JPS6229234A - Privacy telephone equipment - Google Patents

Abstract

PURPOSE:To improve remarkably the deterioration of sound quality without increasing line capacity by converting an input data into a frequency region data having a frequency band assigned independently and sending the spectrum envelope parameter and residual waveform. CONSTITUTION:An input voice signal is fed to a window processor 102 and a delay circuit 103 via an A/D converter 101. A quantization signal inputted to the window procesesor 102 is subject to band compression into prescribed assigned frequency bands 2-3kHz by a linear converter 107 and fed to an adder 102. The quantization signal inputted to the delay circuit 103 is given to a frequency generator 116, which generates a frequency changed in hte assigned frequency band of 1.6-1.9kHz and the result is fed to the adder 120. On the other hand, a residual waveform is fed to a residual waveform normalizing device 113 is subjected to FM modulation being the content using (1.5-0.4)/2kHz as a carrier by a frequency generator 119 and the modulated wave is sent to the adder 120. The adder 120 adds the spectrum envelope parameter of the inputted frequency region and the data relating to the residual waveform and gives the result to the reception side via a transmission line 1201 as the transmission signal.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a speech encrypting device, and particularly to playback in an analog encrypting device that attempts to polarize an input audio signal by representing it with a spectral envelope parameter and a residual waveform. Concerning improving sound quality.

[Conventional technology]

An analog secret communication device that attempts to conceal the contents of a call by expressing the input audio signal as a spectral envelope parameter and a residual waveform instead of transmitting it as it is from the transmitting side to the receiving side. is becoming widely used in various operational fields.

This type of privacy device has high secrecy and is increasingly being used in applications such as shortwave communications privacy, due to its characteristics such as its ability to ensure communication content that can withstand use even on relatively poor lines.

In such a confidential communication device, the spectral envelope parameter indicating the macroscopic spectral distribution of the input audio signal is expressed as LPC (Linear Prediction Coe.
ffi-cient ) coefficient α parameter, etc., the residual waveform as a microscopic spectral distribution, that is, the sound source is a modeled representation of the sound source with voiced/unvoiced information and pitch period, and this voiced/unvoiced information and pitch period are modeled. The normal operating format is to transmit unvoiced information and pitch period information instead of the sound source to deal with the problem of line transmission capacity.

[Problem that the invention seeks to solve]

However, the above-mentioned conventional secret communication device of this type has the following problems.

In other words, in order to perform confidential communication with the limited transmission capacity of 9 lines, the residual waveform is transmitted representing the pitch period and voiced/unvoiced information, so it is difficult to extract the pitch period and determine voiced/unvoiced. Deterioration of sound quality due to errors, etc. is unavoidable, and if a method of transmitting the residual waveform is used to eliminate this problem, the required line capacity will increase significantly.The purpose of the present invention is as described above. It is an object of the present invention to provide a secret communication device which eliminates the above disadvantages and essentially eliminates deterioration of sound quality without increasing line capacity t'.

[Means for solving problems]

The device of the invention converts an input audio signal into its spectral envelope t
[In a private communication device that decomposes and transmits the current spectral envelope parameters and the residual waveform from which the spectral envelope has been removed, the data is converted into frequency domain data of nine frequency bands that are independently assigned to each other in a preset occupied frequency band.] transmits the spectral envelope parameter and the residual waveform, and in this case, the residual waveform has an imaging width proportional to its maximum value, and each time position corresponds to the time position of each maximum value of the residual waveform. The apparatus includes input audio signal converting means for converting nine signals represented by a plurality of impulse trains.

〔Example〕

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of the transmitting side of an embodiment of the privacy device of the present invention, and FIG. 2 is a block diagram showing the configuration of the receiving side of the privacy device of the invention.

