal analysis so as not to perform useless signal processing. In addition, it is desirable that the signal processing itself for finding a signal that is not observed is as simple and highly accurate as possible.

As a matter of course, the radio wave band is assigned to operate in all signal formats all over the world. Therefore, it is natural that analog signals and digital signals are mixed, and in a radio wave monitoring apparatus used in such an environment, it is inevitable that interference of signals outside the observation target will interfere to some extent.
JP 2003-262670 A

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a receiving apparatus capable of detecting and removing an analog audio signal with high accuracy.

According to one aspect of the present invention, an envelope extraction unit that extracts an envelope by speech spectrum envelope estimation from a signal that has received a predetermined radio band and converted into a baseband, and an envelope extracted by the envelope extraction unit based on, if the received signal is determined and the audio signal determination means for determining whether the analog modulated audio signal or a digitally modulated audio signal, and this sound signal determining means is an analog modulated audio signals, There is provided a receiving apparatus comprising audio signal removing means for removing a signal determined as an analog-modulated audio signal from the received signal.

  According to another aspect of the present invention, an envelope extraction unit that extracts an envelope from a signal obtained by receiving a predetermined radio band and converted into a baseband by speech spectrum envelope estimation, and the envelope extraction unit A sound signal determination means for determining whether or not the received signal is a sound signal based on the extracted envelope, performs a sound / silence determination of the received signal, and performs the sound spectrum only for a time estimated as a sound part. A correlation value between an envelope and a signal obtained by converting the received signal into a frequency domain is calculated, and if the correlation value is smaller than a predetermined threshold, the voice signal is determined to be removed, and the voice signal is removed. A receiving device is provided.

  ADVANTAGE OF THE INVENTION According to this invention, the audio | voice signal in a received signal can be discriminate | determined accurately, and the receiver which can analyze only a digital signal efficiently is provided.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the drawings, the same portions are denoted by the same reference numerals, and redundant description is omitted.

  (First Embodiment) First, the first embodiment of the present invention will be described.

  FIG. 1 is a block diagram showing a configuration of a receiving apparatus according to the first embodiment of the present invention. In FIG. 1, the received signal received by the antenna is input to the RF unit 1001. The RF unit 1001 performs signal amplification and frequency selection. The signal processed by the RF unit 1001 is input to a frequency conversion unit 1002 including, for example, an A / D converter. The frequency converter 1002 converts the input signal into a baseband signal. The signal processed by the frequency conversion unit 1002 is sent to the audio signal detection unit 1003 and the audio signal removal unit 1004. The audio signal detection unit 1003 determines whether or not the received signal is an audio signal, and inputs the determination result to the audio signal removal unit 1004. The audio signal removing unit 1004 removes the audio signal from the received signal according to the audio signal determination result input from the audio signal detecting unit 1003. The signal from which the audio signal is removed by the audio signal removal unit 1004 is supplied to the demodulation / analysis unit 1005. The demodulating / analyzing unit 1005 demodulates and analyzes the input signal from which the audio signal has been removed, and then sends it to the display unit 1006. The display unit 1006 is configured by, for example, a monitor or a printer, and is displayed with the detection level corresponding to the frequency band. Note that the display unit 1006 may display the modulation method, the band, or the demodulated signal itself. For example, in the input unit 1007 configured by a keyboard, parameters such as RF frequency or filter switching are set, and the set value is sent to the control unit 1008. The control unit 1008 controls the RF unit 1001, the frequency conversion unit 1002, and the demodulation / analysis unit 1005 based on the set value.

