JPH10336332A - High efficiency voice coder and relay exchange system using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate noise generated at relay exchange in the relay exchange system where digital one link configuration is realized for the high efficiency voice coder. SOLUTION: Two kinds of frame patterns that are differs from each other and subject to bit steal are defined. The 1st frame pattern is inserted to an output of a decoder 7 decoding voice data from a relay line 101 by using a bit steal insertion circuit 9. Furthermore, the 2nd frame pattern is inserted to an output of a rate converter 8 that converts rate of the voice data from the relay line 101 by using a bit steal insert circuit 10. Bit steal detection circuits 3, 4 detect respectively two kinds of the frame patterns, from voice data from the exchange 202 and a path changeover in the exchange 202 (path changeover exit from its own station to relay exchange) is accurately detected by using the detection timing.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-efficiency voice coding apparatus and a transit switching system using the same, and more particularly, to a switch for performing intra-station and inter-station transit switching of digital voice signals at a first transmission rate, and a switching apparatus for the same. It is provided between a digital trunk line for transmitting a digital voice signal encoded at high efficiency at a second transmission rate lower than the first transmission rate, and performs high-efficiency encoding and decoding of the digital audio signal. The present invention relates to a high-efficiency speech coding apparatus and a transit switching system using the same.

[0002]

2. Description of the Related Art When transmitting high-efficiency coded voice information using a high-speed digital private line service or the like, when a digital exchange that exchanges with an internal line is used to carry out inter-station relay of a digital trunk line, a relay is required. The bit rate of a signal (for example, 16 Kbps high-efficiency coded voice signal) transmitted in a line is transmitted in an exchange (for example, 64 Kb).
psPCM audio signal)
At the connection point between the exchange and the trunk line, a high-efficiency speech encoding device for performing code conversion (high-efficiency speech encoding / decoding) between speech signals is provided, and digital speech signals are decoded from high-efficiency speech codes. And high-efficiency speech codes.

In such a high-efficiency speech coding system, input speech data is subjected to batch processing for each block of about 20 ms called a frame in order to realize advanced speech compression. To this end, as shown in FIG. 7, the encoder adds a frame synchronization bit to the coded data and transmits the coded data in order to transmit the frame phase on the transmitting side to the receiving side. Further, the decoder detects a frame synchronization bit from the received encoded data, and performs the synthesizing process at the same frame phase as that on the transmitting side.

When a general toll dial network is considered, a tandem connection for relaying and switching a plurality of stations is generally used. When a high-efficiency speech coding apparatus is used in such a network, there is a serious problem of deteriorating speech quality, such as deterioration of sound quality and increase in delay due to tandem connection.

As a solution, a digital one-link relay switching system has been proposed. This is a transit switching method in which the path of a transit circuit exchange that becomes a tandem connection is transmitted as encoded data, so that voice quality is not degraded due to encoding / decoding and line delay is not increased due to processing delay. Therefore, in principle, it is possible to construct a network without deterioration of voice quality even if relaying is performed many times. The present invention relates to a method for controlling encoded data at the time of relay switching.

FIG. 5 is a block diagram showing an example of the above-described high-efficiency speech coding apparatus, which is disclosed in Japanese Patent Application Laid-Open No. 7-107050. Hereinafter, the case of 16 Kbps will be described.

In the conventional transit switching system, as shown in FIG. 5, taking the high-efficiency speech encoding device 204 as an example, the encoder side encodes the PCM data inputted from the transit exchange 202. , A speed conversion circuit 2 (speed conversion from 64 Kbps to 16 Kbps) for converting the input PCM data, and an output of the speed conversion circuit 2 and an encoder 1
Selector 5 for switching the output of the relay line switch 202
And a bit steal detection circuit 3 for detecting a predetermined frame pattern inserted by bit stealing from the input PCM data from the PC.

On the other hand, the decoder side has a digital line 10
Decoder 7 for decoding the encoded data inputted from 1
And a speed converter 8 (1) for converting the speed of the encoded data.
Speed conversion from 6 Kbps to 64 Kbps), a selector 11 for switching between the output of the speed conversion circuit 8 and the output of the decoder 7, and a bit stealing for inserting a predetermined frame pattern into the output signal by bit stealing. And an insertion circuit 64. The encoding side selector 5 and the decoding side selector 11 are switched by the output of the bit steal detection circuit 3.

