JPS6165635A - Coder and decoder - Google Patents

Abstract

PURPOSE:To minimize deterioration in quality of voice and to transmit voice and data simultaneously by allowing a coding section and a decoding section to switch the code rule synchronously when a forecast coefficient detection circuit of transmission side and a reception side detects respectively that the forecast coefficient is at the outside of a prescribed range. CONSTITUTION:A tone signal generator 1 uses a data transmission request signal RE to output a tone signal of a predetermined specific frequency and a switch 3 switches a voice signal into a tone signal for a prescribed time to input it to a low-frequency coding section 2. When the tone signal is cut off after a prescribed time, a forecast coefficient detected by forecast coefficient detection circuits 5, 9 exceeds a prescribed range. Thus, the other coding rule reducing bit number is switched to the low-frequency coding section 2 and the low-frequency decoding section 10, which are operated in the data simultaneous transmission mode. Further, a multiplex circuit 8 at the transmission side inserts a data signal D to a low-order 1 bit of e.g., a low-frequency coding section 2, outputs the result multiplexedly and a separating circuit 14 at the reception side outputs separately the corresponding bit as the data signal D.

Description

[Detailed Description of the Invention] Technical field to which the invention pertains The present invention provides a code that enables simultaneous transmission of voice and data by allocating a part of encoded bits of a voice encoder to transmission of data signals. - Regarding decoders.

Conventional technology Conventionally, this type of coder/decoder uses a statistical multiplexing method that transmits a data signal instead of encoded voice when there is no voice, and one that uses a statistical multiplexing method that transmits a data signal instead of encoded voice when there is no voice. There is an embedding encoding method in which the bits outside the data signal are reduced and encoded, and the data signal is inserted into the reduced bits.

In the former case, the length of the silent section is 3i depending on the nature of the audio signal.
4, and since silent sections do not appear regularly, it is a national problem to transmit data signals at a constant bit rate, and when switching back from data transmission to audio transmission, a cutoff occurs at the beginning of the 5th episode. Therefore, there is a drawback that the audio transmission quality deteriorates.

In the latter case, since the number of bits for encoding the voice is small, it is natural that the voice quality deteriorates accordingly, but it also has the disadvantage that the above-mentioned quality deterioration occurs even when no data is transmitted.

OBJECT OF THE INVENTION The object of the invention is to solve the above-mentioned conventional drawbacks and to minimize the deterioration of voice quality by using different encoding rules when data is being transmitted and when it is not. The object of the present invention is to provide an encoder/decoder that can simultaneously transmit voice and data.

Structure of the Invention The encoder/decoder according to the invention includes an encoding section that includes an adaptive predictor and outputs a coded audio signal obtained by encoding an analog audio signal, and a coded audio signal sent from the encoding section. In a communication device comprising a decoding unit that decodes an analog audio signal, the encoding unit and the decoding unit are configured to perform encoding and decoding according to a plurality of encoding rules,
On the transmitting side, a tone signal generator generates a tone signal of a specific frequency for a certain period of time according to a data transmission request signal and inputs it to the encoding section, and prediction coefficients of an adaptive predictor built in the encoding section are determined in advance. and a prediction coefficient detection circuit for detecting whether the prediction coefficient of the decoding section is within a predetermined range. A detection circuit is provided, and when the prediction coefficient detection circuits on the transmitting side and the receiving side respectively detect that the prediction coefficient is outside a certain range, the encoding section and the decoding section synchronize and change the encoding rule. By switching the bits, it is possible to simultaneously transmit data signals using specific bits, and by transmitting tone signals of other frequencies, the transmitting and receiving sides can transmit coded audio alone in synchronization. It is characterized by restoring to mode.

Embodiments of the Invention Next, one embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention applied to a communication device using a two-band encoding method.

In the band division transmission method, the transmitting side includes a low pass filter 6 that passes the low frequency side of the analog audio signal V of 1, for example, 8 KH2, and a high pass filter 7 that passes the high frequency side.
A high-frequency side encoding section 4 encodes the output of the high-pass filter 7 into, for example, 2 bits and inputs it to the multiplexing circuit 8, and a low-frequency side analog that has a built-in adaptive predictor and is manually inputted via the switch 3. It includes a low frequency encoding section 2 that encodes and outputs an audio signal according to a certain encoding rule, and a multiplexing circuit 8 that outputs the output bits of the high frequency encoding section 4 and the low frequency encoding section 2 in an array. ing. The low-band encoder 2 reduces the prediction residual by changing the coefficients of the adaptive predictor using a certain algorithm according to the human input signal, and outputs the prediction residual by quantizing it to, for example, 6 bits. . The total of the output bits of the high-band encoder 4 and the output bits of the low-band encoder 2 is 8 bits,
The multiplexing circuit 8 arranges the signals in time slots of a normal PCMI channel and outputs them. However, in this embodiment, when simultaneously transmitting audio and data signals, the low frequency side encoding section 2 is configured to encode and output the low frequency side of the audio signal into, for example, 5 bits. . That is, it is possible to operate with multiple encoding rules. Multiplexing circuit 8
In this embodiment, zs
M (11a q version! 4t low jdillE[) version 11
2 To-c.