  Next, the configuration of the audio signal detection unit 1003 will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the audio signal detection unit 1003. In FIG. 2, the baseband signal output from the frequency conversion unit 1002 is input to a speech spectrum envelope estimation unit 2001 and a fast Fourier transform (FFT) calculation unit 2003, respectively. A speech spectrum envelope estimation unit 2001 estimates a spectrum envelope from a received signal using a filter that simulates a feature of speech, and calculates and outputs an envelope while shifting a predetermined FFT window size and window start time. The FFT calculation unit 2003 performs frequency conversion of the received signal, that is, conversion from the time axis to the frequency axis. Similar to the speech spectrum envelope estimation unit 2001, the FFT calculation unit 2003 outputs a calculation result while shifting the predetermined FFT window size and the window start time. The calculation result of the speech spectrum envelope estimation unit 2001 is input to a formant (maximum value) detection unit 2002 for extracting a formant, which is a spectral feature amount of a speech signal, and the position of a fundamental frequency called a speech formant is calculated. The detected position is output to the correlation value calculation unit 2004. The correlation value calculation unit 2004 calculates a correlation value between the speech spectrum envelope and the signal frequency-converted by the FFT calculation unit 2003 at the position of the fundamental frequency obtained by the formant (maximum value) detection unit 2002. The calculation result in the correlation value calculation unit 2004 is supplied to the audio signal determination unit 2005. The audio signal determination unit 2005 determines whether or not the input reception signal is audio, and outputs the result to the audio signal removal unit 1004.

  Next, the configuration of the speech spectrum envelope estimation unit 2001 will be described with reference to FIG. FIG. 3 is a block diagram illustrating a configuration of the speech spectrum envelope estimation unit 2001. In the embodiment shown in FIG. 3, the spectral envelope is extracted by the mel cepstrum method. Spectral envelope extraction is not limited to this. Linear Predictive Coding (LPC) analysis that extracts the parameters by modeling the spectral density function of the signal using an all-pole filter, and the acoustic properties of speech. Of course, the cepstrum method which is the spectrum analysis used may be used.

  The baseband signal input from the frequency conversion unit 1002 is frequency converted by the FFT calculation unit 3001. The frequency-converted signal is supplied to the power spectrum calculation unit 3002 and converted from waveform data to a power spectrum. Further, the signal frequency-converted by the FFT calculation unit 3001 is supplied to the evaluation function minimizing unit 3005. The signal converted into the power spectrum by the power spectrum calculation unit 3002 is given to the frequency conversion unit 3003 and converted on the frequency axis. The signal converted on the frequency axis is supplied to an inverse fast Fourier transform (IFFT) calculation unit 3004. IFFT calculation section 3004 performs inverse fast Fourier transform to collectively convert frequency component digital modulation data into a time signal. The signal converted into the time signal is sent to the evaluation function minimizing unit 3005. Evaluation function minimizing section 3005 performs correction so that the error between the signal obtained by frequency-converting the baseband signal by FFT calculation section 3001 and the signal converted from the time signal from IFFT calculation section 3004 is minimized. . This corrected signal is supplied to the mel cepstrum logarithmic conversion unit 3006 to perform conversion that increases the characteristics of the speech.

  FIG. 4 conceptually shows the operation of extracting the spectral envelope from the audio signal and obtaining the correlation value. In FIG. 4, the vertical axis represents the magnitude of the power spectrum, and the horizontal axis represents the frequency. The calculation processing in the correlation value calculation unit 2004 of the audio signal detection unit 1003 is performed by the position of the fundamental frequency obtained by the formant (maximum value) detection unit 2002, that is, the spectrum envelope and the FFT calculation unit 2003 at the formant point shown in FIG. The correlation value of the received signal is calculated from the FFT waveform obtained by frequency conversion in step (1). When the correlation value difference is larger than a predetermined threshold value, that is, when the correlation is not recognized, the baseband signal is determined as an analog-modulated audio signal such as AM or FM. When the difference between the correlation values is smaller than the predetermined threshold, it is determined as a digitally modulated audio signal.

  In this way, by detecting the audio signal with high accuracy and removing it from the received signal, it is possible to perform efficient reception without performing unnecessary analysis.