Other high-efficiency speech coding devices 205 to 207
Has the same configuration, and a description thereof will be omitted.
Reference numerals 201 and 203 denote trunk line exchanges.
210 indicates a telephone, and 100, 101, 110 and 111 indicate digital lines, respectively.

Next, an outline of the operation of the prior art will be described. Call dropped by own station, that is, telephone 208 in FIG.
209 (the same applies to the telephones 210 and 209), the voice signal input from the telephone 208 is converted by the trunk line switch 202 into a 64 Kbps PCM.
After being converted to data, it is input to the audio encoding device 204, and the PCM data is encoded by the encoder 1.

In this case, the bit steal detection circuit 3
Is in a non-detection state, the selector 5 selects the output of the encoder 1, and the encoded data is output to the digital line 100. This encoded data is transmitted to the high-efficiency speech encoding device 2
05, and is decoded to PCM data, output to the circuit switch 201, and reproduced by the telephone 209 as an audio signal.

In the opposite direction (from telephones 209 to 20)
The same applies to the direction 8).

On the other hand, in the case of the relay exchange,
9 and 210, if the trunk circuit switch 202 detects that the call is a trunk exchange call, 104 and 1
Connect the 05 path.

Therefore, the output of the bit steal insertion circuit 64 is input to the bit steal detection circuit 23, and the output of the bit steal insertion circuit 72 is input to the bit steal detection circuit 3, and the bit steal detection circuits 3, 2
3, a predetermined frame pattern is detected,
The selectors 11 and 31 on the decoding side and the selectors 5 and 25 on the encoding side are switched to the speed conversion circuits 8 and 28 (decoding side) and 2 and 22 (encoding side).

The trunk line switch 202 is connected to the trunk path 104
In order to output the input data transparently by 105, the encoded data input from the digital line 111 is transparently output to the digital line 100 via the speed conversion circuit 2 and the selector 5. Similarly, the encoded data of the digital line 101 is transparently output to the digital line 110 via the speed conversion circuit 22 and the selector 25.

Thus, digital one-link relay switching is realized.

[0017]

Problems of the prior art will be described with reference to FIG. In the normal state, the high-efficiency speech encoders 204 and 205, and 206 and 207 are in opposition to each other. That is, the frame synchronization between the audio CODECs has been established between the encoder 1 and the decoder 40 and between the encoder 41 and the decoder 7. Similarly, frame synchronization between audio CODECs has been established between the encoder 21 and the decoder 42 and between the encoder 43 and the decoder 27.

Here, consider the case of a call that is dropped by the own station (call between telephones 208 and 209).
Connects the path 106 to the high-efficiency speech coding apparatus 205.

Similarly, the trunk line switch 202 connects the path 102 to the high-efficiency speech encoder 204. At this time, there is no particular problem since the high-efficiency speech encoding devices 205 and 204 are in the facing state and the frame synchronization between the speech CODECs is established. The same applies to telephone calls between telephones 208 and 210.

Next, consider the case of a call for relay exchange. As described above, the normal high-efficiency speech encoders 204, 205, and 2
06 and 207 are facing each other. Here, considering a telephone call between telephones 209 and 210, circuit switch 201 connects path 106 to high-efficiency speech coding apparatus 205 by a connection signal from telephone 209 and transmits the connection signal to trunk circuit switch 202. .

When the trunk circuit switch 202 determines from the received connection signal that the call is not a drop of its own station, it transfers the connection signal to the corresponding circuit switch 203, and at the same time,
Path 104 between the high-efficiency speech encoders 204 and 206,
105 is connected.

Thus, the bit steal detection circuits 3 and 23 detect the frame pattern inserted by the bit steal insertion circuits 72 and 64, and switch the selectors 5, 11 and 25, 31 to achieve high efficiency speech coding. The devices 205 and 207 face each other.

Then, when the circuit switch 203 connects the path 107 between the high-efficiency speech coding apparatus 207 and the telephone 210, a telephone call between the telephones 209 and 210 is established.

