JPS6142699A - Secret talk apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] (Technical Field) The present invention relates to a confidential communication device, particularly a C8M (Composite
e8Inusoidal Modeling: Based on the principle of compound sine wave model), it relates to an S stuffing device.

(Prior art) Generally used as a secret communication device, the C1 audio signal is converted into a digital code by A/D conversion, and the original audio information is transmitted in a confidential manner by performing code conversion, etc. Digital secret speech devices are widely used that reproduce the original audio signal through D/A conversion after performing this inverse code conversion.

However, such digital systems have the disadvantage that high requirements are placed on the transmission quality of the transmission path, such as transmission capacity and error rate.

On the other hand, there are analog secret communication devices that use methods such as inverting the spectrum of the audio signal, or dividing it and transmitting it by changing the relative position, but this generally affects the transmission quality of the transmission path. Although the requirements are low,
Since the spectral envelope of the original audio signal remains in the form of a trace, it generally has the disadvantage of low privacy.

(Object of the Invention) An object of the present invention is to provide a novel analog secret speech device with high speech privacy based on the principles of speech analysis and speech synthesis using CSM, which will be described later.

(Structure of the Invention) The device of the present invention includes a transmitting side having a CSM spectrum generating means for generating an additively synthesized waveform of a plurality of sine waves having a frequency and amplitude specified by CSM analysis of an input audio signal. , spectral parameter extraction means for performing spectrum analysis of the signal from the 03M spectrum generation means on the transmitting side and extracting each parameter specifying the frequency and image width of each of the plurality of sine waves that have been added and synthesized; A plurality of variable frequency oscillators with a phase reset function are set for each frequency base specified by the parameter, and correspondingly, the output of each variable frequency oscillator with a phase reset function is set to the amplitude specified by the extracted parameter. a variable gain amplifier, a variable length window function generator, and a random number generator, and when synthesizing voiced sound, resets the phase of each variable frequency oscillator with a phase reset function in accordance with the pitch period, When synthesizing unvoiced sound, the phase of each of the variable frequency oscillators with a phase reset function is reset in accordance with the period calculated from the random number output from the random number generator, and the frequency generated by the variable length window function generator is reset. and a receiving side in which the start and end points of the window function substantially coincide with the phase reset points.

(K. Theory) As mentioned above, the uninvented secret speech device is based on the principles of CSM speech analysis and sound synthesis.

First, we will explain the principles of CS MO9 voice analysis and speech synthesis.

C8M expresses an audio signal as a sum of a specific number of sine waves having amplitude and frequency as freely transparent parameters. A predetermined number of 4 to 6 sine waves is used at most.

Therefore, when performing CSM speech synthesis, it is first necessary to express the speech signal as a sum of a predetermined number of sine waves by CSM speech analysis.

CSM speech analysis will be explained in detail later, and only the main points will be explained here. Similarly to the case of LPC analysis, in the @Ca1 company analysis,
An autocorrelation coefficient is used as an intermediate paratard for the purpose of ignoring phase information, averaging shadows of sound sources, and reducing instability due to noise components.

That is, in C8M analysis, N low-order taps of the sample autocorrelation coefficient directly calculated from the speech waveform to be expressed for each analysis frame are combined into one low-order tap of the autocorrelation coefficient of the composite wave. The purpose is to determine the frequency of each sine wave to be synthesized and its intensity (power loss@) so that it matches the N sine waves.

Now let the number of sine waves to be synthesized be n, the angular frequency of each sine wave be ω1 (''=1 + 2 t...n), and the intensity of each sine wave be mi, then the CSM synthesis yt multiplied by However, the autocorrelation number 173 of this tap can be easily expressed using ωi and mH, and 171 = Σmi C01 ω1.

On the other hand, if the sample of the audio waveform to be expressed is Xt, the tap l (D sample autocorrelation coefficient Vl in a certain frame is given as.

However, M is the number of samples in one analysis frame.

Now, in the CSM analysis, the value of each mi and ωi is determined so that VJ given the above-mentioned γj is equal to the given VJ for N low-order values.

That is, rl=V1: However, J=0, 1, 2. . . . The purpose is to determine the values of mi and mi so that N holds true.

