JP4900402B2 - Speech code conversion method and apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable both speech communication and data communication between a communication system, having only a speech line, and a communication system having a data line, for other than the speech line. <P>SOLUTION: In a speech code converter 103, a first speech code obtained by encoding input speech by a first speech encoding system is converted into a second speech code according to a second speech encoding system. The first speech code is converted to the second speech code; and when optional data is embedded in the first speech code received from the transmission end, the embedded data is extracted from the first speech code, and the second speech code and the extracted data are transmitted separately to the transmission destination. <P>COPYRIGHT: (C)2009,JPO&amp;INPIT

Description

  The present invention relates to a speech code conversion method and a speech code conversion device, and in particular, a speech code encoded by a speech encoding device used in a network such as the Internet or a speech encoding device used in an automobile / mobile phone system or the like. The present invention relates to a voice code conversion method and a voice code conversion apparatus for converting a voice code into a voice code of another encoding method.

  In recent years, the amount of communication between different communication systems is expected to increase more and more due to the diversification of mobile phone systems, the explosion of subscribers, and the spread of voice communication using the Internet (Voice over IP: VoIP). . In a voice communication system such as a cellular phone or VoIP, a voice coding technique for compressing voice is used to effectively use a communication line. Mobile phones use different voice coding techniques depending on the country or system, and W-CDMA employs an AMR (Adaptive Multi-Rate) system as a world-wide voice coding system. On the other hand, in VoIP, ITU-T recommendation G.729A is widely used as a voice encoding method. Hereinafter, the G.729A encoding method and decoding method will be described, and differences between the G.729A and AMR methods will be described.

The G.729A encoding method and decoding method are as follows.
-Configuration and operation of encoder FIG. 18 is a configuration diagram of an encoder of the ITU-T recommendation G.729A system. In FIG. 18, an input signal (audio signal) X having a predetermined number of samples (= N) per frame is input to the LPC analyzer 1 in units of frames. If the sampling rate is 8 kHz and one frame period is 10 msec, one frame is 80 samples. The LPC analysis unit 1 expresses the human vocal tract as
H (z) = 1 / [1 + Σαi · z ⁻ⁱ ] (i = 1 to P) (1)
The coefficient αi (i = 1,..., P) of this filter is obtained. Here, P is the filter order. Generally, in the case of telephone band voice, a value of 10 to 12 is used as P. The LPC (Linear Prediction) analysis unit 1 performs LPC analysis using a total of 240 samples of 80 samples of the input signal, 40 samples of the pre-reading, and 120 samples of the past signal to obtain LPC coefficients.

  The parameter converter 2 converts the LPC coefficient into an LSP (Line Spectrum Pair) parameter. Here, the LSP parameter is a parameter in the frequency domain that can be mutually converted with the LPC coefficient. Since the quantization characteristic is superior to the LPC coefficient, the quantization is performed in the LSP area. The LSP quantization unit 3 quantizes the converted LSP parameter to obtain an LSP code and an LSP inverse quantization value. The LSP interpolation unit 4 obtains an LSP interpolation value from the LSP inverse quantization value obtained in the current frame and the LSP inverse quantization value obtained in the previous frame. That is, one frame is divided into two first and second subframes of 5 msec, and the LPC analysis unit 1 determines the LPC coefficient of the second subframe, but does not determine the LPC coefficient of the first subframe. Therefore, the LSP interpolation unit 4 predicts the LSP inverse quantization value of the first subframe by interpolation using the LSP inverse quantization value obtained in the current frame and the LSP inverse quantization value obtained in the previous frame.

The parameter inverse conversion unit 5 converts the LSP inverse quantization value and the LSP interpolation value into LPC coefficients, respectively, and sets them in the LPC synthesis filter 6. In this case, as the filter coefficient of the LPC synthesis filter 6, the LPC coefficient converted from the LSP interpolation value is used in the first subframe of the frame, and the LPC coefficient converted from the LSP dequantized value is used in the second subframe. . In the following description, 1 with a subscript, for example, 1 in lspi, li (n),.
The LSP parameter lspi (i = 1,..., P) is quantized by scalar quantization or vector quantization in the LSP quantization unit 3, and then the quantization index (LSP code) is transmitted to the decoder side. The

  Next, sound source and gain search processing is performed. Sound source and gain are processed in subframe units. First, the sound source signal is divided into a pitch period component and a noise component, the pitch code component is quantized using the adaptive codebook 7 storing a past sound source signal sequence, and the noise component is quantized with an algebraic codebook. Or a noise codebook. In the following, a speech coding scheme that uses the adaptive codebook 7 and the algebraic codebook 8 as the excitation codebook will be described.

  The adaptive codebook 7 outputs a sound source signal (referred to as a periodic signal) for N samples sequentially delayed by one sample corresponding to the indexes 1 to L. N is the number of samples in one subframe (N = 40), and has a buffer for storing the pitch period component of the latest (L + 39) samples. Index 1 identifies the periodic signal consisting of the 1st to 40th samples, Index 2 identifies the periodic signal consisting of the 2nd to 41st samples, ... Lth to L + 39th samples based on the index L A periodic signal is identified. In the initial state, the contents of the adaptive codebook 7 contain all signals with an amplitude of 0. For each subframe, the oldest signal in time is discarded by the subframe length, and the excitation signal obtained in the current subframe is applied. It operates so as to be stored in the codebook 7.

  In the adaptive codebook search, the periodic component of the excitation signal is identified using the adaptive codebook 7 storing the past excitation signal. That is, while the starting point read from the adaptive codebook 7 is changed by one sample, the past excitation signal in the adaptive codebook 7 is extracted by subframe length (= 40 samples), and is input to the LPC synthesis filter 6 to be input to the pitch synthesis signal β. Create × A × PL. Here, PL is a past pitch periodic signal (adaptive code vector) corresponding to the delay L extracted from the adaptive codebook 7, A is an impulse response of the LPC synthesis filter 6, and β is an adaptive codebook gain.

The arithmetic unit 9 calculates the input power X and the error power EL of β × A × PL as follows:
EL = | X−β × A × PL | ² (2)
Ask for. When the weighted composite output of the adaptive codebook output is A × PL, the autocorrelation of A × PL is Rpp, and the cross correlation between A × PL and the input signal X is Rxp, the error power in equation (2) is minimized. The adaptive code vector PL for the pitch lag Lopt is given by
P _L = argmax (Rxp ² / Rpp) (3)
Is represented by That is, the reading start point at which the value obtained by normalizing the cross-correlation Rxp between the pitch synthesized signal A × PL and the input signal X by the autocorrelation Rpp of the pitch synthesized signal becomes the optimum starting point. As described above, the error power evaluation unit 10 obtains the pitch lag Lopt that satisfies the expression (3). At this time, the optimum pitch gain βopt is expressed by the following equation: βopt = Rxp / Rpp (4)
Given in.

代数符号帳８は、１サブフレームを構成するＮ(=40)サンプル点を複数のパルス系統グループ１〜４に分割し、各パルス系統グループから１つのサンプル点を取り出してなる全組み合わせについて、各サンプル点で＋１あるいは−１のパルスを有するパルス性信号を雑音成分として順次出力する。この例では、基本的に1サブフレームあたり4本のパルスが配置される。 Next, the noise component contained in the sound source signal is quantized using the algebraic codebook 8. The algebraic codebook 8 is composed of a plurality of pulses having an amplitude of 1 or −1. As an example, Table 1 shows pulse positions when the subframe length is 40 samples.

The algebraic codebook 8 divides N (= 40) sample points constituting one subframe into a plurality of pulse system groups 1 to 4, and for each combination obtained by taking one sample point from each pulse system group, A pulse signal having a +1 or -1 pulse at the sample point is sequentially output as a noise component. In this example, basically four pulses are arranged per subframe.

FIG. 19 is an explanatory diagram of sample points assigned to each of the pulse system groups 1 to 4.
(1) Eight sample points 0,5,10,15,20,25,30,35 are assigned to pulse system group 1,
(2) Eight sample points 1,6,11,16,21,26,31,36 are assigned to pulse system group 2,
(3) Eight sample points 2,7,12,17,22,27,32,37 are assigned to pulse system group 3,
(4) 16 sample points 3,4,8,9,13,14,18,19,23,24,28, 29,33,34,38,39 are assigned to the pulse system group 4 .

