JPS617745A - Privacy device of multiple transmission line - Google Patents

Abstract

PURPOSE:To perform high-level privacy communication with an excellent transmission efficiency, by providing a spectrum converting device of transmission frequencies at the transmitting side and a reversely converting device of spectrum at the receiving side and performing the privacy communication after the content of the spectrum conversion is further complicated. CONSTITUTION:At the transmitting side, spectrum converting circuits 1-1-1-N respectively perform spectrum conversions while inputting filter weighting coefficient data received from a spectrum memory 3 against sound signals inputted from each channel CH1-CHN and send their outputs to the receiving side from a carrier transmitting terminal station 2 through a transmission line 201. At the receiving side, on the other hand, the transmitted signals are received by a receiving terminal station 6 and respectively sent to converting circuits 7-1-7-N for reversely converting frequency spectra and an operation synchronizing signal is sent to a PN signal generating circuit 9. The circuit 9 receives the same operation synchronizing signal as the PN signal generating circuit 4 at the transmitting side receives and is controlled by a keying circuit 10 which is the same circuit as the keying circuit 5 at the transmitting side. Therefore, by only increasing the synchronizing signal data transmission for generation PN code, privacy can be secured at a high level under a complicated condition.

Description

[Detailed Description of the Invention] [Technical Field] The present invention relates to a multiplex transmission path secure device, and more particularly to a multiple transmission path secure device that performs secure communication via multiplex transmission paths using frequency division multiplexing (hereinafter abbreviated as FMD) circuits. .

[Prior art]

The FMD communication method is well known, in which the transmission frequency band is divided into multiple channels whose frequency regions do not overlap with each other, and multiple signals are transmitted and received via these channels.
Various types of confidential voice communication are also being carried out using FMD lines. -However, in order to encrypt voice in an extremely complex manner to make it a high-level confidential message, it is necessary to digitize the voice by quantizing it at a high vibrancy rate.As a result, the transmission efficiency tends to decrease, and especially the line quality A disadvantage is that a poor carrier line, such as O/H (non-line-of-sight) communication, tends to increase the error rate.

[Purpose of the invention]

An object of the present invention is to eliminate the above-mentioned drawbacks, and to provide a multiple transmission line secure device for performing secure communication via multiple transmission lines based on F'DM lines, in which a transmission frequency spectrum conversion means is provided on the transmitting side, and a transmission frequency spectrum conversion means is provided on the receiving side. After receiving this signal, it is equipped with a spectrum inverse conversion means that returns the spectrum to the state before the frequency spectrum conversion, and further complicates the contents of this spectrum conversion to carry out secret communication. It is an object of the present invention to provide a multi-transmission path secret communication device that can realize high-level secret communication with good transmission efficiency.

[Structure of the invention]

The device of the present invention is a multiplex transmission line secure device that performs secure communication via multiplex transmission lines using FDM line tags, in which the transmitting side converts and transmits the frequency spectrum of each channel constituting a frequency multiplexed line, and the receiving side Now let us inversely convert this into this frequency spectrum and output it again, and in this case, the frequency spectrum conversion characteristics for each channel are different from each other, and even for the same channel, the frequency spectrum conversion characteristics constantly change over time. , and frequency spectrum conversion in which this time-varying characteristic can be changed arbitrarily in response to a key code set in advance between the transmitting side and the receiving side. It is configured to include inverse conversion means.

〔Example〕

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

FIG. 1 is a block diagram showing a first embodiment of the present invention.

The first embodiment shown in FIG. 1 is a spectrum conversion circuit 1-1.
．． 1-2. ......1-N, carrier transmitting terminal station 2
, spectral memory 3 , P N (Pseudo N
o1se, pseudo noise) generating circuit 4 and a keying circuit 5, a carrier receiving terminal station 6, and a spectrum inversion circuit 7-1.7-2.・・・・・・・・・7-N,
It is configured to include a spectrum memory s, a PN signal generation circuit 9, and a receiving side that supports a keying circuit 10.

Each spectrum conversion circuit 1-1 to 1-N on the transmitting side is an FMD.
N channels CH1, CH2, which constitute N channels CH1, CH2,.

. . . Converts the frequency spectrum of the CHN transmission signal. The purpose of this frequency spectrum conversion is to hide each audio signal transmitted using each channel from third parties. The spectrum conversion circuits 1-1 to 1-N convert the frequency spectrum of the audio signal to eliminate the unique characteristics of the audio signal.

The characteristics of the audio signal are determined depending on the human vocal cord characteristics and vocal tract characteristics. It is well known that this characteristic is also represented by a spectral envelope indicating macroscopic features and sound source information indicating a fine #1 #II structure. The spectral envelope information is removed from the above-mentioned sound source information line audio signal, and includes various information such as pitch information, voiced/unvoiced/silent information, and strength of the sound source. Therefore, it is possible to eliminate the sound characteristics by converting the characteristics of either or both of the spectral envelope information and the sound source information.