  (Second Embodiment) Next, a second embodiment of the present invention will be described with reference to FIG. In the first embodiment, the correlation value of the received signal is calculated from the speech spectrum envelope and the FFT waveform obtained by frequency conversion by the FFT calculation unit. If the correlation value is calculated in the silent part, the voice feature may not be extracted, and an incorrect result may be output. Therefore, as shown in FIG. 5, the zero-order mel cepstrum coefficient m (0) used in the mel cepstrum is output from the mel cepstrum logarithmic conversion unit 3006 and supplied to the correlation value calculation unit 2004 in the audio signal detection unit 1003. Here, the zero-order mel cepstrum coefficient m (0) is an average value for all frequencies of the spectral envelope estimated by the mel cepstrum method. Since the zeroth-order mel cepstrum coefficient m (0) indicates the power of the spectrum, it is discriminated whether it is sound or silent by this value. FIG. 6 shows a method for detecting sound and silence. If the 0th-order mel cepstrum coefficient m (0) is greater than or equal to a predetermined threshold value, it is determined that there is sound. For example, in FIG. 6, the shaded area is determined to be sound. Conversely, if the 0th-order mel cepstrum coefficient m (0) is smaller than a predetermined threshold, it is determined that there is no sound. In the second embodiment, the correlation value calculation unit 2004 correlates only the sounded part. According to such an embodiment, since the correlation value is not calculated in the silent part, the audio signal can be determined with higher accuracy.

  (Third Embodiment) Next, a third embodiment of the present invention will be described with reference to FIG. Generally, in a modulated digital signal, unless the data is all 1 or 0, the distribution of the maximum value of the spectrum in the time direction spreads according to the data pattern. That is, the variance in the time direction has a relatively large value. Similarly, the variance in the time direction taking into account the Nth spectrum (N is an integer 1, 2, 3...) From the maximum value of the spectrum takes the same value as the maximum value variance. Here, FIG. 8 shows an example in which the dispersion at the position where the spectrum of the multilevel FSK signal has the maximum value is measured. The vertical axis represents frequency and the horizontal axis represents time. FIG. 8A is a plot of the first variance value from the maximum value. FIG. 8B is a plot of variance values taking into consideration the maximum value to the eighth value. However, analog voice often has the highest fundamental frequency. That is, the maximum value of the voice spectrum is the fundamental frequency, and the variance in the time direction is small. On the other hand, the dispersion in the time direction taking into account from the maximum value of the spectrum to the Nth is increased by adding a harmonic structure other than the fundamental frequency. Here, FIG. 9 shows an example in which the variance of the position where the spectrum of the audio signal is maximum is measured. The vertical axis represents frequency and the horizontal axis represents time. FIG. 9A is a plot of the first variance value from the maximum value. FIG. 9B is a plot of dispersion values taking into account the maximum value to the eighth value. This is very different from FIG.

  In the third embodiment, the above characteristics are used. FIG. 7 is a block diagram illustrating a specific configuration example of the audio signal detection unit 1003. In FIG. 7, the baseband signal is input to the speech spectrum envelope estimation unit 2001. The output of the speech spectrum envelope estimation unit 2001 is supplied to a Top-m calculation unit 7001 and a Top-n calculation unit 7003, respectively. In the Top-m calculation unit 7001, m is set to 1, for example, and the maximum value of the spectrum is detected. The statistical value calculation unit 7002 calculates the variance of the detected frequency position. The variance value calculated by the statistical value calculation unit 7002 is sent to the audio signal determination unit 7005. Similarly, assuming that n of the Top-n calculation unit 7003 is 8, for example, the position of the frequency including the eighth value from the maximum value is detected. The statistical value calculation unit 7004 calculates the variance of the detected frequency position. The variance value calculated by the statistical value calculation unit 7004 is sent to the audio signal determination unit 7005. The audio signal determination unit 7005 compares the variance of only the maximum value with the variance value taking into consideration the eighth value from the maximum value, and determines that the signal is an audio signal if the variance difference is greater than a preset value.