3 bits are required to represent the sample points of pulse system groups 1 to 3, 1 bit is required to represent the positive and negative of the pulse, and a total of 4 bits are required. Also, to represent the sample points of pulse system group 4 4 bits, 1 bit and 5 bits in total are required to express the positive and negative of the pulse. Accordingly, 17 bits are required to specify the pulse signal output from the noise codebook 8 having the pulse arrangement shown in Table 1, and there are 217 (= 24 × 24 × 24 × 25) types of pulse signals.
As shown in Table 1, the pulse position of each pulse system is limited, and in the algebraic codebook search, the pulse with the smallest error power from the input speech in the playback area is selected from the combinations of pulse positions of each pulse system. Determine the combination. That is, the optimum pitch gain βopt obtained by the adaptive codebook search is set, and the adaptive codebook output PL is multiplied by the gain βopt and input to the adder 11. At the same time, a pulse signal is sequentially input from the algebraic codebook 8 to the adder 11 and the output of the adder is input to the LPC synthesis filter 6 so that the difference between the reproduced signal and the input signal X is minimized. Identify the signal. Specifically, first, a target vector X ′ for algebraic codebook search is generated from the optimal adaptive codebook output PL obtained from the input signal X by adaptive codebook search and the optimal pitch gain _βopt by the following equation.

X ′ ＝ X−βopt × A × PL (5)
In this example, the position and amplitude (positive / negative) of the pulse are expressed in 17 bits as described above, so there are 2 17 combinations. Here, if the kth algebraic code output vector is Ck, the following equation is used in the algebraic codebook search: D = | X′−GC × A × Ck | ² (6)
The code vector Ck that minimizes the evaluation function error power D is obtained. GC is the algebraic codebook gain. In the search of the algebraic codebook, the error power evaluation unit 10 normalizes the mutual value obtained by normalizing the square of the cross-correlation value Rcx between the algebraic composite signal A × Ck and the input signal X ′ with the autocorrelation value Rcc of the algebraic composite signal. A search is made for a combination of a pulse position and a polarity having the largest correlation value (Rcx * Rcx / Rcc).

Next, gain quantization will be described. In the G.729A system, the algebraic codebook gain is not directly quantized, and the adaptive codebook gain Ga (= βopt) and the correction coefficient γ of the algebraic codebook gain Gc are vector-quantized. Here, there is a relationship GC = g ′ × γ between the algebraic codebook gain GC and the correction coefficient γ. g ′ is the gain of the current frame predicted from the logarithmic gain of the past 4 subframes.
In the gain quantization table (not shown) of the gain quantizer 12, 128 (= 27) combinations of the correction code γ for the adaptive codebook gain Ga and the algebraic codebook gain are prepared. The gain codebook search method is as follows: (1) For the adaptive codebook output vector and the algebraic codebook output vector, one set of table values is extracted from the gain quantization table and set in the gain variable sections 13 and 14. (2) The gain variable units 13 and 14 multiply the respective vectors by the gains Ga and Gc and input them to the LPC synthesis filter 6. (3) The error power evaluation unit 10 has the smallest error power with the input signal X. This is done by selecting a combination.

  From the above, the line encoder 15 (1) an LSP code that is an LSP quantization index, (2) a pitch lag code Lopt that is a pitch lag quantization index, (3) an algebraic code that is an algebraic codebook index, ( 4) Multiplex the gain code, which is the gain quantization index, to create circuit data and transmit it to the decoder.

Decoder Configuration and Operation FIG. 20 is a block diagram of a G.729A decoder. The line data sent from the encoder side is input to the line decoding unit 21, and an LSP code, pitch lag code, algebraic code, and gain code are output. The decoder decodes the audio data based on these codes. The operation of the decoder is partly duplicated because the function of the decoder is included in the encoder, but will be briefly described below.
When the LSP code is input, the LSP inverse quantization unit 22 performs inverse quantization and outputs an LSP inverse quantization value. The LSP interpolation unit 23 interpolates the LSP inverse quantization value of the first subframe of the current frame from the LSP inverse quantization value of the second subframe of the current frame and the LSP inverse quantization value of the second subframe of the previous frame. . Next, the parameter inverse conversion unit 24 converts the LSP interpolation value and the LSP inverse quantization value into LPC synthesis filter coefficients, respectively. The G.729A LPC synthesis filter 25 uses the LPC coefficient converted from the LSP interpolation value in the first first subframe, and uses the LPC coefficient converted from the LSP inverse quantization value in the next second subframe. .

The adaptive codebook 26 outputs a pitch signal of subframe length (= 40 samples) from the reading start position indicated by the pitch lag code, and the noise codebook 27 outputs the pulse position and pulse polarity from the reading position corresponding to the algebraic code. To do. The gain dequantization unit 28 calculates an adaptive codebook gain dequantization value and an algebraic codebook gain dequantization value from the input gain code, and sets them in the gain variable units 29 and 30. The adder 31 adds the signal obtained by multiplying the adaptive codebook output by the adaptive codebook gain inverse quantization value and the signal obtained by multiplying the algebraic codebook output by the algebraic codebook gain inverse quantization value, A signal is created, and this sound source signal is input to the LPC synthesis filter 25. Thereby, reproduced sound can be obtained from the LPC synthesis filter 25.
In the initial state, the contents of the adaptive codebook 26 on the decoder side all contain a signal with an amplitude of 0. For each subframe, the oldest signal in time is discarded by the subframe length, while in the current subframe. It operates so as to store the obtained excitation signal in the adaptive codebook 26. In other words, the adaptive codebook 26 of the encoder and decoder is always maintained in the latest state.
The above is the G.729A encoding and decoding scheme. On the other hand, the AMR system uses a basic algorithm called CELP (Code Excited Linear Prediction) as in the G.729A system, and the differences from the G.729A system are as follows.

-Difference in encoding method between G729A method and AMR method FIG. 21 shows the result of comparing the main specifications of the G.729A method and AMR. Although there are eight types of AMR encoding modes in total, the specifications in FIG. 21 are common to all encoding modes. The G729A method and the AMR method have the same sampling frequency (= 8KHz), subframe length (= 5msec), and linear prediction order (= 10th order) of the input signal, but the frame length is different and the subframe per frame The number of frames is different. As shown in FIG. 22, one frame is composed of two 0th to first subframes in the G.729A system, and one frame is composed of four 0th to third subframes in the AMR system.

FIG. 23 shows a comparison result of bit allocation between the G.729A system and the AMR system. The AMR system shows the case of the 7.95 kbit / s mode closest to the G.729A bit rate. As is clear from FIG. 23, the number of bits in the algebraic codebook per subframe (= 17 bits) is the same, but the distribution of the number of bits necessary for other codes is all different. In addition, in the G.729A system, the adaptive codebook gain and the algebraic codebook gain are vector quantized together, so that there is one type of gain code per subframe, but in the AMR system, the adaptive codebook gain and the adaptive codebook gain per subframe are different. Two types of algebraic codebook gain are required.
As explained above, the basic algorithm is common between the G.729A system widely used in VoIP that communicates voice over the Internet and the AMR system adopted in mobile phone systems, but the frame length is different, The number of bits representing the code is different.

・ Voice code conversion With the spread of the Internet and mobile phones, it is considered that the volume of voice calls between Internet users and mobile phone network users will increase in the future. For voice communication between such different communication systems, a voice code converter 53 is required in the middle as shown in FIG. That is, in the speech code conversion device 53, the speech code encoded by the encoder 52 of one communication system 51 according to the first speech encoding method is used in the second speech encoding method used in the other communication system 54. Convert to speech code. If the speech code conversion is performed in this way, the decoder 55 of the second speech coding system of the communication system 54 can correctly reproduce the user 1 speech.

As such a code conversion technique, (1) a tandem connection system that repeats decoding / codes in each system's speech coding system, and (2) a speech code is decomposed into element codes constituting the speech code, There has been proposed a method for individually converting element codes into codes of different speech encoding methods (see Japanese Patent Application No. 2001-75427). FIG. 25 is an explanatory diagram of the latter method.
The encoding method 1 encoder 71a incorporated in the terminal 71 encodes the audio signal emitted by the user A into the encoding method 1 audio code and sends it to the transmission line 71b. The voice code conversion unit 74 converts the voice code of the encoding method 1 input from the transmission line 71b into the voice code of the coding method 2 and sends it to the transmission line 72b. The decoder 72a of the terminal 72 uses the transmission line 72b. The user B can listen to the reproduced voice by decoding the reproduced voice from the voice code of the encoding method 2 input via the user.