However, converting the sound source information as the fine structure of the sound is often inherently difficult to reproduce accurately. Therefore, in this embodiment, the frequency spectrum is converted illegally by deforming the spectrum envelope.

Such deformation of the spectral envelope can be arbitrarily utilized as long as the deformation can be restored on the receiving side. There are various possible methods for converting the spectrum of the audio signal of each channel on the transmitting side and then performing inverse transformation on the receiving side to restore it to the state before conversion.
For example, the transmitting side is equipped with a conversion filter having a transfer function H(Z) determined corresponding to the desired conversion spectral envelope data, and the receiving side receives the audio signal of each channel that has passed through this conversion filter. The reciprocal of H(Z) H-1(Z
) can be passed through an inverse conversion filter to restore the audio signal before conversion, and it is possible to maintain the confidential state in which the spectrum of the audio signal has been converted during transmission.

The above-mentioned variable and inverse conversion filters can be easily implemented using, for example, a transversal type filter or a furnace filter, and in this example, a transversal filter configuration is also used to realize these and create a spectrum conversion circuit. 1-1 to 1-N, and spectrum inversion circuits 7-1 to 7-N.

Figure 2 shows the spectrum conversion circuit (
It is a block diagram showing the basic composition of A) and a spectrum inversion circuit (B).

Figure 2 (A) shows the unit delay amount Z” (Z= expjwT
, or=2πf, f is the frequency, T is the frequency sampling period) delay elements D(1), D(2), D(3
)・−・・・・D(N), weighting coefficient α1. α2
．．・・・・・・・・・Multiplier P that multiplies α and αn
(1), P(2), ---・-・・-・P (n
-1), P (n) and an adder S1. Weighting coefficient α1. α2・・・・・・αn-1
, αn are nth-order parameters as linear prediction coefficients supplied from the spectrum memory 3, and the spectrum memory 3 is prepared with ms sets of such n-thorder α parameters. LPC (Linear Prediction) of these m types of n-th α parameter branch audio materials
Coefficient (Linear Prediction Coefficient) is extracted in advance through analysis, but other LPCs such as a partial autocorrelation coefficient (partial autocorrelation coefficient) may also be used.

The audio signal supplied from the input line of such a conversion filter is subjected to convolution with the transfer function H(Z), and its frequency spectrum is converted and sent to the output line.

Now, the spectrum inversion circuit shown in Fig. 2(B) is
The output from the spectrum conversion circuit shown in Figure (A) is fed back to the input side, added to the input signal, and supplied to the delay element group.
) has an inverse transfer function H-1(Z). Therefore, the audio signal which has undergone convolution integration using the transfer function H(Z)K is passed through this inverse transformation filter to be restored to the state before transformation. Transmission side spectrum conversion time t6x-1
~I-N and receiving side spectrum inversion circuit 7-1
~7-N are respectively provided with such spectral transformation and inverse transformation filters to actually transform and restore spectral components.

In FIG. 1, spectrum conversion circuits 1-1 to 1-N perform the above-mentioned spectrum conversion while inputting filter weighting coefficient data received from spectrum memory 3 for audio signals input from each channel of CHI to CHN, respectively. The respective outputs are sent to the carrier transmitting terminal station 2.

The spectrum memory 3 stores a set of n-th order α parameters of the maI type extracted and set in advance on the basis of audio materials, etc., in units of time corresponding to processing units of transmission and reception.

Here m))N.

Each of the m types of primary α parameter sets stored in this way is a PN signal sent from the PN signal generation circuit 4.
The spectral envelope information that is different from each other is read out randomly at the timing of signal occurrence, and is sent to the spectral conversion circuit l.
-1 to 1-N.

The PN signal generation circuit 4 is composed of an n-stage shift register and a logic circuit that feeds back the theoretical combination of states of the plurality of stages to the input side of the shift register. The code of the sequence is generated as a PN signal. Further, the PN signal generating circuit 4 generates the PN signal in a state in which the logic circuit is keyed by the keying circuit 5 using a keying code that is previously specified.

In this way, each channel constituting FDM is given a different spectrum conversion characteristic, and even the same channel has its spectrum conversion characteristics changed from time to time. This will be carried out under the control of a keying code known only to the operator.

The spectrum conversion circuits 1-1 to 1-N supply the N-channel frequency spectrum conversion audio signals thus formed to the carrier transmission terminal station 2, and the PN signal generation circuit 4 supplies an operation synchronization signal. These inputs are transmission signals of a predetermined format, in this example SSB (Sin
The signal is transmitted to the receiving side via the transmission path 201 as a carrier transmission wave of the GLE 5ide Band) system.