It constitutes an encoding section.

Then, a tone signal generator l generates a tone signal of a specific frequency indicating the start or end of a data signal in response to a data transmission request signal RE, and a tone signal output from the tone signal generator sl is inputted from a low-pass filter 6. For a certain period of time (for example, 500 S
degree) nti switcher 3 for input to the low band side encoding unit 2
and a prediction coefficient detection circuit 5 for detecting whether or not the prediction coefficients of the lower frequency side encoding section 2 are within a predetermined range.

The receiving side includes a single separation circuit 14, a high-pass decoding section 11, a high-pass filter 13, a low-pass decoding section IO and a low-pass filter +2, a high-pass filter 13, and a low-pass filter. Addition output of 12 outputs (!!+5
and a prediction coefficient detection circuit 9 that monitors the prediction coefficients of the fJll device included in the low frequency side decoding unit 1A10. The low frequency side decoding section 10 can operate by switching between a plurality of coding rules according to the output of the prediction coefficient detection circuit 9, and the one-separation circuit 14 can operate by switching to a plurality of coding rules.

In data simultaneous transmission mode 1, specific bits of the received signal are separated and output as data signals. In this embodiment, a high frequency side decoding section 11 and a low frequency side decoding section 10 constitute a decoding section. .

Next, the operation of this embodiment will be explained. Tone (No. 6 Generator 1 outputs a tone signal of a predetermined specific frequency according to the data transmission request signal RE inputted prior to data transmission, and the switch 3 outputs a tone signal of a predetermined frequency (for example, 500 ag)) is switched to the above-mentioned tone signal and inputted to the lower band encoder 2. Therefore, the lower band encoder 2 encodes and outputs the tone signal.At this time, the lower band encoder 2 The prediction coefficients of the built-in predictor fall within a certain range depending on the frequency of the tone signal.

For example, the prediction coefficient of the all-pole predictor is the first-order coefficient AI = 2co when the sampling frequency is f, and the tone frequency is f.
g(2tcf/fc) + quadratic coefficient A, = -1
However, in reality, 1 for the H2 tone signal is 2.
The second order coefficient A is between -0.1 and +0.1, and the second order coefficient A2 is between -1 and 10.B. Also, 2.4K
For Hz tones, the first order will be between -0,8 and -0,3 and the second order will be between -1 and -0,8. Therefore,
For example, when a 2 KHz tone signal is sent, the primary coefficient detected by the predictive coefficient detection circuit 5 is within the range of -0, 1 and +0.1, and the range of the secondary coefficient is between -1 and +0.1. -0
, 6.

Further, the tone signal is encoded by the low frequency side encoding section 2, sent out to the digital transmission path via the multiplexing circuit 8, separated by the receiving side separation circuit 14, and then decoded on the low frequency side. The low frequency side decoding unit IO creates a pre-8-count from the received signal and uses the built-in pre-aS.
Restore to 0Hz tone signal. Therefore, during this time, a tone signal can be heard on the receiving side, but it is only for a short time, so it is not a big problem.On the other hand, the prediction coefficient detection circuit 9 on the reception side monitors the prediction coefficients of the low frequency side decoding section 10. , its prediction coefficient is usually the same as that on the transmitting side, and the first-order coefficient A1 is −0,1.
~+0.1.The second-order coefficient A2 is in the range of -1 to -0.6.

Then, when the tone signal is cut off after a certain period of time, the prediction coefficients detected by the prediction coefficient detection circuit 5 and the prediction coefficient detection circuit 9 deviate from the above range. As a result, the lower-band encoder 2 and the lower-band decoder 10 each switch to another encoding rule with a reduced number of bits and operate in the data simultaneous transmission mode. Further, the multiplexing circuit 8 on the transmitting side inserts the data signal into, for example, the lower one bit (or multiple bits) of the low-frequency side encoding unit 2 and multiplexes it, and the demultiplexing circuit 14 on the receiving side The bits are separated and output as data signals.

When the transmission of the data signal is completed, a tone signal generator l on the transmitting side sends a specific frequency (for example, 2.
A tone signal of 4 KHz) is outputted to the prediction coefficient detection circuit 5 and the prediction coefficient detection circuit 9.