  The encoding method 1 includes (1) a first LSP code obtained by quantizing an LSP parameter obtained from a linear prediction coefficient (LPC coefficient) obtained by linear prediction analysis for each frame, and (2) a periodic sound source. A first pitch lag code that specifies an output signal of an adaptive codebook for outputting a signal, and (3) an output signal of an algebraic codebook (or a noise codebook) for outputting a noisy excitation signal. 1 is obtained by quantizing an algebraic code (noise code), and (4) a pitch gain representing the amplitude of the output signal of the adaptive codebook and an algebraic codebook gain representing the amplitude of the output signal of the algebraic codebook. This is a method of encoding an audio signal with a gain code of 1. The encoding method 2 is obtained by quantization by a quantization method different from the first speech encoding method (1) a second LSP code, (2) a second pitch lag code, (3) a second In this method, a speech signal is encoded with an algebraic code (noise code) and (4) a second gain code.

  The speech code conversion unit 74 includes a code separation unit 74a, an LSP code conversion unit 74b, a pitch lag code conversion unit 74c, an algebraic code conversion unit 74d, a gain code conversion unit 74e, and a code multiplexing unit 74f. The code separation unit 74a uses a plurality of component codes necessary for reproducing a speech signal from the speech code of the encoding method 1 input from the encoder 71a of the terminal 1 via the transmission line 71b, that is, (1) An LSP code, (2) a pitch lag code, (3) an algebraic code, and (4) a gain code are separated and input to the code conversion units 74b to 74e, respectively. Each of the code conversion units 74b to 74e converts the input LSP code, pitch lag code, algebraic code, and gain code according to the speech coding method 1 into the LSP code, pitch lag code, algebraic code, and gain code according to the speech coding method 2 (pitch gain). The code multiplexing unit 74f multiplexes the converted codes of the audio coding scheme 2 and sends them to the transmission path 72b.

Data Embedding Technology With the spread of computers and the Internet in recent years, “digital watermark technology” that embeds special data in multimedia content (still images, moving images, audio, audio, etc.) has attracted attention. The digital watermark technology is a technology that embeds other arbitrary information in multimedia content itself such as an image, a moving image, and sound by using the characteristics of human perception and hardly affecting the quality. Such technologies are often intended to protect copyrights by embedding the names of creators and sellers in the content to prevent unauthorized copying and data tampering. It is also used for the purpose of improving the convenience when the user uses the content by embedding the attached information.

  In the field of voice communication, attempts have been made to embed such arbitrary information in a voice code for transmission. FIG. 26 is a conceptual diagram of a voice communication system to which the data embedding technique is applied. When encoding the input speech SP into a speech code, the encoder 81 embeds an arbitrary data sequence DT other than speech in the speech code SCD and transmits it to the decoder 82. At this time, since the data is embedded in the voice code itself without changing the format of the voice code, the information amount of the voice code does not increase. The decoder 82 reads out an arbitrary data sequence embedded in the speech code, and performs normal decoder processing on the speech code to output a reproduced speech SP ′. At this time, since the embedding is performed so that the quality of the reproduced sound SP ′ is hardly affected, the reproduced sound is hardly different from the case where the embedding is not performed. With the above configuration, it is possible to transmit arbitrary data separately from voice without increasing the transmission amount. In addition, a third party who does not know that data is embedded can only recognize normal voice communication.

There are various methods for embedding data. In particular, in the high-compression voice coding system based on the CELP system, several methods for embedding arbitrary information in the coded voice code have been proposed. For example, in a speech coding system that performs coding using an algebraic codebook and an adaptive codebook, a technique for embedding arbitrary data in pitch lag codes and algebraic codes has been proposed. This embedding technique embeds an arbitrary data sequence in a code (pitch lag code, algebraic code) quantized with an algebraic codebook or an adaptive codebook according to a certain rule.
Focusing on the pitch lag code corresponding to the pitch sound source and the algebraic code corresponding to the noise sound source, these gains (pitch gain, algebraic codebook gain) can be regarded as factors indicating the contribution of each code, and the gain is small. The contribution of the corresponding code becomes small. Therefore, a gain is defined as a determination parameter, and when the gain falls below a certain threshold, it is determined that the degree of contribution of the corresponding code is small, and the index of the code is replaced with an arbitrary data series. As a result, it is possible to embed arbitrary data while suppressing the influence of replacement small.

  In the future, it is expected that communication between communication systems to which the data embedding technology as described above is applied will increase. At this time, it is necessary for the speech code conversion apparatus to perform code conversion on the speech code on which data is embedded.

・ Problem 1
FIG. 27 shows the principle of code conversion. FIG. 27 shows a case where encoded data Code1 of the first encoding method is converted into encoded data Code2 of the second encoding method. The code conversion unit 91 includes a first quantization table 92 used when encoding by the first encoding method and a second quantization table 93 used when encoding by the second encoding method, respectively. ing. Although the first quantization table 92 and the second quantization table 93 have different table sizes and table values, FIG. 27 shows a case where the table size is the same as 2 bits for simplification of explanation.

In FIG. 27, the encoded data Code1 (“01” in the figure) of the first encoding method input to the code conversion unit 91 represents the index number of the first quantization table 92. Therefore, a value with the smallest error is selected from the second quantization table 93 for the value (2.0 in the figure) of the first quantization table 92 corresponding to the input Code1, and the index of the second quantization table 93 corresponding thereto is selected. The number ("10" in the figure) is output as encoded data Code2 of the second encoding method. In this way, the code conversion unit 91 compares the conversion source and conversion destination quantization tables and associates index numbers so that the error is minimized.
Here, consider a case where the data sequence of the input code Code1 is arbitrary data (referred to as “01”) embedded by the above-described embedding method. The code conversion unit 91 converts the input data series “01” to “10” in order to perform the same conversion processing as described above. However, in this case, the embedded data sequence changes from “01” to “10” and is not retained, and the decoder on the receiving side of the second encoding method can normally restore the embedded data sequence. I can't.
As described above, in the conventional code conversion method, when an arbitrary data sequence is embedded in the input code, the embedded data sequence cannot be held, and as a result, there is a problem that the embedded data is damaged in the code conversion device. .

・ Problem 2
In the future, as represented by third-generation mobile phone systems, it is expected that communication systems targeting multimedia information such as data communications will become widespread in addition to voice communications. For this reason, communication occurs between a conventional communication system having only a voice line and a communication system having a voice line and other data lines. In such a case, voice communication between users becomes possible by performing mutual conversion of voice codes between the two communication systems with a conventional voice code conversion device. However, since one of the data lines does not have a data line, data communication between users is impossible. As described above, there is a problem that only voice communication can be performed between users between a communication system having only a voice line and a communication system having a voice line and another data line.

  Accordingly, an object of the present invention is to enable both voice communication and data communication between a communication system having only a voice line and a communication system having a data line outside the voice line.

The present invention is a speech code conversion method and a speech code conversion device for converting a first speech code obtained by encoding an input speech by a first speech encoding method into a second speech code by a second speech encoding method.
-Voice code conversion method The voice code conversion method of the present invention includes a step of receiving a first voice code, and when arbitrary data is embedded in the first voice code, the first voice code is converted into a second voice code. And converting, extracting the embedded data from the first speech code, and transmitting the second speech code obtained by the conversion and the extracted data separately to the transmission destination.

  When the data embedding condition is satisfied at the transmission source, the data is embedded in the first speech code by replacing a part of the first speech code with the data. The data embedding condition is monitored by referring to a dequantized value of a predetermined element code constituting the first speech code. If the data embedding condition is satisfied, the embedding is performed from the first speech code. Extract data.

-Speech code conversion device The speech code conversion device of the present invention, the code conversion unit for converting the first speech code to the second speech code, if any data is embedded in the first speech code received from the transmission source, An embedded data extraction unit for extracting embedded data from the first speech code, and a means for separately transmitting the second speech code obtained by the conversion and the extracted data to a transmission destination.
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first audio code by replacing a part of the first audio code with the data, the embedded data extraction unit receives the data from the transmission source. A monitoring unit that monitors whether the data embedding condition is satisfied with reference to a dequantized value of a predetermined element code constituting the one speech code; if the data embedding condition is satisfied, the embedding is performed from the first speech code An extraction unit for extracting data is included.