This specific method will be explained in detail later.
Here, it is assumed that mi and mi of the above-mentioned n sine waves are successively obtained for each analysis frame in response to a given audio signal.

An example of the audio feature vector pattern based on the C8M paratotas mi and mi obtained in this way is shown in FIG.

Also, 9th order (N=9) CAM (number of sine waves n=5) analyzed with the window length of the analysis frame as 307FL8EC.
9 obtained from the line spectrum and the same audio sample
An example of correspondence with the following LPC spec) k-hull Wt (frequency transmission characteristics of an LPG synthesis filter) is shown in FIG.

Note that there is a relationship of N=21-1 between the above-mentioned order N and the number n of sine waves, as described later.

From these figures, it can be seen that the CAM contains information extracted from the IR speech of the original voice to be expressed.

However, the n4 obtained as a result of C8M analysis
Using the values of mi and mi of i, create n sine waves with the intensity (the actual amplitude is on the side as described above) and angular frequency specified by ff1j and mi, and simply add and synthesize these. If this is done, the human ear can only hear the sound as a synthesized sine wave, and cannot hear it as the original sound at all.

This means that even if sine waves are simply added, the spectrum of the generated signal is only a discretized n line spectrum, whereas the spectrum of an audio signal has a continuous spectrum with an envelope. Furthermore, voiced sounds are expressed by a pitch structure, and unvoiced sounds have a fine spectral structure expressed by a stochastic process, and this is due to the fact that the spectral spread is completely different from that of a simply added CSM and a speech signal. Conceivable.

Therefore, in order to synthesize audible speech using CSM, it is necessary to spread the line spectrum into a continuous spectrum using some special method.

In other words, C8M speech synthesis is the sound 4I! expressed by a line spectrum as shown in FIG. 1, 2rIA! The purpose is to spread the IF vector pattern in a special way and generate the audio spectrum pattern from it. If such csM speech synthesis is not performed, C
A simple summation composite waveform of multiple sine waves with the frequency and amplitude specified by AM contains the information necessary to reproduce the original composition in its most basic form, but it does not produce any audio at all. It is impossible to hear it as such.

The secret speech device of the present invention utilizes the above-mentioned % ridges of the simple addition composite waveform of CAM.

That is, the transmitting side performs CAM analysis of the input audio signal, generates an analog signal of a simple addition composite waveform of a plurality of sine waves having the angular frequency and amplitude specified by the CAM, and sends this to the transmission path. As mentioned above, although this synthesized waveform contains the information necessary to reproduce the voice in its most basic form, it is highly confidential in that it cannot be heard as voice at all. . Furthermore, if particularly necessary, the confidentiality can be further improved by applying a predetermined specific conversion to the parameters of the CSM, as will be described in detail later.

Now, on the receiving side, special CAM voice synthesis is performed to reproduce the original voice. The method of C8M speech synthesis using spread spectrum on the reception side of the confidential communication device of the present invention is as follows.

That is, since a voiced sound has a pitch structure such as 1811wi, the phase of each of the CSM0n sine waves specified by the above-mentioned method is reset for each pitch period. This makes it possible to easily generate a spectral envelope and a fine spectral structure of the pitch.

Furthermore, by applying the special time window function processing described in detail in the explanation of the embodiment to the above-mentioned phase reset waveform, the discontinuity of the synthesized waveform at the time of phase reset can be removed and the continuity of the audio waveform can be improved. It is secured.

As a result of the above implementation, the CSM line spectrum shown in Figure 2 is diffused as shown in Figure 3, and changed to a spectrum having a spectral envelope and pitch fine structure, which is sufficiently practical for auditory purposes. Experimental results have shown that durable sound quality can be obtained.

For reference, FIG. 3(c) shows the spectrum of a CAM obtained by simple addition without performing the above-mentioned processing. This is the spectrum of the signal transmitted via the transmission path, and a waveform with such a spectrum can only be heard aurally as a synthesized sound of sine waves, and cannot reproduce the original sound.

The above is for voiced sounds, but in the case of unvoiced sounds, it is performed as follows. In other words, in the case of voiced sounds mentioned above, the phase reset and special time window function processing performed for each pitch period are performed, whereas in the case of unvoiced sounds, randomly generated pulses are used as a stochastic process instead of pitch periods. , the above-mentioned processing is performed every time this pulse is generated.

