JP2001111499A - Voice decoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a voice decoding device by which a user confirms the deterioration of a transmission line and a call service is improved by dissolving a problem that the user has incompatibility caused by a sudden output voice interruption when the conventional transmission line is deteriorated and gradually attenuating an output voice before a decoding output is stopped by reduction in the quality of the transmission line. SOLUTION: In this voice decoding device, a channel decoder 4 detects an error from synchronizing data and outputs error information, an attenuator 7 judges whether a probability to stop the decoding output in a voice decoder 5 exists in the situation of the transmission line or not in accordance with error information and the decoding output is gradually attenuated when it is judged that the probability to stop the output exists.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech decoding apparatus in a speech communication system used for wireless communication, and more particularly to a speech service for eliminating a sense of incongruity in which reproduced sound is suddenly stopped when a transmission path condition deteriorates. The present invention relates to a speech decoding device capable of improving the sound quality.

[0002]

2. Description of the Related Art In a speech encoding system used in a speech communication system or the like, an error correction system is generally employed in order to suppress deterioration of speech quality due to transmission errors. The error correction method is a method in which an error correction code is added to coded data in a voice on a transmission side and an error is corrected on a reception side, and the FEC (Forward Error Correction: FEC) having a high error correction capability such as Viterbi decoding. Forward error correction)
The method is widely used.

[0003] However, when the condition of the transmission path deteriorates and an error occurs that exceeds the error correction capability, the residual error after decoding usually occurs in a burst, so that unpleasant noise and noise appear in the decoded output of voice. Output. For this reason, the conventional speech decoding device can recognize the occurrence of an error even in a situation where the error cannot be corrected. A method of stopping audio decoding output by a decoder has been used.

As a result, as shown in FIG. 5, when the error rate in the transmission path gradually increases, the decoding output is stopped (cut off) at the moment when the decoding output is stopped (cut off) due to the setting conditions of the audio decoder. The reproduced sound output from the speaker is stopped (disconnected). FIG. 5 is an explanatory diagram showing a relationship between a transmission path error rate, a speech decoder output, and a speaker output in a conventional speech decoding device.

[0005] As a prior art of the audio decoding apparatus, Japanese Patent Application Laid-Open No. Hei 6-149279, published May 27, 1994, entitled "Audio Decoding Method and Decoder" (applicant: Kokusai Electric Corporation,
Inventor: Hiroki Goto et al.). According to this conventional technique, when the out-of-synchronization of a received audio signal is detected, the decoded audio is attenuated and output at the initial out-of-synchronization stage, and thereafter, the out-of-sync detection signal and the white noise are changed to the decoded audio. This is an audio decoding device that outputs, and can eliminate the occurrence of abnormal sound during the initial out-of-synchronization time when the out-of-synchronization is detected.

[0006]

However, in the above-described conventional speech decoding apparatus, when an error exceeding the capability of the employed error correction system occurs, the speech decoding output is suddenly stopped. From the user's point of view, there is a problem that the deterioration of the transmission path condition cannot be recognized, the voice output is suddenly interrupted, and the user feels strange.

Further, since the audio output is suddenly interrupted, it is misunderstood that the other party has suddenly disconnected the line, which is unpleasant. If the transmission path condition is recovered, the audio decoding output is resumed and the conversation is resumed. Although there is a possibility that the communication can be continued, there is a problem that the call service is deteriorated, for example, the line is disconnected by determining that the other party has disconnected.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and solves the conventional problem that a user has a sense of incongruity due to a sudden stop of output sound when the quality of a transmission path is deteriorated. An object of the present invention is to provide a speech decoding device capable of improving a call service by allowing a user to recognize a decrease in transmission path quality by gradually attenuating output sound before stopping output.