  According to the present invention, voice information and data information transmitted through a conversion source voice line can be separated and transmitted to a conversion destination voice line and data line.

1 is a first system conceptual diagram of the present invention. 1 is a configuration diagram of a speech code conversion device in a first system of the present invention. FIG. FIG. 5 is another schematic configuration diagram of a speech code conversion device in the first system of the present invention. It is a 2nd system conceptual diagram of this invention. It is a schematic block diagram of the speech code converter in the 2nd system of this invention. It is another schematic block diagram of the speech code converter in the 2nd system of this invention. It is a 3rd system conceptual diagram of this invention. It is a schematic block diagram of the speech code converter in the 3rd system of this invention. It is another schematic block diagram of the speech code converter in the 3rd system of this invention. 1 is a configuration diagram of a speech code conversion device in a first system of the present invention. FIG. FIG. 5 is a configuration diagram of another example of the speech code conversion device in the first system of the present invention. FIG. 6 is a configuration diagram of still another example of the speech code conversion device in the first system of the present invention. It is a block diagram of an algebraic code. FIG. 6 is a block diagram of an embodiment of a speech code conversion device in the second system of the present invention. FIG. 6 is a configuration diagram of another example of the speech code conversion device in the second system of the present invention. FIG. 6 is a block diagram of an embodiment of a speech code conversion device in the third system of the present invention. FIG. 10 is a configuration diagram of another example of the speech code conversion device in the third system of the present invention. It is a block diagram of an encoder of ITU-T recommendation G.729A system. It is explanatory drawing of the sample point allocated to each pulse system group 1-4. [Fig. 7] Fig. 7 is a block diagram of a G.729A decoder. It is a comparison explanatory drawing of the main specifications of G.729A system and AMR. It is a G.729A system and AMR frame structure explanatory drawing. It is a comparison explanatory drawing of the bit allocation in a G.729A system and an AMR system. It is voice code conversion explanatory drawing between different communication systems. It is explanatory drawing of the prior art which converts an audio | voice code into the code | symbol of another audio | voice encoding system. It is a conceptual diagram of the audio | voice communication system to which a data embedding technique is applied. It is a principle diagram of code conversion.

(A) Outline of the present invention
(a) First System FIG. 1 is a conceptual diagram of a first system of the present invention, in which a first encoding speech code SP1 in which arbitrary data DT is embedded is converted into second encoding in which the data DT is embedded. This shows the case of conversion to the system voice code SP2.
A speech code conversion apparatus 103 is provided between the communication system 101 using the first encoding scheme and the communication system 102 using the second encoding scheme. When encoding the input speech SP1, the encoder 104 of the first encoding system in the communication system 101 embeds an arbitrary data series DT other than speech data in the speech code SCD1 and sends it to the transmission path 105. At this time, since the data embedding by the encoder 104 is performed in the speech code itself without changing the speech code format, there is no increase in the information amount of the speech code.
When speech code conversion apparatus 103 receives speech code SCD1 encoded in accordance with the first speech encoding method from encoder 104, speech code conversion device 103 uses the speech code of the second speech encoding method used in communication system 102. It is converted to SCD2 and sent to the transmission line 106. At this time, the speech code conversion device 103 performs speech code conversion without damaging the embedded data.
The decoder 107 of the second coding method in the communication system 102 reads out and outputs an arbitrary data sequence DT embedded in the speech code SCD2, and outputs the reproduced speech SP2 by performing normal decoder processing on the speech code. . At this time, since the embedding is performed so that the quality of the reproduced sound SP2 is hardly affected, the reproduced sound is hardly different from the case where the embedding is not performed.

  FIG. 2 is a block diagram of the code conversion apparatus 103 in the first system of the present invention. The speech code SCD1 encoded at the conversion source according to the first encoding method and embedded with the data DT is input to the code conversion unit 111 and the embedded data extraction unit 112 in order in units of frames. The code converting unit 111 has the same configuration as that shown in FIG. 25, and converts the first encoding speech code SCD1 into the second encoding speech code SCD2 '. The embedded data extraction unit 112 extracts the data DT embedded in the speech code SCD1 and outputs it to the data embedding unit 113. The data extraction method by the embedded data extraction unit 112 is the same as the data extraction method of the first encoding decoder. The data embedding unit 113 embeds the data DT in units of frames in the speech code SCD2 ′ when the second encoding method speech code SCD2 ′ converted by the code conversion unit 111 and the data DT extracted from the speech code SCD1 are input. And output as voice code SCD2. The data embedding method by the data embedding unit 113 is the same as the data embedding method of the encoder of the second encoding method.

  FIG. 3 is another block diagram of the code conversion device 103 in the first system of the present invention. The same parts as those of the code conversion device of FIG. The code conversion device 103 adaptively extracts embedded data DT from the speech code SCD1 based on the properties of the speech code and embeds the data DT in the speech code SCD2 ′. For example, as described in the section of the prior art, the encoder of the first encoding scheme has a corresponding code (pitch lag code, algebraic code) if the gain (pitch gain, algebraic codebook gain) is less than a certain threshold. Assuming that the contribution to speech is small, the index of the code is replaced with an arbitrary data sequence DT. For this reason, a section in which data is embedded and a section in which data is not embedded are generated according to the gain, in the first encoding speech code SCD1.

  The embedding determination unit 121 determines whether another data is embedded in the code in units of frames or subframes based on the gain of the audio code SCD1, and when determining that the data is embedded, SW1 is closed and the voice code SCD1 is input to the embedded data extraction unit 112. The embedded data extraction unit 112 extracts data from the audio code SCD1 and inputs the data to the data holding unit 122 having a FIFO buffer configuration. The FIFO buffer is a first-in first-out buffer.

  The embedding determination unit 123 determines whether to embed data in the audio code in units of frames or subframes based on the gain of the audio code SCD2 ′ of the second encoding method output from the code conversion unit 111, and embeds the data If it is determined, the switch SW2 is closed, and the data holding unit 122 inputs the held data from the oldest to the data embedding unit 113 in units of frames or subframes. As a result, the data embedding unit 113 embeds the data DT output from the data holding unit 122 in the second encoding speech code SCD2 ′ in units of frames and outputs the result as the speech code SCD2.

  Each embedding determination method may be the same as the method used in each encoding method. When the embedding determination unit 121 and the embedding determination unit 123 have different embedding determination methods, the closing timings of the switches SW1 and SW2 do not necessarily match. Further, even when the embedding determination method is the same, the same phenomenon occurs because the speech code differs before and after conversion due to the conversion error of the speech code conversion unit 111. The data holding unit 122 in FIG. 3 has a function of absorbing the difference in the switching timing and preventing data loss.

  That is, when the conversion destination is not the embedding target section, the data DT extracted from the first speech code SCD1 by the data holding unit 122 is temporarily held. Conversely, when the conversion source is not the embedding target section, the data held in the data holding unit 122 is extracted and embedded in the second speech code SCD2 ′. Further, when the code data size to be embedded at the conversion source is larger than the conversion destination, only the amount of data that can be embedded is embedded, and the remaining data is temporarily held by the data holding unit 122. When the data holding number of the data holding unit 122 decreases, data conversion destination data embedding is temporarily stopped to restore the data holding number. As described above, the difference in switching timing is absorbed to prevent data loss.

(b) Second System FIG. 4 is a conceptual diagram of the second system of the present invention. The conversion source communication system 101 has a voice line 105 and a data line 108, and the conversion destination communication system 102 has only a voice line 106. It shows the case of having. As shown in the figure, the encoder 104 of the first encoding method in the communication system 101 encodes the input speech SP1 into a speech code SCD1 and sends the speech code to the speech line 105, and any data other than the speech code The sequence DT is transmitted to the data line 108. Actually, the voice code SCD and the data series DT are time-division multiplexed and transmitted to the multiplex line, separated at an appropriate location, and input to the voice code converter 103. Thus, the voice code SCD1 is input from the voice line 105 and the data DT is input from the data line 108 to the voice code conversion apparatus 103. The voice code converter 103 converts the voice code SCD1 of the first coding system into a voice code of the second coding system and embeds the data DT in the voice code as the voice code SCD2 to the destination communication system 102. 106 is transmitted.