The transmission IIl shown in FIG.
, window treatment device 102. Delay circuit 103. Autocorrelation coefficient calculator 104. LPC analyzer 105. L8P analyzer 106
, linear converter 107. Interpolator 108° Frequency generator 1
09. LPC inverse filter 110. Window processor Ill, maximum value searcher 112. Residual waveform normalizer 113. interpolator 1
14. μ converter 115. Frequency generators 116, 11
/ Lineari converter 117. Residual waveform pulse generator 1181 Frequency generation 6119. It is configured with an adder 120.

9. The reception [2 is A/f) converter 201. shown in FIG. Slide 4y DFT (Discrete
Fou-rier Transform) circuit 202
．． Delay circuit 203. Peak picking circuit 204. Interpolator 205. Linear inverse f converter 206, tt/Lin
earK converter 207, ω/α transformer 208, LP
F (Low Pa5s iI'jlter) 20g
.

FM detection 5210, cross correlation calculation 62ti, peak detector 212, pulse generator 2131 multiplier 214°L
It is comprised of a PC combiner 215 and a D/A converter 216.

First, the transmitting side lvc will be explained. Input audio signal HA/
After being sampled by the D converter 101 at a pre-knob frequency of 8 KHz, the quantized signal is converted to a 12-bit bit number, and this quantized signal is sent to the window processor 102 and -J! ! It is supplied to the extension circuit 103.

The window processor 102 stores 30 m5 EC of the input quantized signal, that is, 240 samples each as a window time in the -H internal memory, and performs weight multiplication using a window function such as a Hamming function for each smsgc. Therefore 5m5E
C is the analysis 7V-me period.

The output of the window processor 102 is supplied to an autocorrelation coefficient calculator 104, which calculates a 12th order autocorrelation coefficient for each analysis frame and sends it to an LPC analyzer 105.

LPC analysis W I C) Based on the 12th-order autocorrelation coefficient input for each 5f1 analysis frame, the α parameter as the 12th-order LPC coefficient is extracted via a known 0LPC analysis means, and this 12th-order α parameter ij5m8Ec
is supplied to the LAP analyzer 106 and the LPC inverse filter 110 with an analysis frame period of V-frame.

The LSP analyzer 106H inputs the 12th order LSP (Lin
e Spectrum Pa1rs, line spectrum pair] parameter and supplies it to the linear converter 107.

The 12th-order 0LSP parameter output from the L8P analysis 5106 is a frequency domain parameter, and in the case of this embodiment, it is extracted using a quantized signal sampled at a Sanog frequency of 8 KHz, so its frequency is 0 to It is distributed over the frequency range <KHz. Linear converter 107
performs band compression by linearly converting this frequency distribution O to 4KH2 into a predetermined assigned frequency band of 2 to 3KHz.

The output of linear converter 107 is fed to interpolator 108,
Linear compression LAF input at analysis frame period of m5EC
The parameter input is interpolated at predetermined interpolation steps of 9 for each order, for example, 8 KHz and 125 μSEC, and the output thereof is supplied to the frequency generator 109. In this way, the frequency generator 109 is supplied with 212-order LSP parameters compressed into the assigned frequency band of 2 to 3 KHz, with the data smoothed by interpolation for each order, and whose values are updated for each analysis frame. Ru.

The frequency generator 109 has a function of generating 12 frequencies corresponding to each of the 12th LAP parameters.
Twelve such frequencies are generated within the KHz allocated frequency band and supplied to the adder 120.

In this way, the spectral envelope parameter is converted into 12 time-varying frequencies distributed in the assigned frequency band of 2 to 3 KHz and supplied to the adder 120. Assigned frequency band 2 to 3 KHztj for spectrum envelope parameter: Determined based on the following operating conditions.

That is, in this embodiment, the occupied frequency band of the transmission line is 0.
．． The target frequency is about 3 to 3.4KHz, of which 2
Spectral envelope parameters in the band ~3KHz, or
The band from t0.4 to 1.9 KHz is allocated as a residual waveform portion. The content of such allocation can be arbitrarily set in consideration of the occupied frequency band, audio reproducibility, and other operating conditions.

Now, the delay circuit 103 delays all the input quantized signals by a predetermined time and then supplies them to the LPC inverse filter 110.