By using the above method, it is possible to perform CAM synthesis that is acoustically sufficient for practical use.

Note that since the above 08M approval is a synthesis method that does not use a filter, it does not require consideration of stability on the synthesis side. Therefore, information on nli and ω1 is transmitted to the synthesis side,
When used in a communication means such as this device that reproduces speech on the synthesis side, when the line quality is relatively poor, it may be possible to obtain better sound quality than a border.

(Example> Next, the present invention will be explained in detail using examples.

The fourth @ is a block diagram showing one embodiment of the present invention.

This embodiment consists of a transmitting side 1 and a receiving side 2.

The transmitting side 1 further includes an A/D converter 101, a Hamming window processor 102 . Autocorrelation coefficient needle side device 103, CAM analyzer 104, pitf, V/Ul-analyzer 105, parameter converter 106, n variable frequency oscillators 107-1 to 10
7-n, n11! variable gain amplifier 108-1”
108-n, addition combiner 109, variable gain amplifier 11
0, variable frequency oscillator 111, V/UV switch 112
, and an addition synthesizer 113.

Further, the receiving side 2 further includes a spectrum analyzer 201, a power component W&
4-n, n variable gain amplifiers 205-1 to 205-n
, addition combiner 206, multiplier 207, multiplier 208, V
/UV switch 209, variable length window function generator 210, period calculator 211, and random number generator 212.

Now, the operation of this embodiment is as follows. The audio waveform to be transmitted is sent to the A/D via input line 1000.
converter, 101, where the amplitude and time axes are converted into quantized digital data, the outputs of which are supplied to a )-mining window processor 102, and a pitch/V
/UV analyzer 105.

The digital data supplied to the Hamming window processor 102 is subjected to weight multiplication using a known Hamming window function for each predetermined frame, and is supplied to the autocorrelation coefficient measuring device 103 for each frame of data. Ru.

The autocorrelation coefficient measuring device 103 calculates the lowest N autocorrelation coefficients Vz (where !=l, 2...
Find N).

In other words, the data for one frame is Xt (however, t”
0w1s .

The set of Vl for each frame obtained in this way is expressed as the following C'
It is supplied to the 3M analyzer 104, and vo therein is supplied to the parameter converter 106 as power information in this frame.

Now, the autocorrelation coefficient V for each frame mentioned above! 90 The SM analyzer 104 specifies the intensity and angular frequency of each of the n sinusoids of the CAM of the corresponding frame by performing operations detailed below.
(however, i" 1 t 2 m...n) is determined,
This is supplied to the parameter converter 106.

In addition, the pitch/V/UV analyzer 105, which receives the digital data of the original audio signal from the A/D converter 101, extracts information regarding the pitch period and voiced sound (V)/unvoiced sound (UV). (Supplied to the rammeter exchanger 106.

Now, the parameter converter 106 is a converter that performs parameter conversion of this information and outputs it, but to make the explanation easier to understand, at first this converter does not perform any conversion and simply inputs the input signal as it is. shall be output.

Thus, the n pieces of angular frequency information ωi obtained by the CAM analyzer 104 are supplied to the variable frequency oscillators 107-1 to 107-(l) via the converter 106, and specify the output angular frequency of each of these oscillators. In addition, n pieces of intensity information mi obtained by the C8M analyzer 104 are set to ωi.
is also connected to the variable gain amplifier 108 via the converter 106.
-1 to 1087rl is supplied as gain control information, and the output of each of the oscillators 107-1 to 107-H is set to have a designated intensity mi.

Thus, the output of the additive synthesizer 109 is a waveform obtained by simple additive synthesis of a plurality of sine waves having the amplitude and frequency specified by C8M.

The above composite waveform is further processed by the variable gain amplifier 110.
The total power is controlled to be proportional to the power vo obtained by the correlator 103 and then supplied to the combiner 113.

On the other hand, the pitch period information obtained by the analyzer 105 is supplied to a variable frequency oscillator 111, and the frequency of this oscillator 111 is controlled so as to become the fundamental frequency of a given pitch. Furthermore, voiced/unvoiced (V/UV
) ff information is supplied to the V/UV sketch 112, which supplies the output of the oscillator 111 to the synthesizer 113 in the case of voiced sound (■), and cuts it off in the case of unvoiced sound (UV). to).