[0009]

According to the present invention, in order to solve the above-mentioned problems of the prior art, in a speech decoding apparatus, a receiving means receives a radio speech signal, demodulates the speech signal into speech coded data, and obtains a transmission path condition. The detecting means separates the encoded audio data into synchronous data and audio information data for each encoded frame, detects an error from the synchronous data and outputs error information, and the audio decoding means performs error correction decoding of the audio information data. When the error exceeds the correction capability, the output of the decoded voice is stopped, and the attenuation control unit determines that the decoding output may be stopped in the voice decoding unit in accordance with the error information. It is determined whether or not there is a road condition, and if it is determined that there is no possibility of being stopped, the decoded output is output without attenuating, and if it is determined that there is a possibility of being stopped, the decoded output is output. Gradually decay Since the audio reproducing means reproduces the audio from the decoded output from the attenuation control means, the attenuation control means attenuates the decoded output before the transmission output quality is deteriorated and the decoded output is stopped by the audio decoding means. So,
It is possible to eliminate a sense of incongruity in which the playback sound is suddenly stopped.

[0010]

Embodiments of the present invention will be described with reference to the drawings. Note that the function realizing means described below may be any circuit or device as long as the function can be realized, and some or all of the functions may be realized by software. is there. Further, the function realizing means may be realized by a plurality of circuits, or the plurality of function realizing means may be realized by a single circuit.

[0011] Explaining the general concept, the speech decoding apparatus according to the present invention determines the state of the transmission channel quality from the received encoded speech data, and outputs the decoded output when the transmission channel quality tends to deteriorate. Since the output is gradually attenuated before stopping, it is possible to improve the call service by resolving a sense of incongruity in which the reproduced sound is suddenly stopped, and allowing the user to recognize the deterioration of the transmission path quality. It is.

In terms of function realizing means, the speech decoding apparatus according to the present invention comprises: a receiving means for receiving a radio speech signal and demodulating the speech speech data into speech encoded data; Transmission path condition detecting means for detecting an error from the synchronization data and outputting error information indicating the transmission path condition; and outputting error-corrected decoded speech information data to output a decoded speech. Under a specific condition that exceeds the correction capability, the voice decoding means to stop the output of the decoded voice, and corresponding to the specific condition to stop the output of the decoded voice in the voice decoding means, according to the error information,
In the voice decoding means, it is determined whether or not there is a transmission path condition in which the decoding output may be stopped, and if it is determined that there is no possibility of being stopped, the decoding output is output without attenuating, and the Attenuating control means for gradually attenuating the decoded output when it is determined that there is a possibility that the received audio encoded data is received, and audio reproducing means for reproducing the audio from the decoded output from the attenuating control means; Judging the state of the transmission path quality from, if the transmission path quality is in a deteriorating trend, because the decoding output is gradually attenuated before stopping the decoding output, eliminating the uncomfortable feeling that the playback sound is suddenly stopped, By making the user aware of the deterioration of the transmission path quality, the call service can be improved.

First, the configuration of a speech decoding apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a configuration block diagram of a speech decoding device according to the present invention.

An audio decoding apparatus (this apparatus) according to the present invention includes an antenna 1, a radio 2, a demodulator 3, a channel decoder 4, an audio decoder 5, and a D / A converter (D / A converter in FIG. 1). /
A) It is composed of 6, an attenuator 7, and a speaker 8. Here, the antenna 1, the radio 2, and the demodulator 3 correspond to the receiving means in the claims, and the channel decoder 4
Corresponds to the transmission path condition detecting means in the claims, and the audio decoder 5
And the D / A converter 6 correspond to the voice decoding means in claims.
The attenuator 7 corresponds to the attenuation control means in the claims, and the speaker 8 corresponds to the sound reproduction means in the claims.

Next, each section of the apparatus will be described in detail. The antenna 1 receives a modulated wave in a wireless band and is a general antenna. The wireless device 2 is a general wireless device that converts a modulated wave received by the antenna 1 into a modulated wave in a baseband. The demodulator 3 is a general demodulator for demodulating a baseband modulated wave into a digital signal.

The channel decoder 4 detects error information indicating a transmission path condition from the received digital signal. In general, digital signals are often transmitted in a frame configuration, and the frame configuration includes synchronization data for performing frame synchronization, bit synchronization, channel identification, and the like, and information data such as voice. Therefore, as a specific operation of detecting error information indicating a transmission path condition from the digital signal received by the channel decoder 4, first, synchronization data and audio information data are converted from the digital signal demodulated by the demodulator 3 in frame units. Is detected and separated, and the voice information data is output as it is.