The delay time in this case approximately corresponds to the processing time from when the quantized signal is input to the window processor 102 until it is output from the LPC analyzer 105 as a 12th-order α parameter while being updated every 5m5EC. Inverse filter 1
10 is provided with a quantized audio signal and a 12th-order α parameter for each same analysis frame period.

LPC inverse filter tt014. The LPC synthesis filter is modeled using LPC parameters as filter coefficients and is driven by the residual waveform etc. to synthesize the input audio signal.It is a nine filter with frequency characteristics opposite to that of the LPC synthesis filter, and has a 12th order α parameter that is updated every 5m5EC. A quantized audio signal is inputted with the filter coefficients as the filter coefficients, a residual waveform is outputted, and this is supplied to the window processor 1lllC.

Window processor 111H1 Residual waveform input in this way? 20m
It is cut out using a rectangular function with a window time of 5Ec and supplied to the total maximum value searcher 112 and the residual waveform normalizer 113. Therefore, 2omsEC becomes the analysis frame period for the residual waveform.

The maximum value searcher 112 searches for the maximum value k2 of the input residual waveform.
While searching every analysis frame of 0m5EC, this is sequentially supplied to the interpolator 114.

The interpolator 114 smoothes the data by interpolating between the input maximum values at predetermined interpolation increments of 9, for example 8 KHz and 125 μSEC, and supplies these to the μ converter 115. It is necessary to compress the dynamic range of several tens of dB or more in the analysis frame when the signal is changed and sent to the transmission line. This compression must be performed with the goal of optimizing the S/N ratio in the transmission path and the signal deterioration caused by compression. A μ converter 115 performs nonlinear compression using μ conversion, which allows small compression to ensure the necessary S/N even at a low level, and supplies this to a frequency generator 116 and a μ/Linear converter 117.

The frequency generator 116 generates a frequency that changes within the allocated frequency band of 1.6 to 9 KHz in response to the magnitude of the input residual waveform maximum value, and supplies this to the adder 120. Of the frequency band 0.4-1.9 KHz assigned to the residual waveform, 1.6-1.9 KHz is provided for the maximum value every 7 V-me period of analysis of 2 omsEc of the residual waveform.

On the other hand, the μ supplied to μ/L near f converter 117
The maximum value after f conversion is thereby restored to the state before f conversion and is supplied to the residual waveform normalizer 113. The residual waveform normalizer 113 utilizes the maximum value of the difference waveform input in this way, The residual waveform input from the window processor Ill is normalized. This normalization represents the positive residual waveform within each analysis frame of 20 msgc.

The maximum values in both negative directions are normalized with respect to the maximum value, and these normalized maximum value sequences are supplied to the residual waveform pulse generator 118 for each analysis frame. These normalized local maximum value sequences are shown as impulse trains indicated by positive and negative lines shown in FIG. For example, one of the pulses shown by the dotted line is a residual waveform pulse in the negative direction.Giving such a bandwidth means that each normalized local maximum value sequence is adjusted to the pulse center frequency. In this case, the generated waveform is a Gaussian waveform or a cosine square wave source, which simulates the residual waveform and occupies as little bandwidth as possible.

Looking at it from a different perspective, the provision of the fractional bandwidth mentioned above is intended to reduce the occupied bandwidth. Analysis: The number of residual waveform pulses for each period should be set as large as possible for the purpose of reproducing input audio signals, while the occupied frequency band should be as narrow and strict as possible. and bandwidth are set. The residual waveform pulse thus formed is a north pseudo residual waveform in which the waveform structure approximates the residual waveform and each extremely thick value including the maximum value is expressed as a normalized value, which is then used as the frequency generator 5 t l 9 into frequency domain data. In this conversion to frequency domain data, a frequency band of 0.4 to 1.5 KHz is allocated, and f Chivalry is performed.

The frequency generator 119 further converts the residual waveform into an FM modulated wave (1.5-0.4)/zKHz carrier wave, and sends the modulated wave to the adder 120. Send.