The decoder 107 of the second encoding method in the communication system 102 reads out and outputs an arbitrary data series DT embedded in the speech code, and performs a normal decoder process on the speech code to output the reproduced speech SP2. At this time, since the embedding is performed so that the quality of the reproduced sound SP2 is hardly affected, the reproduced sound is hardly different from the case where the embedding is not performed.
FIG. 5 is a block diagram of the code conversion device 103 in the second system of the present invention. The same reference numerals are given to the same parts as those of the code conversion device in the first system of FIG. The difference is that (1) the data DT is input via a route different from that of the speech code SCD1, and (2) there is no embedded data extraction unit, and the embedded data DT is directly input to the data embedding unit 113.

  The communication system that is the conversion source time-division-multiplexes the voice code SCD1 and data DT encoded according to the first encoding method and sends them to the multiplex line 200. The line separation unit 201 separates these voice codes SCD1 and data DT. Are input to the code conversion device 103 via the voice line 105 and the data line 108. The data embedding unit 113, when the audio code SCD2 'of the second encoding method converted by the code conversion unit 111 and the data DT are input, embeds the data DT in units of frames in the audio code SCD2', and as the audio code SCD2 The data is sent to the voice line 106.

FIG. 6 is another configuration diagram of the code conversion apparatus 103 in the second system of the present invention. The same reference numerals are given to the same parts as those of the code conversion apparatus in the first system of FIG. 3 differs from FIG. 3 in that (1) the data DT is input via a different path from the speech code SCD1, and (2) there is no embedded determination unit and embedded data extraction unit, and the embedded data DT is directly sent to the data holding unit 122. It is a point to input.
The communication system that is the conversion source time-division-multiplexes the voice code SCD1 and data DT encoded according to the first encoding method and sends them to the multiplex line 200. The line separation unit 201 separates these voice codes SCD1 and data DT. Are input to the code conversion device 103 via the voice line 105 and the data line 108.

  The code conversion device 103 adaptively embeds the data DT in the speech code SCD ′ based on the nature of the speech code. That is, the code conversion unit 111 converts the first encoding speech code SCD1 into the second encoding speech code SCD2 ', and the FIFO buffer configuration data holding unit 122 holds the input data DT. The embedding determination unit 123 determines whether to embed data in the audio code in units of frames or subframes based on the second encoding method audio code SCD2 ′ output from the code conversion unit 111, and determines to embed the data Then, the switch SW2 is closed, and the data holding unit 122 inputs the held data from the oldest to the data embedding unit 113 in units of frames or subframes. As a result, the data embedding unit 113 embeds the data DT output from the data holding unit 122 in the second encoding method voice code SCD2 ′ in units of frames, and sends it as the voice code SCD2 to the voice line 106.

(c) Third System FIG. 7 is a conceptual diagram of the third system of the present invention. Contrary to the second system, the conversion source communication system 101 has only the voice line 105 and the conversion destination communication system. A case where 102 has a voice line 106 and a data line 109 is shown.
The encoder 104 of the first encoding system in the communication system 101 encodes the input speech SP1, embeds an arbitrary data sequence DT other than speech data in the code, and sends it as a speech code SCD1 to the speech line 105. The voice code conversion device 103 converts the voice code SCD1 of the first coding system into the voice code SCD2 of the second coding system, and extracts the data DT embedded in the voice code SCD1, these voice codes SCD2, Data DT is transmitted to each of the lines 106 and 109. The communication system 102 outputs the data input via the data line 109 and also decodes the speech code SCD2 by the decoder 107 and outputs the reproduced speech SP2. Actually, the voice code SCD2 and the data DT are time-division multiplexed at an appropriate place, transmitted to the communication system 102, and separated by the communication system.

FIG. 8 is a block diagram of the code conversion apparatus 103 in the third system of the present invention. The same reference numerals are given to the same parts as those of the code conversion apparatus in the first system of FIG. The difference is that (1) there is no data embedding unit, and the data DT extracted by the embedded data extraction unit 112 is not embedded in the audio code SCD2 of the second encoding method output from the code conversion unit 111, (2) the data DT Is transmitted separately from the audio code SCD2 of the second encoding method.
The speech code SCD1 encoded at the conversion source according to the first encoding method and embedded with the data DT is input to the code conversion unit 111 and the embedded data extraction unit 112 in order in units of frames. The code conversion unit 111 converts the voice code SCD1 of the first coding system into the voice code SCD2 of the second coding system and sends it to the voice line 106. The embedded data extraction unit 112 extracts the data DT embedded in the speech code SCD1 and sends it to the data line 109. The line multiplexing unit 203 time-division-multiplexes the voice code SCD2 and the data DT input via the voice line 106 and the data line 109 and sends them to the multiplexing line 204.

  FIG. 9 is another configuration diagram of the code conversion apparatus 103 in the third system of the present invention. The same reference numerals are given to the same parts as those of the code conversion apparatus in the first system of FIG. 3 differs from FIG. 3 in that (1) there is no data holding unit, embedding determination unit, and data embedding unit, (2) data DT is not embedded in the audio code SCD2 output from the code conversion unit 111, and (3) data The point is that DT is sent separately from the voice code SCD2.

  When the gain (pitch gain, algebraic codebook gain) is below a certain threshold, the encoder of the communication system on the transmitting side assumes that the contribution of the corresponding code (pitch lag code, algebraic code) to speech is small, and Replace the index with an arbitrary data series DT. As a result, a section in which data is embedded and a section in which data is not embedded are generated in the speech code SCD1 of the first encoding method. The embedding determination unit 121 determines whether another data is embedded in the code in units of frames or subframes based on the gain obtained from the audio code SCD1, and when determining that the data is embedded, The switch SW1 is closed and the voice code SCD1 is input to the embedded data extraction unit 112. The embedded data extraction unit 112 extracts embedded data from the voice code SCD 1 and sends it to the data line 109. In parallel with the above, the voice code conversion unit 111 converts the voice code SCD1 of the first coding system into the voice code SCD2 of the second coding system and sends it to the voice line 106. The line multiplexing unit 203 time-division multiplexes the voice code SCD2 and the data DT inputted via the voice line 106 and the data line 109 and sends them to the multiplexing line 204.

(B) Example in the first system
(a) First Embodiment FIG. 10 is a configuration diagram of a code conversion device in the first system of the present invention, and shows a configuration in the case of embedding control.
The first embodiment shows an example in which an AMR speech code in which arbitrary data is embedded is converted into a G.729A speech code without losing the embedded data. Further, in the first embodiment, the AMR encoder of the conversion source embeds arbitrary data in all 17 bits / subframes assigned to the algebraic code if the algebraic codebook gain is smaller than the set value, If the book gain is larger than the set value, the original algebraic code data is embedded. Similarly, the conversion destination G.729A encoder also embeds data in all 17 bits assigned to the algebraic code according to the algebraic codebook gain.

  In FIG. 10, when the line data bst1 (m), which is the AMR encoder output of the m-th frame, is input to the code separation unit 114 through the terminal 1, the code separation unit 114 converts the line data bst1 (m) into an AMR element. It is separated into codes (LSP code 1, pitch lag code 1, pitch gain code 1, algebraic code 1, algebraic gain code 1). Then, these element codes are input to each code converter (LSP code converter 111a, pitch lag code converter 111b, pitch gain code converter 111c, algebraic gain code converter 111d, algebraic code converter 111e) in the code converter 111. To do. Each of the code conversion units 111a to 111e converts the code of the first coding method into the code of the second coding method, but the operation thereof is the same as that of the prior art, and thus description thereof is omitted here. Below, only the part relevant to data embedding is demonstrated.

  The embedding determination unit 121 obtains an algebraic gain inverse quantization value (algebraic gain) from the algebraic gain code 1, and switches the switch SW1 according to the gain value. That is, if the AMR algebraic gain value is smaller than a certain threshold, it is determined that there is embedded data, the switch SW1 is closed, and the algebraic code 1 is input to the embedded data extraction unit 112. The embedded data extraction unit 112 extracts embedded data Dcode included in the algebraic code and outputs it to the data holding unit 122. In this embodiment, since data is embedded in all AMR algebraic codes (17 bits / subframe), a 17-bit data series is cut out as embedded data Dcode. A data holding unit 122 having a FIFO configuration stores and holds the input data series in chronological order.