Thus, the transmission loss 12001 connected to the combiner 113 means that the pitch information and the summed composite waveform of each sine wave of the power-controlled CAM are transmitted in the form of an analog signal, but as described above, this signal is It has a confidential nature in that even if it is converted directly into audio, some parts cannot be heard as audio.

Now, on the receiving side 2, the signal transmitted in this way is subjected to spectrum analysis by a spectrum analyzer 201, and each angular frequency information ωi specified by the C8M line spectrum and each intensity information loaded with the total power are mi' is obtained, and the latter is obtained in the power separator 202 as the total power information vo
and intensity information mi of each sine wave of the CSM.Furthermore, pitch period information and V/UV information are extracted in the spectrum analyzer 201, and each of the above information is supplied to a parameter inverse transformer 203. be done.

Now, the parameter inverse converter 203 is a converter that performs the inverse conversion of the parameter converter 106 on the transmitting side described above, but in order to make the explanation easier to understand as described above, first,
Corresponding to the parameter conversion unit 106 on the transmitting side which does not perform any conversion and outputs the input signal as it is, it is assumed that the parameter inverse converter 203 also outputs the input signal as it is.

Thus, the output of spectrum analyzer 201, C3M
ωi (ω1~ωn) specifies the angular frequency of each of the 011 waves
) is added to the frequency control input of the n variable frequency oscillators with phase reset function 204-1 to 204-n,
The output angular frequencies of these oscillators are set to specified angular frequencies ω1 to ωn.

Also, the intensity (1ii, force amplitude) of each n wave of CSM
m1 to mH specifying the n variable gain amplifiers 2
It is supplied to the gain control terminals of 05-1 to 205-n, thereby controlling the oscillation power of each frequency to a specified value.

The n outputs obtained in this manner are subjected to addition and combination in the addition and combination unit 206, and then supplied to the next multiplier 207.

Now, the pitch period data and V/UV information (voiced sound/unvoiced sound information) extracted by the spectrum analyzer 201 are as follows.
V/UV through the parameter inverse converter 203, respectively.
The signal is supplied to the switch 209.

On the other hand, the random numbers generated by the random number generator 212 are supplied to the period calculator 211, where they are converted so that the random number distribution width and its lower limit value become specific values, and the phase reset time interval for unvoiced sounds is determined. It is supplied to the V/UV switch 20-9 as a data string to be determined.

Thus, if the aforementioned V/UV information of the spectrum analyzer 201 specifies a voiced sound (g), the switch 209
selects the aforementioned pitch period data side and supplies it to the variable length window function generator 210.

In addition, if the above-mentioned V/UV information specifies unvoiced sound ([JV), the switch 209 and the above-mentioned period calculator 211
The window function generator 210 selects the data string representing a random time interval generated in the stochastic process of the output of
supply to.

Now, the window function generator 210rt is used to generate a window function that ensures the continuity of the audio waveform by removing discontinuities that occur in the output waveform by the phase reset meter, and also generates a time function that is closely related to this window function. It also generates a phase reset pulse.

As mentioned above, the window function generator 210 includes the switch 209
A data string specifying the interval between the next phase reset pulses is input via the window function generator 21.0, and the window function generator 21.0 successively generates impulses having the time intervals specified by this data. This is connected to variable frequency oscillators 204-1 to 204-H with phase reset function via line 2101.
oscillators, thereby resetting the phases of these oscillators.

Now, the window function generator 210 generates a variable length window function W (z
) occurs.

That is, let T be the value of the phase reset pulse interval at that point specified by the input data,
If the elapsed time since the previous phase reset pulse was generated is X, then W(z) = 0.5 + 0.5 c, os (π〒
) However, a window function expressed as 0 < ay ≦T is generated. This window function W(
Figure 5 (5) shows the construction. The above-mentioned value of T represents the pitch period in the case of voiced sounds, and represents a variable that occurs in a stochastic process in the case of unvoiced sounds, so it changes over time in both cases. Therefore, this □ window function W (:t)
has a variable length, and is synchronized with the generation of the phase reset pulse described above in the relative time relationship shown in Figure 5 (6) (the start and end points of the window function are the same as the generation of the phase reset pulse). (almost coincident with the time).