In this way, the residual waveform pulse H0 is the normalized maximum value of the frequency band of 4 to 1.5 KHz, and the residual waveform pulse is 1.6 to 1.6 KHz.
It will be expressed by the maximum value of the residual waveform in the 9 KHz frequency band. Generally speaking, it is often convenient for processing on the receiving side to handle data as frequency domain data in this way.

Addition 5120 adds the input frequency domain spectral envelope parameters and data regarding the residual waveform in a predetermined manner, and supplies the resultant signal to receiving side 2 via transmission line 1201i as a transmission signal.

On the receiving side 2, the transmitted signal input to the A/D converter 201 is sampled at the Sanoprinog frequency of 13 KHz, quantized with a predetermined number of bits, and then subjected to 30 m5 Ec Hamming windowing every 5 m5 Ec.
I) FT circuit 202 and delay circuit 203 are supplied with sliding data for each analysis frame period of m S EC.

Sliding DFT circuit 202H, input 12
Shift rate to preset processing for each analysis frame of 5μSEC, #iδKHz, 125μS in this example
The input data is subjected to kliill dispersive Fourier transform while performing sliding D F1', which processes the data in as small processing increments as possible while shifting the data by EC.
．． 5 m of data for frequency components between 6 and 1.9 KHz and subsequent <2 and 3 KHz i.e. 1.6 and 3 KHz
It outputs 5EC each and sends it to the peak picking circuit 20.
Supply to 2.

A peak picking circuit 204H performs peak picking to detect the maximum value of input data.
g) The maximum value of the residual waveform that exists as a maximum value included in the k'ii 1.6-1.9 KHz band, and the 12th order that exists as 12 maximum values included in the 2-3 KHz band. The LAP parameters of are detected and all corresponding frequencies are determined, and these are then interpolated 52051C for each analysis frame.

The interpolator 205 interpolates the maximum value of the input residual signal and the spectral envelope parameter at predetermined analysis increments of 8 KHz and +25 μSEC between analysis frames and supplies them to the linear inverse f-converter 206.

Note that in this embodiment, the peak pick-nog circuit 20 is
The output of 4 is output every 5m5EC, but this is 'k1
When outputting every 25 μSEC, the interpolator 205 becomes unnecessary.

The linear inverse f calculator 206 converts the input 12th-order LSP parameter into linear inverse f as a constant spectral envelope parameter. This linear inverse transformation is to inversely transform and restore the 0 to 4 KHz band compressed to 2 to 3 KHz again to 0 to 4 KHz. Mat1
The maximum value of the residual waveform has also been compressed by the μ transform, but this is supplied to the μ/L, 1near f converter 207 to cancel the μ transform, and then converted to the power 1i4. 214.

ω/ to l 2 LAP parameters subjected to linear inverse transformation
The LSP parameter (ω
) is converted into an α parameter f and this is supplied to the LPC synthesis 5215.

The LPC synthesizer 215i receives a 12th-order α parameter updated every 125μSEC as a filter coefficient of an all-pole digital filter from the ω/αjKm unit 208, and receives a residual waveform from the adder 214 to be driven. However, this residual waveform is reproduced as follows.

That is, the output of the A/D converter 201 is sent to the delay circuit 2.
03, but the delay circuit 203 is μ/Lin
A delay time is given so that the maximum value of the residual waveform output from the ear K converter 207 and the residual waveform pulse train output from the pulse generator 5213 have the same frame period, and this output is sent to the LPF 209. Supplied.

The LPF 209 performs FM modulation in the frequency band of 0.4 to 1.5 KF-1z, extracts only the t residual pulse, and supplies it to the total FM detector 210 .

The FM detector 210 receives the input signal '(l-FM
The residual waveform pulse shown in (■) in FIG. 2 is detected and supplied to the cross-correlation calculator 211.

