JP3163567B2 - Voice coded communication system and apparatus therefor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice coded communication system.
In particular, sound related to transmission error countermeasures in wireless lines
The present invention relates to a voice coded communication system and a device thereof .

[0002]

2. Description of the Related Art Conventionally, data transmission has been mainly performed by using a wired network in both analog and digital systems, and errors during transmission in the wired network are extremely small, and the error rate is 10 ^-4 to 1.
It is said to be ^0-8 . Against this background, speech codecs have been developed for application to digital data transmission over wired networks. As an example, an adaptive codec (ADPCM) speech codec (speech codec)
Communication system and its device) have been developed assuming an error rate of 10 ⁻⁴ or less. In recent years, applications have been made to replace a part of a wired network with wireless transmission (for example, a cordless telephone). In this case, in consideration of the compatibility with the wired network, it is appropriate to use a conventional speech codec, for example, the ADPCM system. However, the occurrence of transmission errors is unavoidable in wireless transmission. This can be reduced to some extent by a technique such as retransmission if the real-time property is ignored, but the retransmission technique cannot be applied to audio signal processing such as conversation since real-time property (delay time within 10 to 20 ms) is essential.

FIG. 4 shows 32 examples of the effects of transmission errors.
It is a waveform diagram which shows the waveform of the reproduction | regeneration audio signal in a kbps ADPCM audio codec. FIG. 4A shows the case where there is no error.
FIG. 4B is a waveform when an error having an error rate of 10 ^-2 occurs. In FIG. 4B, it is observed that an impulse-like signal having a very large amplitude not shown in FIG. 4A (the main signal in the figure is indicated by an arrow) is superimposed.

[0004]

SUMMARY OF THE INVENTION
In order to apply a speech codec to a system in which a part of a wired network is a wireless section, some error countermeasures are required. As one of the methods, use of an error correction code is conceivable. However, use of the error correction code has a disadvantage that the amount of transmission information increases. Therefore, if the voice codec itself can take measures against errors, its utility value is great. Conventionally, such a technology is not known because a voice codec used for a wired network as described herein is not used for a wireless network.

SUMMARY OF THE INVENTION It is an object of the present invention to reduce the influence of a transmission error in a radio section on reproduced sound, which is a problem of the prior art, and reduce the deterioration of reproduced sound quality on the receiving side. Voice note
An encryption communication system and an apparatus therefor are provided.

[0006]

SUMMARY OF THE INVENTION A speech coded communication system according to the present invention comprises an input speech signal and one step from a first adaptive predictor.
Tsu the difference between the predicted value of the previous flop using an adaptive quantizer in accordance with the level again and inverse quantizing the coded signal sends out a coded signal obtained by adaptive differential coding quantized to the transmission path
A reproduced signal obtained by adding the raw residual and the predicted value by a first adder
To the first adaptive predictor, and a reproduction residual obtained by inversely quantizing a coded signal received via a transmission path by an adaptive inverse quantizer, and a second
The adaptive predictor in one step prior to the predicted value and the second pressure
And outputs a reproduced audio signal by adding
In speech coding communication system that includes a decoder for providing an audio signal to said second adaptive predictor, wherein said encoder
An output of a first adder and the second addition of the decoder
A comparative attenuator is provided at both or one of the output sides of the
The amplitude of the reproduced signal output from the first adder or
The amplitude of the reproduced audio signal output from the second adder is
When the predetermined threshold value is exceeded, the reproduced audio signal and / or
Attenuating a raw audio signal to produce said first adaptive predictor and / or
The second adaptive predictor is provided.
Is what you do.