On the other hand, the number of error bits is detected from the separated synchronization data, and is output as error information indicating the number of error bits and a frame pulse indicating frame timing is output. Note that the detection of the number of error bits in the synchronization data
Since the synchronous data pattern is already known, it is calculated by a bit pattern title or the like.

The audio decoder 5 is an audio decoder that corrects errors in the audio information data separated by the channel decoder 4, performs audio decoding, and outputs decoded audio data. However, as a feature of the audio decoder 5 of the present apparatus, when the state of the transmission path is poor, errors are remarkably large, and when the error correction capability is exceeded, the quality of the audio output is prevented from deteriorating due to the residual decoding error. This is an audio decoder having a function of stopping decoding output. It is assumed that the condition for stopping the decoding output (output stop condition) is determined in advance from the number of errors per frame and the frequency of occurrence thereof.

The D / A converter 6 is a general digital / analog (D / A) converter for converting digital data of decoded audio data into an analog signal.

The attenuator 7 responds to the output stop condition in the voice decoder 5 and determines whether or not the decoding output of the voice decoder 5 is in a transmission path state where the output may be stopped in accordance with the input error information. Then, when it is determined that there is no possibility of being stopped, the input decoded output is output without attenuation, and when it is determined that there is a possibility of being stopped, the input decoded output is gradually reduced. This is an attenuator that performs an attenuation process to attenuate the output. Details of the attenuator 7 will be described later. The speaker 8 is a general speaker that reproduces and outputs audio according to an input audio signal.

Here, the attenuator 7 of the present apparatus will be described in more detail. The attenuator 7 of the present invention implements a method of determining a condition for starting attenuation (attenuation start condition) and an attenuation method in response to an output stop condition in the audio decoder 5. Specifically, in the attenuator 7 of the present invention, the start of attenuation and the determination of the attenuation are determined based on the relationship between the predetermined number of errors and the time during which the error occurs, based on the error information and the frame pulse output from the channel decoder 4. This determines the processing method.

Here, the relationship between the predetermined number of errors and the time during which an error occurs is a relationship determined by an output stop condition under which the error correction function of the audio decoder 5 stops decoding output. For example, it is considered that the error of the entire frame can be estimated from the error of the synchronization data in each frame, and the synchronization when the output of the audio decoder 5 is stopped due to an error is determined in advance by an experimental value or a specified value. Check the number of data errors (output stop condition).

Here, as an example, when the output stop condition in the audio decoder 5 is such that the synchronization data of each frame is 20 bits and an error of 2 bits or more continues for 6 frames or more, the audio decoder 5 has an error correction capability. It is assumed that the output is stopped when it is determined that the output has been exceeded.

First, an example of the internal configuration of the attenuator 7 of the present apparatus will be described with reference to FIG. FIG. 2 is a block diagram showing an example of the internal configuration of the attenuator 7 of the speech decoding device according to the present invention. Here, the condition for stopping the decoding output in the audio decoder 5 is “when an error of 2 bits or more continues for 6 frames or more”. In the audio decoder 5, for error correction and audio decoding processing, It is assumed that about 6 frames are decoded while being buffered.

The attenuator 7 of the speech decoding apparatus according to the present invention comprises a memory 11, a decision unit 12, a coefficient switch 13, and a multiplier 1.
And 4. Here, the memory 11 corresponds to the error information storage means, the determiner 12 corresponds to the determination means, the coefficient switch 13 corresponds to the coefficient switch means, and the multiplier 14 corresponds to the claim information. It corresponds to a term multiplication means.

Each part will be described specifically. The memory 11
It is a memory such as a shift register that stores error information from the channel decoder 4 while shifting the information in order of a plurality of frames. FIG. 2 shows a case where error information for 5 frames is stored as an example, where t is error information which is the number of errors of synchronous data in the current frame, and t-1 is error information of the immediately preceding frame. The current error information is input to t every time a frame pulse indicating the timing of frame processing is input from the channel decoder 4, where t is t−1 and t−1 is t The shift is performed as -2, and error information from the current data to the past t-4 is always stored.