On the receiving side, the transmission signal transmitted in this way is received by the carrier reception terminal station 6, and a frequency spectrum converted audio signal is extracted, which is passed through spectrum inverse conversion circuits 7-1, 7-2.・・・・・・
...7-N, and the operation synchronization signal is supplied to the PN signal generation circuit 9, respectively.

The spectrum inverse conversion circuits 7-1 to 7-N perform spectrum inversion to return the spectrum conversion performed on the transmitting side to its original state.
A spectral inverse transform filter with an inverse transfer function of 1 to 1-N is used to receive the frequency spectrum transform on the transmitting side and perform inverse filtering processing to remove the convolution integral of the audio signal, so-called deconvolution processing.
i ou ) to restore and output the audio signal before frequency spectrum conversion.

The filter coefficients of the spectral inverse transform transversal filter required in the above-mentioned inverse transform filtering process are obtained from the spectral memory 8 having the same contents as the spectral memory 3 at the timing of code occurrence constituting the PN signal output from the PN signal generation circuit 9. The signals are read out randomly to each of the spectrum inverse conversion circuits, but the code generation timing and code content of the PN signal generation circuit 9 are the same due to the same operation synchronization signal and configuration as the PN signal generation circuit 4 on the transmitting side. Furthermore, since the keying circuit 5 is controlled by the same key ink circuit 10, the spectral envelope data read from the spectral memory 3 for the spectral conversion processing in each of the spectral conversion circuits 1-1 to 1-N on the transmitting side is The same spectrum envelope data read from the spectrum memory 8 is always used for spectrum inversion processing in each of the spectrum inversion circuits 7-1 to 7-N on the receiving side.

In this way, complex and sophisticated communication that secures a high level of confidentiality can improve line quality without causing a decline in basic transmission efficiency by increasing the transmission of synchronization signal data for PN code generation at most. It can be easily implemented even in poor transport lines.

FIG. 3 is a block diagram showing the configuration of a multiple transmission line privacy device in a second embodiment of the present invention.

The configuration of the second embodiment shown in FIG. 3 is such that only the frequency switching circuit 11 on the transmitting side and the frequency switching circuit 12 on the receiving side are added to the first embodiment shown in FIG. Since the contents of the same symbol numbers are the same, a detailed explanation regarding them will be omitted.

In the second embodiment shown in FIG. 3, in the first embodiment, FM
The frequency band of each channel constituting D is subdivided into frequency intervals so minute that the pitch structure of the voice cannot be distinguished, and the frequencies are swapped across channels between these subdivision intervals between each channel to conceal it. It is a more complex and sophisticated version of the standard. From a different perspective, the first
In the second embodiment, the vocal tract characteristics are concealed through conversion of the spectral envelope characteristics, and in the second embodiment, the vocal tract characteristics are also concealed by making the pitch structure invisible. .

FIG. 4 is an explanatory diagram of a channel subdivision structure for explaining the contents of the channel subdivision structure in the second embodiment.

Figure 4 shows an example of a subdivision structure for arbitrary three channels of N FMD channels CHI to CHN, CHI, CHJ, and CHK channels, and an example of using frequencies in the subdivision structure of these three channels by replacing them. This shows the subdivision structure of the channel CHZ as follows. The embodiment shown in FIG. 3 will be described below with reference to FIG. 4.

The frequency switching circuit 11 includes spectrum conversion circuits 1-1 to 1-.
When a spectrally converted audio signal from N is input, it is passed through a filter bank to perform channel subdivision. In this example, the transmission frequency band is 0.3kHz to 3.4kHz.
An audio signal of z is transmitted through each channel with a width of 4 kHz, but the subdivision of the channel is such that the pitch interval is unknown due to the maximum number of pitches to be targeted.In this embodiment, the channel band is #Mibo This is implemented using a filter bank composed of 13 filters whose subdivision number is 13, which is the value divided by the maximum pitch frequency.

F shown in FIG. 4 indicates 13 consecutive subdivision band groups formed by such a filter bank, and CHI
~Spectral envelope Si of each audio signal using CHK,
Sj and Skl and Pi.

pj and Pk are each divided into such subdivision bands. Note that the characteristics of each subdivision band constituting the subdivision band group F overlap with each other, and the overall frequency range is from 0.3kHz to 3kHz.
．． It is set to pass frequencies up to 4 kHz with almost uniform passing characteristics.

In this way, the spectral envelope and pitch information of each audio signal using CHI-CHK are subdivided by the filter bank, and then these subdivision information are frequency swapped with each other across channels as follows. Ru.