  On the other hand, the embedding determination unit 123 obtains an algebraic gain dequantized value from the algebraic gain code 2 of the converted G.729A input from the algebraic gain code converter 111d, and switches the switch SW2 according to the gain value. Do. That is, when the algebraic gain value of G.729A is smaller than a certain threshold value, it is determined that data is embedded, the switch SW2 is closed, and data is input from the data holding unit 122 to the data embedding unit 113. In this embodiment, the data holding unit 122 inputs 17-bit data to the data embedding unit 113 in order to embed data in all G.729A algebraic codes (17 bits / subframe). The data embedding unit 113 embeds data input to 17 bits assigned to the algebraic code 2. That is, all the algebraic codes (17 bits) of G.729A are replaced with the data series (17 bits).

The algebraic code 2 in which the data is embedded is multiplexed by the code multiplexing unit 115 together with other element codes, and is output from the terminal 2 as line data bst2 (n) of the nth frame of G.729A including the embedded data. The
According to the first embodiment, when arbitrary data is embedded in the algebraic code in the AMR speech code bst1 (m), the data is embedded in the algebraic code of G.729A without damaging the embedded data. It can be converted into a voice code bst2 (n). As a result, data communication can be performed in addition to voice communication without changing the voice format between AMR and G.729A.
Although the conversion from AMR to G.729A has been described above, the configuration of the portion related to data extraction and data embedding according to the first embodiment can also be applied at the time of reverse conversion from G.729A to AMR.

(b) Second Embodiment FIG. 11 is another configuration diagram of the code conversion device in the first system of the present invention, showing a configuration in the case of embedding control, and in the same part as the first embodiment of FIG. Are given the same reference numerals. The difference is that in the first embodiment, if the algebraic gain is smaller than the set value, arbitrary data is embedded in all 17 bits / subframes assigned to the algebraic code. If the gain is smaller than the set value, arbitrary data is embedded in all 8 bits or 5 bits / subframe assigned to the pitch lag code.

  The embedding determination unit 121 obtains a pitch gain inverse quantization value (pitch gain) from the pitch gain code 1, and switches the switch SW1 according to the gain value. That is, if the AMR pitch gain value is smaller than a certain threshold value, it is determined that there is embedded data, the switch SW1 is closed, and the pitch lag code 1 is input to the embedded data extraction unit 112. The embedded data extraction unit 112 extracts embedded data Dcode included in the pitch lag code and outputs the extracted data to the data holding unit 122. In this embodiment, since data is embedded in all AMR pitch lag codes (8 bits or 6 bits / subframe), an 8-bit or 6-bit data series is cut out as embedded data Dcode as it is. A data holding unit 122 having a FIFO configuration stores and holds the input data series in chronological order.

  On the other hand, the embedding determination unit 123 obtains a pitch gain inverse quantization value from the converted G.729A pitch gain code 2 input from the pitch gain code conversion unit 111c, and switches the switch SW2 according to the gain value. Do. That is, when the pitch gain value of G.729A is smaller than a certain threshold value, it is determined that data is embedded, the switch SW2 is closed, and data is input from the data holding unit 122 to the data embedding unit 113. In this embodiment, since data is embedded in all pitch lag codes (8 bits or 5 bits / subframe) of G.729A, the data holding unit 122 stores data of 8 bits or 5 bits according to the subframe. To enter. The data embedding unit 113 embeds data input in 8 bits or 5 bits assigned to the pitch lag code 2.

The pitch lag code 2 in which the data is embedded is multiplexed by the code multiplexing unit 115 together with other element codes, and is output from the terminal 2 as line data bst2 (n) of the nth frame of G.729A including the embedded data. The
According to the second embodiment, when arbitrary data is embedded in the pitch lag code of the AMR speech code bst1 (m), the voice is embedded in the pitch lag code of G.729A without damaging the embedded data. It can be converted into the code bst2 (n). As a result, it is possible to perform data communication in addition to voice communication without changing the voice format between AMR (7.95 kbps) and G.729A.
Although the conversion from AMR to G.729A has been described above, the configuration of the part related to data extraction and data embedding can also be applied at the time of reverse conversion from G.729A to AMR and other code conversions. .

(C) Third Embodiment FIG. 12 is another configuration diagram of the code conversion apparatus in the first system of the present invention, and shows a configuration when embedding control is not performed. In the third embodiment, an example in which an AMR speech code is converted into a G.729A speech code without losing embedded data is shown. The AMR speech code is 20 msec per frame with reference to FIGS. 21 to 23, and includes four subframes every 5 msec, and each subframe has a 17-bit algebraic code. On the other hand, a G.729A speech code is 10 msec per frame, has two subframes every 5 msec, and each subframe has a 17-bit algebraic code. In both AMR and G729A, the pulse positions m0 to m3 and the polarities s0 to s3 of the four pulse systems (see Table 1) are expressed by these 17 bits. Bit assignments for the pulse positions m0 to m3 and the polarities s0 to s3 are as shown in FIG.

  In the third embodiment, the AMR encoder of the conversion source embeds data Dcode in 5 bits of m3 and s3 indicating the pulse position and polarity of the fourth path system, for example. The embedded data extraction unit 112 always extracts the embedded data Dcode included in the algebraic code 1 and inputs it to the data embedding unit 113. The data embedding unit 113 embeds the data Dcode input in 5 bits m3 and s3 out of 17 bits assigned to the algebraic code 2. The algebraic code 2 in which the data is embedded is multiplexed by the code multiplexing unit 115 together with other element codes, and is output from the terminal 2 as line data bst2 (n) of the nth frame of G.729A including the embedded data. The

As described above, according to the first system, the embedded data DT is once extracted from the speech code SCD1 of the first encoding method as the conversion source, and the data DT is added to the speech code SCD2 'of the second encoding method after the code conversion. By embedding again, the data DT embedded in the speech code SCD1 can be converted into the speech code SCD2 embedded in the data without damaging the data DT.
According to the first system, when adaptive embedding control is performed at the conversion source and the conversion destination, due to a difference in the embedding control method of each coding method or due to a conversion error in the conventional speech code conversion unit By absorbing the difference between the timing of data extraction and embedding that occurs by the data holding unit, the data embedded in the audio code SCD1 can be converted into the audio code SCD2 embedded in the same without damaging the data.
In addition, according to the first system, between voice communication systems having a voice line to which the data embedding technology is applied, the embedded data is not lost and the voice code format is not changed via the voice line. It is possible to perform both voice and data communication.

(C) Embodiment of the second system of the present invention
(a) First Example FIG. 14 is a configuration diagram of a speech code conversion device in the second system of the present invention, in which data Dcode is not embedded in speech code bst1 (m), and the data is a speech code. It is different from the first system embodiment in that it is input to the speech code conversion device via a separate line. The line multiplexing unit 201 separates the voice code bst1 (m) and the data Dcode from the multiplexed data received via the multiple line 200, and inputs the voice code bst1 (m) from the terminal 1 to the code separation unit 114. Data Dcode is directly input to the data holding unit 122.
The code separation unit 114 separates the line data bst1 (m) into AMR element codes (LSP code 1, pitch lag code 1, pitch gain code 1, algebraic code 1, algebraic gain code 1), and converts these element codes. Input to each code conversion unit (LSP code conversion unit 111a, pitch lag code conversion unit 111b, pitch gain code conversion unit 111c, algebraic gain code conversion unit 111d, algebraic code conversion unit 111e) in unit 111. Each code conversion unit 111a to 111e converts the code of the first coding method into the code of the second coding method.

  The embedding determination unit 123 obtains an algebraic gain inverse quantization value from the algebraic gain code 2 of G.729A after conversion input from the algebraic gain code conversion unit 111d, and switches the switch SW2 according to the gain value. That is, when the algebraic gain value of G.729A is smaller than a certain threshold value, it is determined that data is embedded, the switch SW2 is closed, and data is input from the data holding unit 122 to the data embedding unit 113. The data embedding unit 113 embeds data input to 17 bits assigned to the algebraic code 2. The algebraic code 2 in which the data is embedded is multiplexed by the code multiplexing unit 115 together with other element codes, and is output from the terminal 2 as line data bst2 (n) of the nth frame of G.729A including the embedded data. The

  According to this embodiment, in the communication system on the AMR side, when there is a data line in addition to the voice line, the voice code bst1 (m) and the data Dcode input separately via the voice line and the data line are Can be converted to a voice code bst2 (n) embedded in the signal and transmitted to a communication system on the G.729A side having only a voice line. As a result, a communication system capable of voice communication and data communication, such as a third generation mobile phone system (AMR is adopted as a voice encoding method), a communication system having only a voice line, for example, a conventional second generation that performs only voice communication. In addition to voice communication, data communication can be performed with the mobile phone system (G.729A).