Further, the speech coded communication system of the present invention
The difference between the input speech signal and the predicted value one step before from the first adaptive predictor is adjusted according to the level according to the quantization scale factor from the first quantization scale factor adaptation unit. Conversely said coded signal with a coded signal obtained by adaptive differential coding quantized sent to the transmission path by using the encoder
Adding the quantized reproduction residual and the predicted value by a first adder
And a second quantization scale factor adaptation unit, which encodes the encoded signal received via the transmission path by an adaptive inverse encoding method, and an adaptive differential encoding method for applying the reproduced signal to the first adaptive predictor. From the reproduction residual obtained by inverse quantization according to the quantization scale factor from the second adaptive predictor.
One step before the prediction value and the second the regeneration audio signal with are added by the second adder and outputs a reproduced sound signal
And a decoder to be applied to the adaptive predictor , wherein the output of the first adder of the encoder is
Side and the output side of the second adder of the decoder.
Is provided with a comparison attenuator in either one of the first adder
Of the reproduced signal output from the second adder.
Threshold the amplitude of the reproduced audio signal force in accordance with the quantization scale factor from the first quantization scale factor adaptation unit or the second quantization scale factor adaptation unit
The playback signal and / or the playback audio signal
Attenuate the first adaptive predictor and / or the second adaptive predictor
The characteristic is that it is given to the response predictor.
You.

[0008] Also, the transmitting side encoder of the present invention provides an input sound
An input terminal for a voice signal, an adaptive predictor for generating a predicted value,
An input audio signal input from the input audio terminal and
Find the difference from the predicted value from the predictor and output the input residual signal
Subtractor, and the input according to the quantization scale factor.
Coded signal obtained by quantizing the power residual signal and adaptive difference encoding
An adaptive quantizer for transmitting to a transmission path, and
A quantization scale for generating the quantization scale factor
A factor adaptation unit, and dequantizes the encoded signal to
An adaptive inverse quantizer that outputs a reproduction residual to an adaptive predictor
And the reproduction residual and a prediction value from the adaptive predictor.
An adder for outputting a reproduced audio signal by
When the level exceeds the threshold, the reproduced audio signal is attenuated.
And a comparative attenuator for supplying the adaptive predictor
It is characterized by the following.

Further, the receiving side decoder according to the present invention has a transmission line
From the coded signal received from
A quantization scale factor adaptation unit to be generated;
Inversely quantizes the encoded signal according to the
An adaptive inverse quantizer for outputting a difference signal;
An adaptive predictor for generating a predicted value from an audio output;
Adding a value and the residual signal to output a reproduced audio signal
Adder and the level of the reproduced audio signal exceeds a threshold
When the reproduced audio signal is attenuated,
And a comparison attenuator for supplying the adaptive predictor.
It is characterized by the fact that

[0010]

FIG. 2 is a block diagram of a comparative attenuator which is an essential part of the present invention. First, the input reproduction signal Sr (k) is compared with the threshold value by the comparator 1, and when the level of the reproduction signal Sr (k) exceeds the threshold value, the reproduction signal Sr '(k) is attenuated through the attenuator 2. The output is switched by the switch 3 so as to output the reproduced signal Sr (k) when the threshold value is not exceeded.
The reproduced signal Sr ″ (k) finally output as described above
Is limited within a certain level.

This comparison attenuator is used for the speech encoder and the decoder.
Incorporation into either one or both reduces comparison
A speech codec with attenuator is constructed, Poor sound quality due to transmission errors
Speech coded communication with stable quality with less
You. That is, the encoder also generates an impulse sound signal.
To the level due to transmission errors at the decoder.
It makes it easier to detect pulse noise. Do this
And suppression of impulse sound signals by an encoder
However, such cases are rare in general call voice.
And the difference caused by adjusting the attenuation of the comparative attenuator.
Is not felt, and by providing the comparative attenuator of the present invention.
There is almost no deterioration in reproduction sound quality.

Next, as an embodiment of the present invention, ADPCM
Method described case incorporating the comparator attenuator described above in both the decoder encoder speech encoder Goka communication by.