The spectral envelope information and pitch information subdivided by the filter bank are once stored in the memory and then read out one after another according to the specified contents under the control of a program stored in the ROM in advance. C shown in Figure 4
The subdivision structure of Hl shows an example of implementing frequency swapping in this way.p, the spectral envelope subdivided by the seventh subdivision band CHK7 of CHK for the 13 subdivision bands of CHI. and pitch information on CH
C) (The contents of the fifth subdivision band CHJ5 of J are in the second subdivision band of CH1,
Also, the contents of the third subdivision band CHI3 of CHI are CH
Frequency shuffling across channels is then performed in such a way that each third subdivision band of I is assigned to the third subdivision band.

In this way, frequency swapping across channels is also performed for each channel CH2 to CHN. Note that this frequency swapping can be performed for all of the subdivided bands of each channel, or it can be performed for a part of the subdivided band, and it is possible to apply either one as desired. Can be set to

In this way, when the pitch structure can be kept secret, rI
After performing frequency swapping in subdivision band units for each channel divided at intervals of 5 times, the output of each channel is sent to the carrier transmitting terminal station 2, and is transmitted using a predetermined transmission method, in this example, SSB (Single 5ide Ban).
d) is transmitted as a carrier transmission signal to the receiving side via the transmission path 201.

On the receiving side, when the carrier transmission signal is received by the carrier reception terminal station 6, the baseband signal for each channel, that is, the output of the frequency switching circuit 11, is extracted from the carrier transmission signal and is supplied to the frequency reverse switching circuit 12.

The frequency inverse switching circuit 12 temporarily stores the subdivided band input for each channel in the memory via the filter bank, and then converts it to the state before the frequency switching operation, that is, CHI shown in FIG. 3, under the control of the program in the built-in ROM. ~Restored to subdivision structure state like CHK, spectral envelope 5i-8
After obtaining the k1 pitches Pi to Pk, etc., audio output is sent to each channel of CHI to CHN in substantially the same manner as in the first embodiment shown in FIG. In this way, the second embodiment conceals the pitch information without damaging it, and
Furthermore, the content of the concealment in the first embodiment is made even more complex and sophisticated by subdividing each channel and performing frequency switching across channels in units of subdivision.

Note that in the first and second embodiments shown in FIGS. 1 and 3, the PN signal based on the M-sequence code is used as the first timing setting for spectrum conversion; For example, a Gold code sequence may be used, and the number of channel subdivisions in the second embodiment is 13.
However, this can also be set arbitrarily, taking into account the audio transmission band, the maximum number of target pitches, etc.

〔Effect of the invention〕

As explained above, according to the present invention, a multiplex transmission path secure communication device using an FDM line is provided with means for converting the spectrum of the transmission frequency and its inverse conversion, and further complicates the content of the spectrum conversion to carry out secure communication. By providing this means, it is possible to realize a multi-transmission line secret communication device that can perform extremely complicated and sophisticated secret communication with good transmission efficiency even on an FDM line with poor line quality.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a first embodiment of a multiplex transmission line privacy device according to the present invention, and FIG. 2 shows a spectrum conversion circuit (A) and a spectrum inverse conversion circuit (B) in the embodiment of FIG. FIG. 3 is a block diagram showing the configuration of a second embodiment of the multiplex transmission line privacy device according to the present invention, and FIG. 4 shows the contents of the channel subdivision structure in the second embodiment. FIG. 2 is an explanatory diagram of a channel subdivision structure for explaining. 1-1 to 1-N...spectrum conversion circuit, 2.
...Carrier transmitting terminal station, 3...Spectrum memory, 4...PN signal generation circuit, 5...
・Keying circuit, 6...Carrier receiving terminal station, 7-
1 to 7-N... Spectrum inverse conversion circuit, 8...
... Spectrum memory, 9 ... PN signal generation circuit, 10 ... Keying circuit, 11 ...
...Frequency switching circuit, 12... Frequency switching circuit. CB) No. 21￥I Rod 4 Figure

Claims (2)

[Claims] (1) In a multiple transmission line secure device that performs secure communication via multiple transmission lines using frequency division multiplex lines, the transmitting side converts and transmits the frequency spectrum of each channel that makes up the frequency division multiplex line, and the receiving side converts the frequency spectrum for each channel that makes up the frequency division multiplex line. is to be inversely converted into this frequency spectrum and outputted again, and in this case, the frequency spectrum conversion characteristics for each channel are different from each other, and even if the same channel is used, the frequency spectrum conversion characteristics constantly change over time. Frequency spectrum conversion whose time-varying characteristics can change arbitrarily in response to a preset key code between the transmitter and receiver. 1. A multiplex transmission path secret communication device, characterized in that it is equipped with an inverse conversion means and performs secret communication. (2) The frequency band of each channel constituting the frequency division multiplex line is subdivided at frequency intervals so minute that the pitch structure of communication voice cannot be distinguished, and the frequency band between the subdivided bands of each channel is divided across channels. The multiplex transmission path secret communication device according to claim (1), characterized in that the secret communication is carried out by exchanging the following.