(a) Second Embodiment FIG. 15 is another configuration diagram of the speech code conversion apparatus in the second system of the present invention, and shows a configuration when embedding control is not performed. In the second embodiment, the data Dcode is not embedded in the speech code bst1 (m), and the data is input to the speech code conversion device through a separate line from the speech code. In addition, since the algebraic code of G729A represents each pulse position m0 to m3 and the polarity s0 to s3 of the four pulse systems by 17 bits, in the second embodiment, for example, m3 indicating the pulse position and polarity of the fourth path system , Data Dcode is embedded in 5 bits of s3.
The line multiplexing unit 201 separates the voice code bst1 (m) and the data Dcode from the multiplexed data received via the multiple line 200, and inputs the voice code bst1 (m) from the terminal 1 to the code separation unit 114. Data Dcode is directly input to the data embedding unit 113.
The code separation unit 114 separates the line data bst1 (m) into AMR element codes (LSP code 1, pitch lag code 1, pitch gain code 1, algebraic code 1, algebraic gain code 1), and converts these element codes. Input to each code conversion unit (LSP code conversion unit 111a, pitch lag code conversion unit 111b, pitch gain code conversion unit 111c, algebraic gain code conversion unit 111d, algebraic code conversion unit 111e) in unit 111. Each code conversion unit 111a to 111e converts the code of the first coding method into the code of the second coding method.

  The data embedding unit 113 embeds the data Dcode input in 5 bits m3 and s3 out of 17 bits assigned to the algebraic code 2. The algebraic code 2 in which the data is embedded is multiplexed by the code multiplexing unit 115 together with other element codes, and is output from the terminal 2 as line data bst2 (n) of the nth frame of G.729A including the embedded data. The

As described above, according to the second system, it is possible to perform voice communication and data communication without changing the voice code format from a communication system having a data line separately from a voice line to a communication system having only a voice line.
Although the conversion from AMR to G.729A has been described above, the present invention can also be applied during reverse conversion from G.729A to AMR and other code conversions. In the above description, the case where data is embedded in an algebraic code in accordance with an algebraic gain has been described. However, data may be embedded in a pitch lag code in accordance with a pitch gain.

(D) Third system of the present invention
(a) First Embodiment FIG. 16 is a configuration diagram of a speech code conversion apparatus in the third system of the present invention, and shows a configuration when adaptively extracting embedded data. In this embodiment, the first encoding scheme is G.729A, the second encoding scheme is AMR (7.95 kbps), and the code converter converts the G.729A speech code into an AMR speech code. At the same time, the data embedded in the G.729A speech code is extracted and transmitted separately from the speech code. Also, the conversion source G.729A encoder (not shown) embeds arbitrary data in all 17 bits / subframes assigned to the algebraic code if the algebraic gain is smaller than the set value, and the algebraic gain is If it is larger than the set value, the original algebraic code data is embedded.

  When the line data bst1 (m), which is the G.729A encoder output of the m-th frame, is input to the code separation unit 114 through the terminal 1, the code separation unit 114 converts the line data bst1 (m) into the G.729A element. It is separated into codes (LSP code 1, pitch lag code 1, pitch gain code 1, algebraic code 1, algebraic gain code 1). Then, these element codes are input to each code converter (LSP code converter 111a, pitch lag code converter 111b, pitch gain code converter 111c, algebraic gain code converter 111d, algebraic code converter 111e) in the code converter 111. To do. Each code conversion unit 111a to 111e converts the G.729A code into an AMR code, and the code multiplexing unit 115 multiplexes each AMR code and inputs it to the line multiplexing unit 203 as a voice code bst2 (n).

In parallel with the above, the embedding determination unit 121 obtains an algebraic gain dequantized value (algebraic gain) from the algebraic gain code 1, and switches the switch SW1 according to the gain value. That is, when the algebraic gain value of G.729A is smaller than a certain threshold value, it is determined that there is embedded data, the switch SW1 is closed, and the algebraic code 1 is input to the embedded data extraction unit 112. The embedded data extraction unit 112 extracts embedded data Dcode included in the algebraic code and inputs it to the line multiplexing unit 203. Since data is embedded in all G.729A algebraic codes (17 bits / subframe), a 17-bit data series is cut out as embedded data Dcode and input to the line multiplexing unit 203 as it is.
The line multiplexing unit 203 multiplexes the input voice code bst2 (n) and the data Dcode and sends them to the multiplexing line 204.

(b) Second Embodiment FIG. 17 is another configuration diagram of the speech code conversion apparatus in the third system of the present invention, in which embedded data is always inserted in an algebraic code. In this embodiment, the first encoding method is G.729A, the second encoding method is AMR (7.95 kbps), and the speech code converter converts the G.729A speech code into an AMR speech code. The data embedded in the G.729A speech code is extracted and transmitted on a separate line from the speech code. Also, the G.729A encoder as the conversion source embeds data Dcode in 5 bits (see FIG. 13) of algebraic codes m3 and s3.

  When the line data bst1 (m), which is the G.729A encoder output of the m-th frame, is input to the code separation unit 114 through the terminal 1, the code separation unit 114 converts the line data bst1 (m) into the G.729A element. It is separated into codes (LSP code 1, pitch lag code 1, pitch gain code 1, algebraic code 1, algebraic gain code 1). Then, these element codes are input to each code converter (LSP code converter 111a, pitch lag code converter 111b, pitch gain code converter 111c, algebraic gain code converter 111d, algebraic code converter 111e) in the code converter 111. To do. Each code conversion unit 111a to 111e converts the G.729A code into an AMR code, and the code multiplexing unit 115 multiplexes each AMR code and inputs it to the line multiplexing unit 203 as a voice code bst2 (n).

In parallel with the above, the embedded data extraction unit 112 extracts embedded data Dcode included in the algebraic code and inputs it to the line multiplexing unit 203. Since data is embedded in the algebraic code m3, s3 bit position of G.729A, the data is cut out and input to the line multiplexing unit 203 as embedded data Dcode. The line multiplexing unit 203 multiplexes the input voice code bst2 (n) and the data Dcode and sends them to the multiplexing line 204.
According to the third system, voice communication and data communication can be performed without changing the voice code format from a communication system having only a voice line to a communication system having a data line separately from the voice line.
Although the conversion from G.729A to AMR has been described above, the present invention can also be applied to other code conversions. In the above description, the case where data is embedded in an algebraic code in accordance with an algebraic gain has been described. However, data may be embedded in a pitch lag code in accordance with a pitch gain.