The determiner 12 detects the state of the transmission line from the error information stored in the memory 11 and determines whether or not a condition for starting attenuation (attenuation start condition) is met (attenuation start determination). ) To output a determination result, and to output an attenuation start instruction as a determination result when it is determined that the attenuation process should be started.

Here, an example of a criterion for determining the attenuation start condition will be described. For example, if the condition that the speech decoder 5 determines that the error correction capability has been exceeded and the output is stopped is that an error of 2 bits or more continues for 6 frames or more, the timing at which the attenuation process should be started is determined. Condition (decay start condition)
Is determined to be the start of attenuation when all the contents of the five pieces of error information stored in t to t-4 of the memory 11 indicate errors of 2 bits or more.

The coefficient switch 13 is a coefficient switch for switching and outputting a coefficient to be multiplied to attenuate the decoded output in accordance with the determination result of the determiner 12, and usually outputs a coefficient (value 1) that is not attenuated. Triggered when the result of determination from the determiner 12 is an attenuation start instruction, the value of the coefficient is gradually reduced, and returns to 1 when the coefficient becomes 0 (zero).

Here, a specific example of the coefficient switch 13 will be described with reference to FIG. FIG. 3 is a block diagram showing a specific configuration example of the coefficient switch 13 in the attenuator 7 of the speech decoding device according to the present invention. The coefficient switch 13 in the attenuator 7 according to the present invention includes a coefficient memory 21, a counter 22, and a comparator 2.
3 and a switch 24.

Each component will be specifically described. Coefficient memory 21
Is a memory that stores a counter value output from a counter 22 described later in advance and a coefficient corresponding thereto in a one-to-one correspondence. FIG. 3 shows an example in which the attenuation is performed in six stages. If the counter value is 0, the corresponding coefficient 1.0 is selected and does not affect the amplitude of the multiplied audio output, that is, is not attenuated, but is incremented when the counter starts counting up. The coefficients 0.8, 0.6,..., 0 are selected according to the counter value, and the audio output is attenuated in amplitude over five frames, and finally the output is set to zero.

The counter 22 has an initial state and the judgment unit 12.
Although the counter value is fixed to 0 (zero) until the determination result from becomes an attenuation start instruction, once the determination result from the determiner 12 becomes an attenuation start instruction, the timing of the frame pulse from the channel decoder 4 To set the counter value to 1
Count up by 1 and return to 0 when counting to 5. Then, the counter value is fixed to 0 until the next determination result from the determiner 12 becomes an attenuation start instruction.

The comparator 23 is a comparator for switching the coefficient of the attenuation process according to the counter value of the counter 22. The switch 24 is a switch for selecting a coefficient output from the coefficient switch 13 from the coefficient memory 21, and is controlled by the comparator 23.

The operation of the coefficient switching unit 13 is such that the counter value of the counter 22 is fixed to 0 (zero) until the initial state and the judgment result from the judgment unit 12 indicate an attenuation start instruction, and the comparator 23 switches the switch 24. Is controlled to select a coefficient of 1.0. Then, when the determination result from the determiner 12 indicates an attenuation start instruction, the counter 22 counts up the counter value by one at the timing of the frame pulse from the channel decoder 4 and returns to 0 when counting to five. At this time, the comparator 23 controls the switch 24 in accordance with the counter value that is counted up for each frame from the counter 22, and switches the output coefficients to 0.8, 0.6,. When the count returns to 0, the coefficient is returned to 1.0.

The multiplier 14 is a general multiplier that multiplies the coefficient output from the coefficient switch 13 by the decoded audio signal from the D / A converter 6.

Next, the operation of the attenuator 7 of the present apparatus will be described with reference to FIG. In the attenuator 7 of the present apparatus, the error information from the channel decoder 4 is stored while being sequentially shifted in the memory 11 at the timing of the frame pulse from the channel decoder 4.