The window function thus generated is supplied to multiplier 207 via line 2102. As a result, multiplier 207
, the product of n sine waveforms whose phase is reset for each phase reset pulse synthesized by the addition synthesizer 206 and the above-mentioned window function W←) generated in synchronization with each phase reset pulse. is obtained. □The waveform obtained in this way is that each sine wave has a window function W just before the phase is reset.
(s) is continuously converged to O by the multiplication, and each sine wave also rises to O at the time of phase reset, so the continuity of the waveform is ensured, and thus the phase reset waveform is generated by multiplication by the window function WIP). Discontinuities can be removed.

The output of the multiplier 207 from which discontinuities have been removed is supplied to the next multiplier 20B, where it is weighted by the power information o of each frame transmitted from the transmitting side 1, separated by the power separator 202. line 2000 as synthesized voice.
is output from.

As explained above, on the receiving side 2 of this English embodiment, the CSM speech synthesis necessary for the above-mentioned speech synthesis is executed, and as a result, the input to the transmitting side 1, which was hidden on the transmission path 1200, is The original voice will be reproduced with good sound quality.

In the above explanation, it is assumed that the parameter converter 106 on the transmitting side 1 and the parameter inverse converter 203 on the receiving side 2 output the input data as is to the output side, and do not perform any conversion. . Of course, even in this case, as mentioned above, confidentiality on the transmission path is maintained. In other words, the parameter converter 106 in the transmission @l and the parameter inverse converter 20 in the receiving side 2
3 may be completely omitted to form a confidential communication device.

However, in order to obtain a higher degree of confidentiality, the parameter converter 106 performs parameter conversion, and the parameter inverse converter 203 performs inverse conversion, so that each data input to the parameter converter 106 is It is only necessary that the signal be obtained on the output side of the device 203.

As an example of parameter transformation, a simple transformation such as ω = ωi + θ i nl' i := mi

Alternatively, considering the set of 019m1 (i=1, 2...n) as one vector (ωi, mi), and multiplying this by a predetermined constant Ma), the vector (ω 'igrn'i) to set the frequencies of the variable frequency oscillators 107-1 to 107-n and the variable gain amplifiers 108-1 to 108-n.

In addition, as for parameter inverse transformation, the extracted (ω' i , m'i )
It is also possible to restore the original set of vectors (ωi, mi) from the set of vectors by using the inverse matrix of the above-mentioned matrix.

Alternatively, a plurality of corresponding combinations of parameter conversion and parameter inverse conversion may be prepared, and the user may be able to use any combination by inputting data.

Further, it is also possible to further improve confidentiality by changing the parameter inverse transformation corresponding to the parameter transformation described above with time.

Furthermore, in the embodiment described above, the pitch frequency information is transmitted from the transmitting side 1, but instead, the transmission of the Rette frequency information from the transmitting side 1 can be omitted as follows. You can also do it.

In other words, by utilizing the property of voice that when the voice energy is high, the pitch frequency becomes high, and when the voice energy is low, the pitch frequency is low.
[Experimentally create a correspondence table between voice energy and pitch frequency. Then, the receiving side 2 is provided with a pseudo pitch period generating means that generates a pitch period from the total audio power information transmitted from the transmitting side 1 using this correspondence table, and the pseudo pitch period generated thereby is used in the receiving gA2. It is also possible to omit the transmission of the pitch frequency information from the transmitting side.

Next, an example of a variable frequency oscillator 2040 circuit with a phase reset function is shown in FIG. The voltage applied to frequency control terminal 2041 causes constant current flow to flow into sources 2042 and 2043. V: Controls the charging/discharging current value for ft2044, thereby making the oscillation frequency variable. The 7-point oscillation voltage waveform has a triangular shape that linearly rises and falls between +vr and -Vr of the reference voltage. Phase reset terminal 2045
When an impulse is applied to , the 7 points are momentarily grounded,
The potential is forcibly returned to 0, and oscillation is started from there to perform a phase reset. These seven triangular wave oscillation outputs are input to a sine wave converter 2046, converted to a sine wave, and output from a terminal 2047, which is used as the output of the oscillator 204. The sine wave converter 2046 can be easily realized, for example, by reading out a sine function value stored in a ROM as an input waveform.