FIG. 1 is a block diagram of an ADPCM speech codec with a comparison attenuator (encoder and decoder for speech coded communication) according to the present invention. The processing flow is shown below. First,
The encoder calculates the difference ( input residual signal a) between the input speech signal and the prediction signal f adaptively predicted from the reproduced signal one step before. The adaptive residual quantizer 4 adaptively quantizes the input residual signal a according to the level, and outputs the quantized output c to the transmission path . The quantization width used at that time, that is, the quantization scale factor b is derived from the encoded output c by the quantization scale factor adaptation unit 5. An adaptive inverse quantizer 6 reproduces and outputs a reproduction residual d based on the value (encoded output c) quantized by the adaptive quantizer 4. Request by <br/> Ri reproduced signal e to adding its reproduced residual d and one step before the prediction signal f. The reproduced signal e is input to the comparison attenuator 8 added in the present invention, and when the value is equal to or larger than the threshold value, the value of the reproduced signal e is attenuated and output as the reproduced signal e '. The reproduced signal e 'is input to the adaptive predictor 7 to generate a predicted signal f used for prediction in the next step. The difference between the predicted signal f output from the adaptive predictor 7 and the input speech in the next step is quantized by the above procedure. This is the flow of processing on the encoder side.

On the decoder side, the coded output c is transmitted from the coder, and the quantization scale factor adaptation unit 9 determines the quantization width, based on the code c 'affected by the transmission path (for example, a radio line). that calculates and outputs the quantization scale-factor g. The residual h is decoded by the adaptive inverse quantizer 10 using the quantization scale factor g and the code c ′ as inputs.
The adaptive predictor 11 outputs a prediction signal i using the decoded residual h and the reproduced speech j 'one step before. The reproduced signal j is synthesized by adding the prediction signal i and the reproduction residual h.
The reproduced sound j is input to the comparison attenuator 12 added in the present invention, and when the value is equal to or larger than the threshold value, the reproduced sound j is attenuated and output assuming that a transmission error is included. This is the reproduced audio signal j '. This reproduced sound j 'is similar to the encoder.
Is input to the applied predictor 11 to generate a prediction signal i used for prediction in the next step. Is performed.

FIG. 3 is a block diagram of a comparative attenuator which is a main part of the present invention. In the case of an adaptive coding method such as the ADPCM method, it is better to determine the threshold value from a variable having a large relationship with the reproduced sound, such as a quantization scale, instead of making the threshold unchanged. Hereinafter, as a specific example, an implementation result when the threshold value is determined based on the quantization scale factor Y (k) will be described.

FIG. 5 is a waveform chart showing the effect of the present invention.
It shows the result of implementing an ADPCM speech codec with a 32 kbps comparative attenuator. FIG. 5 (A) shows the result in the case of the related art without using the comparative attenuator, and FIG. 5 (B) shows the result in the case of the present invention in which the comparative attenuator is inserted. The threshold value at this time is 7 × 2 ^{Y (k)} using the quantization scale factor Y (k), and the coefficient of the attenuation is 0.9. As shown in the figure, the effect of the transmission error is reduced by providing the comparative attenuator even in the reproduced voice waveform (the portion where the noise is reduced is indicated by an arrow in FIG. 5A). Quantitatively, an improvement of about 1.7 dB in the segmental SNR according to the present invention and about 0.2 dB in the cepstrum distance (spectral envelope distortion) could be realized. Impulsive crackling discomfort was also reduced in hearing. Furthermore, the processing amount of the present invention was as small as 0.11 MIPS, and could be realized with about 2.5% of the processing amount of the ADPCM codec (about 4.4 MIPS). The above embodiment is the case of the voice codec of the ADPCM system, but can be applied to the voice codec of another system.

[0017]

As has been described [of the effect the invention described above in detail, the speech marks Goka
Incorporating a comparison attenuator in the decoder and speech decoder reduces impulsive noise generated when a transmission error occurs, and improves the quality of reproduced speech by about 21% in segmental SNR and about 14% in cepstrum distance. . This allows the system to have a wireless section where transmission errors are likely to occur.
Higher quality communication becomes possible. Further, since the processing amount is extremely small, even if a comparative attenuator is added, the increase in power consumption is small, which is extremely effective for a wireless device requiring low power consumption.

[Brief description of the drawings]

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

FIG. 2 is a block diagram showing a main part of the present invention.

FIG. 3 is a block diagram showing a main part of the present invention.