Then, the content of the memory 11 is determined by the determiner 12 at the timing of the frame pulse, and the determination result is output to the coefficient switch 13. For example, memory 1
If the error information of the past five frames stored in No. 1 has two bits or more, and if the error of the next frame also has two bits or more, the condition for stopping the decoding output of the audio decoder 5 is “ When the error of 2 bits or more continues for 6 frames or more ", the decoding output is stopped, and before that, it is determined that the attenuation process should be started, and an attenuation start instruction is output as the determination result.

When the decision unit 12 outputs the decision of the attenuation start instruction, the coefficient switching unit 13 sets the channel decoder 4
Counter 22 for each frame pulse timing from
Is counted up from 0, and the comparator 23 is accordingly counted up.
And a switch 24 is provided so that the coefficient is gradually reduced from 1.0, and a coefficient that is reduced is output. The multiplier 14 multiplies the decoded sound output from the D / A converter 6 by the coefficient at the timing of the frame pulse. This is output as a reproduced audio signal.

According to this operation of the attenuator 7, while the audio information data is buffered for decoding in the audio decoder 5 according to the error information indicating the status of the transmission path detected by the channel decoder 4, The output of the decoded sound is gradually attenuated by the attenuator 7 before the output of the decoded audio is suddenly stopped by the audio decoder 5, and the output is stopped by the attenuation. Since the output of the audio decoder is stopped, the situation where the output is suddenly stopped is eliminated.

Next, the operation of the present apparatus will be described with reference to FIG. In the speech decoding apparatus of the present invention, a modulated wave in a radio band is received by an antenna 1, converted into a modulated wave in a baseband band by a radio 2, and demodulated into a digital signal by a demodulator 3.

The digital signal is separated by a channel decoder 4 into synchronization data and audio information data in frame units, the number of error bits detected from the synchronization data is output as error information, and a frame pulse indicating frame timing is output. Is output.

On the other hand, the audio information data separated by the channel decoder 4 is error-corrected by an audio decoder 5 and subjected to audio decoding processing, and the decoded audio data is converted to a D / A converter 6.
Is output to However, since the speech decoder 5 is also provided with an error correction function, transmission errors due to bad transmission path conditions are also corrected. However, when an error exceeding the error correction capability is detected, the output of decoded speech data is stopped. It has become so.

The decoded audio data output from the audio decoder 5 is converted from a digital signal to an analog signal by the D / A converter 6 and input to the attenuator 7 as a decoded audio signal. If there is a possibility that the decoding output may be stopped in the audio decoder 5 according to the error information indicating the status of the error in the transmission path detected by the channel decoder 4, the audio decoder 5 attenuates the reproduced audio signal gradually. The reproduced sound is output from the speaker 8.

As a result, as shown in FIG. 4, when the error rate in the transmission path gradually increases, the output of the attenuator 7 is stopped before the decoding output is stopped due to the setting conditions of the audio decoder 5. The attenuation process is started, the reproduced sound output from the speaker 8 is gradually attenuated, and the demodulation output is stopped by the audio decoder 5 soon, as in the conventional audio decoding device shown in FIG. A situation in which the output of the speaker 8 is suddenly stopped can be avoided. FIG.
FIG. 5 is an explanatory diagram showing a relationship between a transmission path error rate and outputs of a speech decoder 5 and a speaker 8 in the speech decoding device of the present invention.

According to the speech decoding apparatus according to the embodiment of the present invention, the attenuator 7 determines the transmission path condition in accordance with the error information indicating the transmission path condition detected by the channel decoder 4, and the decoded signal is output by the speech decoder 5. If it is determined that the situation is about to stop, the decoded output is gradually attenuated, thereby eliminating the phenomenon that the reproduced sound output from the speaker 8 is not output suddenly, and the line is suddenly disconnected for the user. This has the effect of eliminating such discomfort.

