JPS63209338A - Privacy call equipment - Google Patents

Abstract

PURPOSE:To insert pseudo voices different from each other at every interblock into the interblock by updating and storing the sound part of a voice data and using a part of the stored voice data for the insertion to the interblock. CONSTITUTION:If a value representing no input of prescribed number in a voice data outputted from an FFT privacy call equipment 1 is consecutive, an interblock detection means 2 discriminates the consecutive interval of the value as the interblock. Moreover, only the sound part of the voice data outputted from the device 1 is stored in a ring buffer 4 according to the result of discrimination. In the case of the discrimination as the interblock, the data written at the discrimination of the data written in the buffer 4 is subjected to sequence replacement by a data sequence replacement means 5. The replaced data, that is, the pseudo voice is inserted to the interblock by a pseudo voice input means 3 and outputted as the privacy call voice data.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a confidential communication device, and more particularly to an improvement in silent part removal means.

[Conventional technology]

Figure 2 shows a conventional FFT secret speech device that hides silent parts by inserting a dummy spectrum, where (al is normal voice, (b) is a dummy spectrum, and (C) is normal voice ta).
dummy spectrum (b) is superimposed on l, and (d) is a diagram showing the spectrum distribution of the secret speech output, respectively, (
el is a block diagram showing the configuration of a conventional device. In FIG. 2 (el), 11 is an encoding means for sampling and encoding audio, 12 is an FF7 means for calculating the sampled audio and creating FFT coefficients, and 13 is an FET coefficient for replacing the FFT coefficient output by the FFT means 12. A switching means 14 converts the FFT coefficients obtained by the switching means 13 into an inverse FF
15 is a D/A converter, 16 is a decoding signal generating means, and 17 is a waveform shaping L.
It is PF. Of these, means 12 to 14 and 16 are provided in the same computer.

This type of secret speech device (FFT scrambler) inputs a signal including the call voice and a dummy spectrum into a computer 11, subjects it to FFT means 12, and exchanges the FFT coefficients obtained thereby with each other by means of interchanging means 13. The FFT coefficients are returned to an audio signal by the inverse FFT means 14, a decryption key generated by the decryption signal generation means 16 is added thereto, and the signal is returned to the D/A converter 15.
It converts the signal back into audio and sends it out to the transmission path using the LPF 17.

Figures 2+al to (d) show how the audio spectrum changes with this device. In other words, the encoding means 11 encodes the speech (C1) obtained by adding the normal speech (a) and the dummy spectrum (b).
and after the above process, (d
You will get a sound like +. Dummy spectrum (b
) may be input in addition to the normal voice, or may be generated within 12 to 14 computers. When there is silence, this dummy spectrum hides the silence.

[Problem that the invention seeks to solve]

Conventional devices are configured as described above, and since silent parts are only dummy spectra, the presence of silent parts can be easily detected by a spectrum analyzer. "yes""
There was a problem that the text (such as "No") would be deciphered.

This invention was made in order to solve the above-mentioned problems, and provides a secret speech device that can insert pseudo-voices that cannot be distinguished from normal voices and that are different for each silent part into silent parts. The purpose is to

[Means for solving problems]

The secret conversation device according to the present invention updates and stores a sound part of audio data, and uses a part of the stored audio data to insert into a silent part'.

[Effect]

In this invention, since the audio data of the voiced part is used as the pseudo-sound for inserting the silent part, the spectrum distribution is close to that of normal speech, making it difficult to distinguish it from normal speech. Also, the recorded audio data is constantly rewritten, so
A different pseudo-sound will be inserted for each silent section into which a pseudo-sound should be inserted.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 (al is a block diagram showing a confidential communication device according to an embodiment of the present invention, and FIG. 1 (bl is a flowchart showing its operation. 11 to 15, no dummy spectrum is added here. 2 is a silent part detecting means for detecting a silent part of the audio data from the FFT encrypting device 1; An 8 Kbyte circular buffer that stores confidential voice data while updating only the active part; 5 is a data order changing means for changing the order of predetermined data in the circular buffer 4; 3 is a data order changing means with the circular buffer 4; 2 to 5 are on the same computer as 12 to 15 of the conventional device shown in FIG. 2(e). , constitutes a silent part removing means.

Next, the operation will be explained with reference to the flowchart shown in FIG. 1(b).