FIG. 4 is a reproduction waveform diagram of a 32 kbps ADPCM voice codec;

FIG. 5 is a reproduction waveform diagram showing the effect of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 comparator 2 attenuator 3 switch 4 adaptive quantizer 5, 9 quantization scale factor adaptation unit 6, 10 adaptive inverse quantizer 7, 11 adaptive predictor 8, 12 comparative attenuator

Continuation of front page (56) References JP-A-2-53335 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04B 14/00-14/06 H03M 7/38

Claims (4)

(57) [Claims] 1. An input speech signal and a signal from a first adaptive predictor.
The difference from the prediction value of the previous step is quantized using an adaptive quantizer in accordance with the level, and an adaptive difference-encoded coded signal is transmitted to a transmission line, and the coded signal is dequantized.
And the predicted value is added by the first adder.
An encoder of an adaptive difference encoding system for supplying a raw signal to the first adaptive predictor, and a reproduction residual obtained by inversely quantizing an encoded signal received via a transmission path by an adaptive inverse quantizer < br /> and the predicted value of one step before the second adaptive predictor first
And output the reproduced audio signal by adding
A speech encoding communication system comprising: a decoder for providing the reproduced speech signal to the second adaptive predictor; an output side of the first adder of the encoder;
And / or one of the outputs of the second adder of the adder
Is provided with a comparison attenuator, and the output of the first adder is
Amplitude of the raw signal or reproduced sound output from the second adder
When the amplitude of the voice signal exceeds a predetermined threshold value,
Signal and / or the reproduced audio signal to attenuate the first
Adaptive predictor and / or said second adaptive predictor.
A speech coded communication system characterized by: 2. An input speech signal and a signal from a first adaptive predictor.
The difference between the predicted value one step before and the level according to the quantization scale factor from the first quantization scale factor adaptation unit is quantized using an adaptive quantizer, and the encoded signal subjected to the adaptive difference encoding is quantized. wherein it sends out to the transmission path
A reproduction residual obtained by inversely quantizing a coded signal and the predicted value are represented by a first
The reproduced signal added by the adder of
And a coded signal received via the transmission path, which is inversely quantized according to the quantization scale factor from the second quantization scale factor adaptation unit. A second adder calculates the reproduction residual obtained and the predicted value one step before from the second adaptive predictor .
Regeneration speech with in adds outputs reproduced audio signal
A speech encoding communication system comprising: a decoder for providing a signal to the second adaptive predictor; an output side of the first adder of the encoder;
And / or one of the outputs of the second adder of the adder
Is provided with a comparison attenuator, and the output of the first adder is
Amplitude of the raw signal or reproduced sound output from the second adder
When the amplitude of the voice signal exceeds a threshold value according to the quantization scale factor from the first quantization scale factor adaptation unit or the second quantization scale factor adaptation unit ;
Attenuating the reproduced signal and / or the reproduced audio signal.
A first adaptive predictor and / or a second adaptive predictor.
A voice coded communication system characterized by the above. 3. An input terminal of an input audio signal and a predicted value are generated.
Adaptive predictor to be formed, and input from the input voice terminal.
The input speech signal and the predicted value from the adaptive predictor Find the difference between
Subtractor that outputs the input residual signal
Quantizes the input residual signal according to the
Adaptive quantizer for sending encoded signal to transmission line
And the quantization scale factor from the encoded signal
A quantizing scale factor adapting unit to be generated;
The reproduction residual given to the adaptive predictor by dequantizing the signal is
Output adaptive inverse quantizer, the reproduction residual and the adaptive prediction
That outputs the reproduced audio signal by adding the predicted value from the
And the level of the reproduced audio signal exceeds a threshold value.
Attenuates the reproduced audio signal and supplies it to the adaptive predictor
A speech coder comprising a comparison attenuator. 4. Quantization from an encoded signal received from a transmission path.
Quantized scale factor to generate quantized scale factor
An adaptation unit and the code according to the quantization scale factor.
Inverse quantizer that inversely quantizes the quantized signal and outputs the residual signal
And generating a predicted value from the residual signal and the reproduced audio output.
An adaptive predictor, adding the predicted value and the residual signal to
An adder for outputting a raw audio signal, and a level for the reproduced audio signal;
When the audio signal exceeds the threshold, the playback audio signal is attenuated.
Output to the adaptive audio predictor as well as the reproduced audio output
And a comparison attenuator.