Further, such a variable frequency oscillator with a phase reset function can be easily realized using a quadrogram of a calculator.

Next, a circuit example of the variable gain amplifier 205 is shown in FIG. By applying a signal to be amplified to a terminal 2051 and a control signal to a terminal 2052, the amount of negative feedback is controlled, and an output having a controlled amplitude is obtained at an output terminal 2053.

In addition to this, it can also be realized by using an analog multiplier, or by using an analog waveform input as the reference voltage of the D/A converter and using control information expressed in digital quantities as the digital input. This can also be easily realized by

Variable gain amplifier ILJ8 can also be realized in a similar manner.

An example of a circuit of the random number generator 212 is shown in FIG. 8 by the arrow. 1
5 stage shift register 2121 and 1 Gyuto a calculator 21
22, a 15th order M-sequence of pseudo-random numbers having a period M of 2ts-1 is generated. Four terminals 21 at the required time
By adding a shift pulse to 23, the next random value is obtained.

Next, a block diagram of the period calculator 211 is shown in FIG. 9(5). In addition to this, the random numbers uniformly distributed in the range of 0 to 211 1 output from the random number generator 211 described above are suitable for use as random numbers that specify the time interval of the phase reset pulse during unvoiced speech. It is made up of a constant multiplier 2.121 and a constant adder 2122. As a result, as shown in the ninth [N(Bl), the random number distribution width and lower limit value can be set to appropriate values.

Next, an example of window function generation @210 is shown in FIG. 101. This includes a register 2101, a presettable down counter 2102. It includes a counter 2103 and a read-only memory (ROM) 2104.

Data T specifying the phase reset pulse interval supplied from the switch 209 is stored in the register 2101. The down counter 2102 is a high-speed clock C with a constant period.
The counter that counts LK, first of all, Regis J210
The content T of 1 is preset and counted down using the clock CLK. The content of the counter 2102 is O
When this happens, a pulse is generated from the output terminal, which presets the contents of the register 2101 again and starts counting down this value. Thus, the down counter 210
A pulse train having a period proportional to T (for example, T/k) is generated at the output 2102-1 of No. 2. This pulse train is added as a clock to counter 2103. The counter output 2103-1 of the counter 2103, which is incremented by four clocks, is added to the ROM 2104 as an address designation signal, and the window function W (3:
) are read out in order and output to /2200. When the contents of the counter 2103 reach k, the ROM 21
The last data of 04 is read and outputs a reset pulse on line 2100. This reset pulse is used as the aforementioned phase reset pulse supplied to the phase reset terminals of the oscillators 204-1 to 204-fl, and is also used to set the last input data in the register 2101. In this way, phase reset pulses whose interpulse intervals have successively specified values and a variable length window function W (s) synchronized with these pulses as shown in FIG. 5(2) are generated.

Finally, CAM analysis will be explained.

As mentioned above, CSM analysis uses, for each analysis frame, N low-order tap values of the sample autocorrelation coefficient directly calculated from the speech waveform to be represented, and a composite wave (n sine waves). The angular frequency ωi of each sine wave to be synthesized and its intensity (power amplitude) are determined so that the N low-order tap values of
The purpose is to determine mi.

Now, if the autocorrelation coefficient of tap l of the composite wave is fl, then
As mentioned above, rJ =Σ Ini Cog 1G)i.

On the other hand, from the sample Xt of the audio waveform to be expressed, the sample autocorrelation coefficient Vl of tap l in a certain frame is.

From this, 11 = VJ... (2) J
=0,1,2. ...-N, where N=2n-1, the following matrix expression is obtained.

However, the above equation cannot be solved by simple matrix operations because ωi and mi are unknown. Therefore, ωi；CO3xl ・・・・・・(4) and cos ノQli=CO8l cos xl= TJ
(:I:i) H6+ −·(5) is substituted. This TJIx) is Tchebychcff'
(Chiebishev) polynomial. When this substitution is performed, equation (3) is converted as follows.

However, in general, l is 'ro(s), 'I's
It can be expressed as a linear combination of (x)...T a (I:).

That is, ゛where j <l) is the inverse T chebycheff (Chebycheff) coefficient.

Using this S (1), the above-mentioned sample self-reciprocity relationship ■“
The linear combination Al is defined as follows.