Both have linear coefficients between -0,8 and -0,3, and -1 and -
Detect a quadratic coefficient between 0 and 6. Then, when the tone signal is cut off and the audio signal is input, the prediction coefficient deviates from the above range. At this time, the low frequency side encoding section? The low frequency side decoding section 10 operates in synchronization by restoring to the original encoding rule that is suitable for transmission of audio signals alone.

Also, the frequency of the tone signal is 31! By preparing multiple types of modes, for example, a simultaneous data transmission mode in which the low frequency audio signal is encoded into 4 bits and a 2 bit data signal is inserted per sample, and a low frequency audio signal is encoded into 5 bits. Switching and switching back of motes can be performed by arbitrarily selecting one of three types: a data simultaneous transmission mode that encodes and inserts a 1-bit data signal, and an audio signal-only transmission mode that encodes a low-frequency audio signal into 6 bits. It is also possible. In this case, the three types of tones need to be selected at frequencies such that the prediction coefficients are exclusive. In this case, it is possible to increase or decrease the transmission value of the data signal as necessary.

In either case, the operation modes of the lower frequency encoder 2 and the lower frequency decoder IO are synchronously switched or switched back depending on the presence or absence of transmission data, so that audio bits can be transmitted during data transmission. By reducing the number of bits and simultaneously transmitting data signals, and when not transmitting data, all bits are used to encode the voice signal, which has the effect of transmitting without degrading the voice quality.

J: The above description assumes that the tone signal is input to the low-frequency encoding section, and the data signal is inserted by reducing the output bits of the low-frequency encoding section, but the tone signal is input to the high-frequency side. Alternatively, the data signal may be inserted after reducing the output bits on the high frequency side.In either case, the audio signal is optimally encoded with the given number of bits. Of course, it is encoded according to the encoding rule as follows.

The present invention is not limited to the above-mentioned two-band encoding method, but can be similarly applied to a three-band encoding method, and can also be applied to a single-band transmission encoding method. Can be applied.

Effects of the Invention As described above, the present invention includes an encoding unit and a decoding unit that can operate in multiple modes using multiple encoding rules with different numbers of output bits, and a tone signal of a specific frequency according to a data transmission request. and a switch for switching and inputting the output of the tone signal generator to the encoding unit for a certain period of time, and a switch for inputting the output of the tone signal generator to the encoder for a certain period of time, and a switch for detecting that the prediction coefficient of the value ≠ sign is within a predetermined range. The prediction coefficient detection circuit includes a prediction coefficient detection circuit for detecting a prediction coefficient, and a prediction coefficient detection circuit for monitoring the prediction coefficient of the decoding section, and the transmission and disconnection of tone signals are detected by the prediction coefficient detection circuits on the transmitting side and the receiving side. Since the encoding rule is configured to be switched synchronously between the transmitting side and the receiving side, it is possible to arbitrarily switch between the audio-only transmission mode and the data simultaneous transmission mode. Therefore, in the audio-only transmission mode, the audio quality is not degraded, and in the data simultaneous transmission mode, the audio signal can be transmitted with the minimum necessary quality degradation. Moreover, since mode switching and switching back are performed in a short time and in synchronization with transmission and reception, there is little quality deterioration during that time. Furthermore, it is also possible to increase or decrease the number of bits allocated to the data signal depending on the data transmission.

[Brief explanation of drawings]

The figure is a block diagram showing one embodiment of the present invention. (In 4, 1° tone signal generation base, 2: low frequency side encoding section, 3: cutter, 4: high frequency side encoding section, 5: prediction coefficient detection circuit, 6, 12: low pass filter, 7.13: high pass filter, 8: multiplexing circuit, 9: prediction coefficient detection circuit, lO: low band side decoding unit, ll: high band side decoding unit, 1
4: Separation circuit, 15 two adders.

Claims (1)

[Claims] An encoding unit that includes an adaptive predictor and outputs a coded audio signal obtained by encoding an analog audio signal, and a decoding unit that decodes the encoded audio signal sent from the encoding unit into an analog audio signal. In the communication device, the encoding section and the decoding section are configured to perform encoding and decoding according to a plurality of encoding rules, and on the transmitting side, a data transmission request signal is used to select a specific frequency. A tone signal generator that generates a tone signal for a certain period of time and inputs it to the encoding unit, and a prediction that detects whether a prediction coefficient of an adaptive predictor built in the encoding unit is within a predetermined range. a coefficient detection circuit, and the receiving side includes a prediction coefficient detection circuit for detecting whether or not the prediction coefficient of the decoding unit is within a predetermined range; When each of the detection circuits detects that the prediction coefficients are out of a certain range, the encoding section and the decoding section synchronously switch the encoding rule, thereby using specific bits to simultaneously convert the data signals. 1. An encoder/decoder characterized in that the transmitting side and the receiving side are similarly synchronized and restored to the coded voice only transmission mode by transmitting a tone signal of another frequency.