・ Supplementary Note (Supplementary Note 1) In the speech code conversion method for converting the first speech code obtained by encoding the input speech by the first speech coding method into the second speech code by the second speech coding method, the first speech code is arbitrarily selected Is embedded, the first speech code is converted into a second speech code, embedded data is extracted from the first speech code, and the extracted data is extracted into the second speech code obtained by the conversion. A speech code conversion method characterized by embedding.
(Appendix 2)
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first speech code by replacing a part of the first speech code with the data, a predetermined constituent of the received first speech code Additional information characterized by monitoring whether the data embedding condition is satisfied with reference to an inverse quantized value of an element code, and extracting the embedded data from the first speech code if the data embedding condition is satisfied The speech code conversion method according to 1.
(Appendix 3)
3. The speech code conversion method according to appendix 2, wherein the extracted embedded data is stored in a data holding unit, and the embedded data is read from the data holding unit and embedded in the second speech code.
(Appendix 4)
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first audio code by replacing a part of the first audio code with the data, the first audio code received from the transmission source is configured. Monitoring whether the data embedding condition is satisfied with reference to a dequantized value of the predetermined element code, extracting the embedded data from the first speech code if the data embedding condition is satisfied, and extracting Holding the embedded data, and monitoring whether the data embedding condition is satisfied with reference to the inverse quantization value of the predetermined element code constituting the second speech code obtained by the conversion, The speech code conversion method according to appendix 1, wherein the data is embedded in the second speech code by replacing a part of the second speech code with the retained data.
(Appendix 5)
In the voice code conversion method for converting the first voice code obtained by encoding the input voice by the first voice coding method into the second voice code by the second voice coding method, the first voice code and the data are separately transmitted from the transmission source. A speech code conversion method comprising: receiving, converting a first speech code into a second speech code, embedding the data in a second speech code obtained by the conversion, and transmitting the data to a transmission destination.
(Appendix 6)
The supplementary note 5, wherein the first voice code is received from the voice line, the data is received from the data line, and the second voice code in which the data is embedded is transmitted to the destination via the voice line. Voice code conversion method.
(Appendix 7)
When the received data is stored in a data holding unit, and monitored whether the data embedding condition is satisfied with reference to a dequantized value of a predetermined element code constituting the second speech code, 6. The speech code conversion method according to appendix 5, wherein the data is embedded in the second speech code by replacing a part of the second speech code with the data stored in the data holding unit.
(Appendix 8)
In a speech code conversion method for converting a first speech code obtained by encoding an input speech using a first speech encoding method into a second speech code using a second speech encoding method, the first speech code is received, and the first speech code is received. When arbitrary data is embedded in the code, the first speech code is converted into a second speech code, and the embedded data is extracted from the first speech code, and the second speech code obtained by the conversion and the A speech code conversion method, wherein the extracted data is separately transmitted to a transmission destination.
(Appendix 9)
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first speech code by replacing a part of the first speech code with the data, a predetermined constituent of the received first speech code Monitoring whether the data embedding condition is satisfied by referring to an inverse quantization value of an element code, and extracting the embedded data from the first speech code if the data embedding condition is satisfied The voice code conversion method according to attachment 8.
(Appendix 10)
In a speech code conversion apparatus that converts a first speech code obtained by encoding an input speech using a first speech encoding method into a second speech code using a second speech encoding method, arbitrary data is embedded in the first speech code. A code conversion unit that converts the first speech code into a second speech code, an embedded data extraction unit that extracts embedded data from the first speech code, and the extracted data in the second speech code obtained by the conversion A speech code conversion device comprising a data embedding unit to be embedded.
(Appendix 11)
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first speech code by replacing a part of the first speech code with the data, the embedded data extraction unit receives the received first speech code Monitoring whether or not the data embedding condition is satisfied with reference to an inverse quantization value of a predetermined element code constituting the code, and extracting the embedded data from the first speech code if the data embedding condition is satisfied; The speech code conversion device according to supplementary note 10, characterized by the above.
(Appendix 12)
And a data holding unit for storing the extracted embedded data. The embedded data extracting unit stores the extracted embedded data in the data holding unit, and the data embedding unit stores the embedded data from the data holding unit. The speech code conversion device according to appendix 11, wherein the speech code conversion device is read out and embedded in the second speech code.
(Appendix 13)
The embedded data extraction unit monitors whether a data embedding condition is satisfied with reference to an inverse quantization value of a predetermined element code constituting the second speech code. 13. The speech code conversion device according to appendix 12, wherein the data is embedded in the second speech code by replacing a part of the second speech code with the stored data.
(Appendix 14)
In a speech code conversion device for converting a first speech code obtained by encoding an input speech using a first speech encoding method into a second speech code using a second speech encoding method, the first speech code and data are separately transmitted from a transmission source. Receiving means for receiving, a code converting unit for converting the first speech code into the second speech code, and a data embedding unit for embedding the data in the second speech code obtained by the conversion and transmitting it to the transmission destination A speech code conversion device.
(Appendix 15)
The speech code conversion apparatus further includes a data holding unit that stores the data, and the data embedding unit refers to a dequantized value of a predetermined element code constituting the second speech code and satisfies a data embedding condition. Means for monitoring whether or not, if satisfied, means for embedding data in the second voice code by replacing a part of the second voice code with the data stored in the data holding unit. The speech code converter according to appendix 14.
(Appendix 16)
In a speech code conversion apparatus that converts a first speech code obtained by encoding an input speech using a first speech encoding method into a second speech code using a second speech encoding method, the first speech code received from a transmission source is arbitrarily set. When data is embedded, a code conversion unit that converts the first speech code into a second speech code, an embedded data extraction unit that extracts embedded data from the first speech code, and a second speech code obtained by the conversion And a means for separately transmitting the extracted data to a transmission destination.
(Appendix 17)
When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first audio code by replacing a part of the first audio code with the data, the embedded data extraction unit receives the data from the transmission source. The data embedding condition is monitored by referring to a dequantized value of a predetermined element code constituting one voice code, and if the data embedding condition is satisfied, the embedded data is obtained from the first voice code. The speech code conversion device according to supplementary note 16, characterized in that it is extracted.

As described above, according to the present invention, the embedding data is once extracted from the speech code of the first encoding method as the conversion source, and the data is re-embedded in the speech code of the second encoding method after the code conversion. Without damaging the data embedded in the speech code of one encoding method, it can be converted into the speech code of the second encoding method embedded with the same data.
Further, according to the present invention, when adaptive embedding control is performed between the conversion source and the conversion destination, data generated due to a difference in the embedding control method of each encoding method or due to a conversion error in the conventional speech code conversion unit The difference between the timing of extraction and embedding is absorbed by the data holding unit, so that the data embedded in the first encoding speech code is converted to the second encoding speech code embedded with the same data without damaging the data. can do.

Further, according to the present invention, between voice communication systems having a voice line to which the data embedding technique is applied, voice and voice can be transmitted via the voice line without damaging the embedded data and without changing the voice code format. Both data communications can be performed.
Further, according to the present invention, when the voice code and data of the first encoding method are input to the voice code conversion unit through separate lines from the conversion source system, the voice code conversion unit performs the second code after code conversion. By embedding the data in the voice code of the encoding method, it becomes possible to transmit to the conversion destination only by the voice line.
Further, according to the present invention, when a speech code of the first encoding method in which arbitrary data DT is embedded via a speech line is input from the conversion source system, the speech code conversion unit converts the speech code from the speech code. The embedded data is extracted and sent to the data line, and the voice code of the first coding method is converted into the voice code of the second coding method and sent to the voice line. It is possible to transmit the voice information and the data information separately on the voice line and the data line to be converted.

Also, according to the present invention, it is possible to perform voice communication and data communication between a communication system having only a voice line and a communication system having a data line separately from the voice line, without changing the voice code format. .
In the future, with the spread of multimedia information communication, data will be used in communication between various communication systems, such as communication between conventional mobile phone systems and next-generation mobile phone systems, or communication between mobile systems such as VoIP and mobile phones. Since there is a high need for a technique using both the embedding technique and the voice code conversion technique, the effect of the present invention is great.

DESCRIPTION OF SYMBOLS 101 Conversion-source communication system 102 Conversion-destination communication system 103 Speech code conversion apparatus 104 First encoding system encoder 105 Audio line 106 Audio line 107 Second encoding system decoder 108 Data line

Claims (4)

In the speech code conversion method for converting the first speech code obtained by encoding the input speech by the first speech encoding method into the second speech code by the second speech encoding method,
Receiving the first voice code,
When arbitrary data is embedded in the first speech code, the first speech code is converted into a second speech code, and the embedded data is extracted from the first speech code,
Sending the second voice code obtained by the conversion and the extracted data separately to the destination,
A speech code conversion method characterized by the above. At the transmission source, when the data embedding condition is satisfied, the data is embedded in the first speech code by replacing a part of the first speech code with the data,
Monitoring whether the data embedding condition is satisfied with reference to an inverse quantization value of a predetermined element code constituting the received first speech code;
If the data embedding condition is satisfied, the embedded data is extracted from the first speech code,
The speech code conversion method according to claim 1, wherein: In a speech code conversion device that converts a first speech code obtained by encoding an input speech by a first speech encoding method into a second speech code by a second speech encoding method,
A code conversion unit for converting the first audio code into the second audio code;
When arbitrary data is embedded in the first speech code received from the transmission source, an embedded data extraction unit that extracts embedded data from the first speech code,
Means for separately transmitting the second voice code obtained by the conversion and the extracted data to a destination;
A speech code conversion device comprising: When the data embedding condition is satisfied at the transmission source, when the data is embedded in the first speech code by replacing a part of the first speech code with the data, the embedded data extraction unit is
A monitoring unit that monitors whether the data embedding condition is satisfied with reference to a dequantized value of a predetermined element code constituting one speech code received from a transmission source;
An extraction unit that extracts the embedded data from the first speech code if the data embedding condition is satisfied,
4. The speech code conversion apparatus according to claim 3, further comprising:

Images