! However, J=0,1129...12n-1 If you do this, (
By using the relationship of formula (7) and formula (8) on the left side of formula 6), the following relationship is established.

...(9) Now, here, X3. Machi... Define the polynomial for n by taking the zero point pJni.Using this Pn(x), create Ω ΣmI Pn(si)xN, which is similar to the left side of equation (θ), and write this as Try to be frugal. It is clear that the above formula is O, but it can be changed by 4 as shown in the arrow.

From the above, J=0.1, 2. ...The following formula is obtained as n.

Shika, Runi I) (n' = 1 ``ie, so it holds true. The matrix formed by Ai on the left side is generally) 1 a
This is called the nke 1 (Hangul) matrix. As described above, each Ai is given by equation (8) from the sample autocorrelation coefficient Vj of the expressed Bemi speech waveform and is known.

Therefore, by solving equation (10), “ol)e p(
? As a tile of 't "", (χ1 # :rJ!・
., .p''n) are calculated.

From this, each C8M frequency ω1 is ωi C in equation (4)
The C8M intensity mi obtained from O821 is obtained using the following equation derived from equation (9).

Note that the matrix on the left side of the above equation is generally VanclerMon
It is known as a matrix.

To summarize the above, the analysis algorithm for C8M analysis is as follows.

(1) Calculate the five autocorrelation coefficients.

(2) Conversely, AJ is defined using the Chebyshev coefficient.

(3) Find pχn) by solving the Wankel matrix equation by AJ.

(4) Solve an n-dimensional algebraic equation with pi'' as a coefficient to find n :x:i.

Pn")=-"9"', n-1+p(n) gn-
2+, ,.

n-1n-2 p(n), +p. =0 凰(5) Perform cos inverse transformation to obtain C8M angular frequency (ω
Find i).

ωi ; Cog −fl (6) The C8M intensity (mi) is determined by solving the Van der Monde (77) matrix equation.

By executing each step above, each angular frequency (Sushi, ω, ..., ωn) of the CSM and the intensity of each wave (f
nim mte...mH) can be obtained.

Note that, as an efficient method for solving the above-mentioned Hanke 1 (Wankel) matrix equation, a method is known in which initial conditions are given and solutions are sequentially obtained.

Furthermore, since it has been proven that the zero-order algebraic equation has only real roots, the roots can be found using the Neaton-Raphson method or the like.

Furthermore, as an efficient solution method for the above Vander Monde (7y Vander Monde) matrix equation, triangular matrixization and
It is possible to use a method in which the solutions are found sequentially by performing the following steps.

In the CSM analysis described above, in this example,
We used a method of solving an equation that makes the sample autocorrelation coefficient equal to the CAM autocorrelation coefficient, but instead of LP
It is also possible to use a method of calculating the line spectrum frequency by making the C coefficient lossless and calculating the number of tones.

Further, in this embodiment, a variable length window function having a specific function form is used, but this function form is merely an example, and it is clear that other function forms may be used.

Furthermore, the random number generator, period calculator, etc. are also shown as examples, and there is no need to be limited thereto.

In addition, in the above actual ha example, mi was used as the quantity that specifies the intensity (power amplitude) of each C8M sine wave, but as a signal to control the actual variable gain amplifier, a snoring signal proportional to the amplitude is used. It is clear that it may be carried out using

(Effects of the Invention) As described above, by using the present invention, it is possible to provide a new confidential communication device that transmits analog signals based on the principle of speech analysis and synthesis using C8M. It has a relatively low fP requirement for quality, and it is possible to provide high confidentiality.

[Brief explanation of drawings]