The cross-correlation calculator 211 stores the normalized value of the residual waveform pulse shown in
K is stored, and after confirming that ■ is the desired residual waveform pulse through cross-correlation calculation between this and ■, it is supplied to the peak detector 212.

Peak detection 5212H detects all the peak values of the input cross-correlation calculator 211 output and sends them to the pulse generator 2.
13.

The pulse former 212 generates time-domain data of an impulse train of t level normalized to the maximum value at the time position indicated by the vertical line of the residual waveform shown in FIG.

Multiplier 214 i'i Multiplies the normalized level of the input impulse train by the maximum level to generate a residual waveform expressed by the actual level impulse train for each analysis frame,
This is supplied to the 1LPC synthesizer 215. The LPC synthesizer 215 uses the thus input 12th-order α parameter gold filter coefficient and is driven by the residual waveform to re-collect the digital audio.
supply to.

1) /A converter 216 converts the fl digital input into analog, removes unnecessary high frequency components with an LPF, and sends it out as an output audio signal.

In this way, confidential communication based on transmission of residual waveforms can be carried out without increasing the line 'Jft.

In this example, the residual waveform is expressed by a plurality of impulse trains that have amplitudes proportional to the maximum value of the residual waveform for each analysis frame and are located at positions that match the time positions of the residual waveforms. It is clear that other similar expressions, such as multipulses formed by substantially the same impulse train with limited pulse intervals, may be used in substantially the same manner.

Further, the α parameter, the analysis order of the LSP parameter, etc. may be set arbitrarily.

〔Effect of the invention〕

As explained above, according to the present invention, in a confidential communication device that decomposes an input audio signal into spectral envelope parameters and residual waveforms and transmits them, converts the input audio signals into frequency domain data of frequency bands assigned independently from each other. A plurality of ear pulses transmitting the spectral envelope parameter and the residual waveform, the residual waveform has an amplitude proportional to its maximum value, and each time position corresponds to the time position of each local maximum value of the residual waveform. By carrying out secret communication using a means of representing the data by rows, it is possible to realize a secret communication device that can significantly improve the deterioration of sound quality without increasing the line capacity.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the transmitting side of the secure communication device of the present invention, and FIG. 2 is a block diagram showing the configuration of an embodiment of the receiving side of the secure communication device of the present invention. l...Sending side, 2...Receiving side, lot
...A/f) Conover j', 102...
・Window processor, 103...Delay circuit, 104...
...Autocorrelation coefficient calculator, 105...LP
C analysis 6.106...L, SP analyzer, 107
・・・・・・Linear converter, 108・Old・Interpolator, 10
9...Frequency generator, 110...LP
C inverse filter, 1.1...Window processor, 112.
... Maximum value searcher, 113 ... Residual waveform normalizer, 114 ... Interpolator, 115 ...
...μ converter, 116 ... Frequency generator, 11
7.・Dan・μ/Linear conversion 6,118・・・・
... Residual waveform pulse generator, 119 ... Frequency generator, 120 ... Adder, 201 ...
・A/D converter, 202...Sliding DFT [gl path, 2o3...Delay circuit, 2o
4...Peak Bikki 7g circuit, 2o5/old...
Interpolator, 2o6...Linear inverse transformer% 207.
...μ/Linear converter, 208...
...ω/α converter, 2o9...LPF, 210
...FM detector, 211...Cross correlation calculator, 212...Peak detector, 213...
... Pulse formation 5.214 ... Multiplier, 2
15...Dan/LPC synthesizer, 216...D/A
converter. -＋Taku

Claims (1)

[Claims] In a confidential communication device that decomposes and transmits an input audio signal into a spectral envelope parameter expressing its spectral envelope and a residual waveform from which the spectral envelope is removed, frequency bands are assigned independently from each other in a preset occupied frequency band. In this case, the residual waveform has an amplitude proportional to its maximum value, and each time position corresponds to each local maximum value of the residual waveform. What is claimed is: 1. A secret speech device comprising an input speech signal converting means for converting an input speech signal represented by a plurality of impulse trains corresponding to time positions.