According to the audio decoding apparatus of the present invention, if the condition for stopping the output in the audio decoder 5 is, for example, that an error of 2 bits or more continues for 6 frames or more, the channel decoder 4 converts First, the error information is detected quickly, and the attenuator 7 sequentially stores the error information for five frames in the memory 11, and when all the stored error information is 2 bits or more, the decision unit 12 outputs Since it is determined that the demodulation output may be stopped in 5 and the attenuation process is started, the coefficient is switched by the coefficient switch 13 so as to gradually increase the attenuation rate for each frame, and the multiplier 14 multiplies the decoded voice by the coefficient. The effect that the attenuator 7 having a simple configuration corresponding to the condition for stopping the output in the audio decoder 5 is subjected to the attenuating process, and the output can be gradually attenuated before the stop of the reproduction output. A.

[0047]

According to the present invention, the transmission path condition detecting means detects an error from the synchronous data and outputs error information, and the attenuation control means stops the decoding output in the voice decoding means according to the error information. It is determined whether or not there is a possibility of a transmission path condition, and when it is determined that there is a possibility of being stopped, the decoding output is gradually attenuated, so that the audio decoding device is used.
This eliminates the sense of incongruity in which the reproduced sound is suddenly stopped, and makes the user feel that the quality of the transmission path has deteriorated, thereby providing an effect of improving the call service.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of a speech decoding device according to the present invention.

FIG. 2 is a block diagram showing an internal configuration of an attenuator of the speech decoding device according to the present invention.

FIG. 3 is a block diagram illustrating a specific configuration example of a coefficient switch in the attenuator of the speech decoding device according to the present invention.

FIG. 4 is an explanatory diagram showing a relationship between a transmission path error rate, a speech decoder output, and a speaker output in the speech decoding device of the present invention.

FIG. 5 is an explanatory diagram showing a relationship between a transmission path error rate, a speech decoder output, and a speaker output in a conventional speech decoding device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Antenna, 2 ... Radio, 3 ... Demodulator, 4 ... Channel decoder, 5 ... Audio decoder, 6 ... D / A converter, 7 ... Attenuator, 8 ... Speaker, 11 ... Memory, 12 ... Judgment device , 13: coefficient switcher, 14: multiplier, 21: coefficient memory, 22: counter, 23
... Comparator, 24 ... Switch

Claims (3)

[Claims] 1. An apparatus for receiving error-encoded coded voice data, performing error correction according to the error-corrected code,
An audio decoding apparatus for stopping the output of the decoded audio when an error exceeding the correction capability of the error correction code occurs in the received encoded audio data in a process of decoding the audio and outputting the decoded audio. Determining the status of the transmission channel quality from the received encoded voice data, and when the transmission channel quality tends to deteriorate, gradually attenuating the output before stopping the output of the decoded voice. Audio decoding device. 2. A receiving means for receiving a wireless audio signal and demodulating it into encoded audio data, separating the encoded audio data into synchronous data and audio information data for each encoded frame, and detecting an error from the synchronous data. Transmission path condition detection means for detecting and outputting error information indicating the transmission path condition, and outputting a decoded speech by performing error correction decoding of the audio information data, under specific conditions where the error exceeds the correction capability, There is a possibility that the audio decoding means for stopping the output of the decoded audio, the decoding output corresponding to the specific condition for stopping the output of the decoded audio in the audio decoding means, and the decoding output being stopped in the audio decoding means in accordance with the error information. Determine whether there is a certain transmission path conditions, if it is determined that there is no possibility of being stopped, output the decoded output without attenuating,
An attenuation control unit that gradually attenuates the decoded output when it is determined that there is a possibility of being stopped, and an audio reproduction unit that reproduces audio from the decoded output from the attenuation control unit. Voice decoding device. 3. An attenuation information storage means for storing error information while shifting the error information in order of a plurality of frames, and detecting a state of a transmission line from the error information stored in the error information storage means. When there is a possibility that the decoding output is stopped in the audio decoding means from the result, a judging means for judging the start of attenuation, and a coefficient switching means for switching and outputting a coefficient to be multiplied to attenuate the decoding output according to the judgment result. 3. A speech decoding apparatus according to claim 2, wherein said speech decoding apparatus comprises: a multiplication means for multiplying the decoded speech by a coefficient output from said coefficient switching means.