゛The first example is a diagram showing an example of the audio feature vector pattern based on the C8M parameter. Figure 2 is a diagram showing an example of the correspondence between the C81vi line spectrum and the nine LPC spectrum envelope obtained from the same audio sample. CA
FIG. 3 (B) is a diagram showing the C8M spectrum obtained by simple addition; FIG. 4 is a block diagram showing an embodiment of the present invention; and the fifth device is a diagram showing the functional form of a variable length window function, Figure 5 (G)
is a diagram showing the relative time relationship between the variable length window function and the phase reset pulse, FIG. 6 is a diagram showing an example of a circuit of a variable frequency oscillator with a phase reset function, and FIG. 7 is an example of a circuit of a variable gain amplifier. Figure 8 is a diagram showing an example of a circuit of a random number generator, Figure 9 is a block diagram of a period calculator, Figure 9 (
2) is a diagram showing a random number analysis distribution of the output of the period calculator, and FIG. 10 is a block diagram showing an example of a variable length window function generator. In the figure, 1... Sending side, 2... Receiving side, 101
. . . Human/D converter, 102 . . . Hamming window processing device, 103 . . . Autocorrelation coefficient measuring device, 104
......C8M analyzer, 105...Pitch/V/UV analyzer, 106...Parameter converter, 107-1 to 107-n...Variable Frequency oscillator, 108-1 to 108-n...-Variable gain amplifier, 109... Addition synthesizer, 110...
...Variable gain amplifier, 111...Variable frequency oscillator, 112...V/UV switch, 113
・Old... Addition combiner, 201... Spectrum analyzer, 2o2... Power separator, 203...
...Parameter inverse converter, 204-1 to 204-n...
...Variable frequency oscillator with phase reset function, 205
-1~205-n ・Old...Variable gain amplifier, 206・
... Addition synthesizer, 207 ... Multiplier, 2
08... Multiplier, 209... V/U
V switcher, 210...Variable length window function generator, 2
11...Period calculator, 212...Random number generator. Note 2 呵诵 ″ to one hour □ Body 5 Concave-Kaya Δ Times $7 Zukyo I Picture Book 7 (A)

Claims (6)

[Claims] (1) A transmitting side having a CSM spectrum generating means for generating an additively synthesized waveform of a plurality of sine waves having a frequency and amplitude specified by CSM analysis of an input audio signal; spectral parameter extraction means for performing spectral analysis of the signal and extracting each paratota specifying the frequency and amplitude of each of the plurality of sine waves summarily synthesized; a variable frequency oscillator with a phase reset function; a variable gain amplifier that correspondingly sets the output of each of the variable frequency oscillators with a phase reset function to an amplitude specified by the extracted parameter; a variable length window function generator; , a random number generator, and when synthesizing voiced sounds, resets the phase of each of the variable frequency oscillators with a phase reset function in accordance with the pitch period, and when synthesizing unvoiced sounds, resets the phase of the variable frequency oscillators, and when synthesizing unvoiced sounds, the random number generator The phase of each variable frequency oscillator with a phase reset function is reset in accordance with the cycle calculated from the random number of the output, and the start and end points of the window function generated by the variable window function generator are set to the phase reset point. 1. A confidential communication device characterized by having a receiving side that is arranged to substantially coincide with a time point. (2) When the pitch information extracted from the input audio signal by the pitch information extraction means provided on the transmitting side is transmitted to the receiving side and voiced sound is synthesized on the receiving side, the transmitted pitch information 2. The confidential communication device according to claim 1, wherein the phase of the variable frequency oscillator with a phase reset function is reset using a pitch period that is more determined. (3) When the receiving side includes a pseudo pitch generating means and the receiving side includes a pseudo pitch generating means and synthesizes voiced sound on the receiving side, the pitch period generated by the pseudo pitch generating means is 2. The confidential communication device according to claim 1, wherein the variable frequency oscillator with a phase reset function is reset by using the variable frequency oscillator. (4) A parameter conversion means provided on the transmitting side performs a specific conversion on the frequency value and amplitude value specified by the CSM analysis, and a plurality of A transmitting side having the CMS spectrum generating means for generating an additive composite waveform of sine waves, and a parameter inverse converting means provided on the receiving side inversely converting the parameters extracted by the spectrum parameter extracting means. A plurality of variable frequency oscillators with a phase reset function are set to frequencies specified by the parameters, and correspondingly, the outputs of the variable frequency oscillators with a phase reset function are set to respective frequencies specified by the inversely converted parameters. 2. The essential call device according to claim 1, further comprising a variable gain amplifier for setting. (5) A plurality of combinations of the parameter conversion means and corresponding parameter inverse conversion means are provided, and an arbitrary combination among these can be selected by data input by an operator. 4) Confidential communication device described in section 4). (6) Item (4) characterized in that it has a plurality of combinations of the parameter conversion means and corresponding parameter inverse conversion means, and the combinations change over time.
) Confidential communication device described in section.