First, the silent part detection means 2 determines that when a certain number of values in the audio data output from the FFT secret speech device 1 continue, indicating that there is no audio input, the period during which the values continue is determined to be a silent part. (Step 31). Also circular buffer 4
According to the determination result of the silent part, only the sound part of the audio data output from the FFT secret speech device 1 is stored (S4
).

When it is determined that it is a silent part, the order of the data written in the circular ring sofa 4 after the data written at the time of the determination is rearranged by the data order rearrangement means 5 (S2), and the rearranged data is That is, the pseudo-sound is inserted into the silent portion by the pseudo-sound insertion means 3 (S3).

The confidential voice data thus obtained is then output to the line through a D/A converter and an LPF (35).

The reason why the circular buffer is set to 8 Kbytes is to correspond to the silent time (about 1 second) of a normal telephone call. Furthermore, the reason why the data is read after the last written data during silence is to use the audio data located as far away from the start of the silence as possible. It is necessary to send a signal indicating the existence of a silent part to the receiving side in some way together with the voice into which this pseudo-voice has been inserted, and to restore the silent part, but this is shown in Figure 2 (e
It is sufficient to send it as the decoded data 16 shown in l.

In this way, in this embodiment, the updated and stored audio data of the voiced part is used as the dummy voice to be inserted into the silent part. It is possible to insert different pseudo-I sounds into silent parts.

In the above embodiment, the silent part removal means is added next to the inverse FFT means (Fig. 2 + e) 14''), but this is similar to that shown in Fig. 2 (el encoding means 11 and FF7 means 12,
FF7 means 12 and replacement means 13. Replacement means 13
It may be added anywhere in the inverse FFT means 14 and the same effect will be achieved.

Furthermore, in the above embodiment, the data taken in from the inverse FFT means 14 is inserted after the inverse FFT means 14, but
This is an inverse FF for the data taken in after the encoding means 11
The data may be inserted next to the T means I4, or vice versa, and the location where the data is taken in and the location where it is inserted may be different.

Further, in the above embodiment, the data order is changed and inserted into the silent part, but it is not necessarily necessary to change the data order. Further, although the data stored after the time when it is determined to be a silent part is inserted, the invention is not necessarily limited to this. Further, the storage device for the sound portion does not necessarily have to be a circular buffer, but may be any storage device whose contents are constantly updated. Further, the data order may be changed before being written to the storage device.

Further, the present invention can be used not only in FFT secret speech devices but also in general speech secret devices when it is desired to hide silent parts.

[Effects of the Invention] As described above, according to the secret speech device of the present invention, the silent part removing means is composed of the silent part detecting means for detecting the silent part of the audio data and the silent part detecting means for detecting the silent part of the audio data. Since it is constituted by a storage means for updating storage and a pseudo-sound insertion means for inserting the audio data in the storage means into the detected silent portion, the pseudo-sound cannot be distinguished from normal speech and is different for each silent portion. This has the effect of allowing you to insert the text into a silent section.

[Brief explanation of the drawing]

Fig. 1 (al is a block diagram showing a confidential communication device according to an embodiment of the present invention, Fig. 1 (bl is a flow chart showing its operation, Fig. 2 (a), (b), (C1, (dl
2 is a diagram showing the state of the audio in the conventional confidential communication device, and FIG. 2(e) is a block diagram showing the conventional confidential communication device. In the figure, 1 is an FFT secret speech device, 2 is a silent part detection means, 3 is a pseudo voice insertion means, 4 is a circular buffer (storage means), and 5 is a data order changing means.

Claims (4)

[Claims] (1) In a confidential communication device used for secret communication, a silent part detection means for detecting a silent part of voice data, a storage means for updating and storing a sound part of the voice data, and the above-mentioned storage in the silent part of the voice data. 1. A secret speech device comprising a silent part removing means comprising a pseudo voice inserting means for inserting voice data of the means. (2) A patent characterized in that the silent part removing means is equipped with data order changing means for changing the order of audio data stored in the storage means or the order of audio data inserted into the silent part. A secret communication device according to claim 1. (3) The confidential communication device according to claim 1 or 2, wherein the storage means is a circular buffer. (4) The pseudo voice insertion means inserts the voice data stored in the storage means after the silent portion detection means detects the silent portion. The secret communication device according to any one of paragraph 3.