  Also in the present embodiment, it is possible to accurately determine an audio signal, and it is possible to perform efficient reception without performing unnecessary analysis.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

The block diagram which shows the structure of the system of the receiver which concerns on the 1st Embodiment of this invention. The block diagram which shows the structure of the audio | voice signal detection part of the receiver which concerns on the 1st Embodiment of this invention. The block diagram which shows the structure of the audio | voice spectrum envelope estimation part of the receiver which concerns on the 1st Embodiment of this invention. The figure which shows notionally the operation | movement which extracts a spectrum envelope from an audio | voice signal, and calculates | requires a correlation value. The block diagram which shows the structure of the mel cepstrum 0th order coefficient output part of the receiver which concerns on the 1st Embodiment of this invention. The figure which shows the sound silence detection method notionally. The block diagram which shows the structure of the audio | voice signal detection part of the receiver which concerns on the 1st Embodiment of this invention. The figure which plotted the dispersion | distribution of the position where the spectrum of a multi-value FSK signal becomes the maximum value. The figure which plotted the dispersion | distribution of the position where the spectrum of an audio | voice signal becomes the maximum value.

Explanation of symbols

1001: RF unit, 1002: Frequency conversion unit, 1003: Audio signal detection unit, 1004: Audio signal removal unit,
1005: Demodulation / analysis unit, 1006: Display unit, 1007: Input unit, 1008: Control unit, 2001: Speech spectrum envelope estimation unit, 2002: Formant (maximum value) detection unit, 2003: FFT calculation unit, 2004: Correlation Value calculation unit, 2005: Audio signal determination unit, 3001: FFT calculation unit, 3002: Power spectrum calculation unit, 3004: IFFT calculation unit, 3005: Evaluation function minimization unit, 3006: Mel cepstrum logarithmic conversion unit

Claims (5)

An envelope extraction means for extracting an envelope from a signal received from a predetermined radio band and converted into a baseband by speech spectrum envelope estimation;
Audio signal determination means for determining whether the received signal is an analog-modulated audio signal or a digital-modulated audio signal based on the envelope extracted by the envelope extraction means;
When the audio signal determination unit determines that the audio signal is analog-modulated, the audio signal determination unit includes an audio signal removal unit that removes a signal determined to be an analog-modulated audio signal from the received signal. Receiver device.   The sound signal determination means calculates a correlation value between the sound spectrum envelope and a signal obtained by converting the received signal into a frequency domain, and determines that the sound signal is a sound signal if the correlation value is smaller than a predetermined threshold. The receiving apparatus according to claim 1, wherein an audio signal is removed.   The receiving apparatus according to claim 1 or 2, wherein a cepstrum method is used for speech spectrum envelope estimation in said envelope extraction means.   The sound signal determination means performs sound / silence determination of the received signal, and calculates a correlation value between the sound spectrum envelope and a signal obtained by converting the received signal into a frequency domain only for a time estimated as a sound part. The receiving apparatus according to claim 2, wherein: An envelope extraction unit that extracts an envelope from a signal obtained by receiving a predetermined radio band and converted into a baseband by speech spectrum envelope estimation;
A peak detector that outputs the frequency of the maximum value from the largest of the maximum values of the envelope extracted by the envelope extraction unit; and
A signal processing unit for obtaining a dispersion value of the signal output by the peak detection unit;
A difference detection unit for obtaining a difference between variance values obtained by the signal processing unit based on at least two different sample numbers;
A comparison unit that compares the difference signal output by the difference detection unit with a predetermined threshold; a determination unit that determines whether the received signal is an audio signal based on a comparison result of the comparison unit;
A receiving apparatus comprising: an audio signal removing unit that removes a signal determined to be an audio signal from the received signal when the determining unit determines that the signal is an audio signal.

Images