However, the problem at this time is that, as described above, the opposing states (self-dropped states) of the high-efficiency speech encoders 204 and 205 and 206 and 207 until immediately before the relay exchange are performed. That is, the high-efficiency speech encoding devices 205 and 207 face each other from the moment when the speech is encoded. That is, it is necessary for the high-efficiency speech encoding devices 205 and 207 to newly establish frame synchronization.

At this time, the noise generated at the time of the transition from the immediately preceding synchronization establishment state to the out of synchronization state and the transition from the out of synchronization state to the synchronization establishment state in the decoder 40 is output to the telephone set 209, There was a problem that the quality of the call at the moment of switching to exchange was degraded.

In the trunk exchange 202, the path 1
If the switching between 04 and 105 is performed at the same time, the deterioration of the communication quality can be minimized.
When there is a time difference between the switching of the path 104 and the switching of the path 104, the time difference decoder 40
7 will be input, and the deterioration will be increased.

The same applies to the decoder 42. However, since the telephone 210 is not in a talking state at this time, there is no particular problem. FIG. 6 shows a time chart of the above description.

It should be noted that, to give a supplementary explanation, the transition from the synchronization-established state to the out-of-synchronization state and the transition from the out-of-synchronization state to the synchronization-established state refer to the encoder 1, the decoder 40, and the encoder 41. From the opposition of the encoder 21 to the decoder 7 and from the opposition of the encoder 21 to the decoder 42 and from the encoder 43 to the decoder 27, the encoder 43 and the decoder 40, the encoder 41 and the This is a state transition to the opposition to the gasifier 42.

An object of the present invention is to provide a high-efficiency speech coding apparatus capable of eliminating noise generated at the time of transit switching and a transit switching system using the same.

[0030]

According to the present invention, there is provided an exchange for performing intra-station and inter-station relay exchange of a digital voice signal at a first transmission rate, and a second switch having a lower rate than the first transmission rate. A high-efficiency audio encoding device that is provided between a digital trunk line that transmits a digital audio signal that has been encoded at a high transmission rate and performs high-efficiency encoding and decoding of the digital audio signal, Encoding means for transcoding and outputting a digital voice signal from the exchange; first rate conversion means for converting the digital voice signal from the exchange to the second transmission rate; Decoding means for decoding and outputting the audio signal, second speed conversion means for converting the digital audio signal from the trunk line into the first transmission rate, and output from the decoding means. First bit steal insertion means for inserting a predetermined first frame pattern added by bit stealing with respect to the first bit stealing means, and a predetermined bit stealing added to an output of the second speed conversion means. Second to insert the obtained second frame pattern
Bit stealing means, first detecting means for detecting the first frame pattern from the exchange, second detecting means for detecting the second frame pattern from the exchange, First switching means for selecting the output of the first and second bit steal insertion means in accordance with the detection result of the second detection means and guiding the selected output to the exchange; and detection result of the second detection means And a second switching unit for selecting an output of the encoding unit and the output of the first speed conversion unit and outputting the selected output to the trunk line.

Further, according to the present invention, there is provided an exchange for performing intra-station and inter-station relay exchange of digital voice signals at a first transmission rate, and a high-speed switch at a second transmission rate lower than the first transmission rate. A digital trunk line for transmitting a digital voice signal encoded with efficiency, and the above-described high-efficiency voice encoding device provided between each of the plurality of trunk lines and the exchange. A transit switching system is obtained.

In the present invention, switching of the signal path inside the exchange can be detected by the high-performance speech coding apparatus by having two types of frame patterns by bit steal output from the high-performance speech coding apparatus. And that

This makes it possible to make the data switching timing of each of the high-efficiency speech coding apparatuses that are through at the time of the relay switching almost simultaneously, to facilitate re-pulling in the frame synchronization between the voice CODECs, and to facilitate the relay switching. The problem caused by the signal path switching time difference inside the transit exchange at the time can be avoided.

[0034]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, and the same parts as those in FIG. 5 are denoted by the same reference numerals. A description will be given of the high-efficiency speech encoding device 204.

On the decoding side, a decoder 7 for decoding the digital audio signal from the trunk line 101 and a predetermined first frame pattern added by bit stealing to the decoded output are inserted. A bit stealing insertion circuit (# 1) 9 and a speed conversion circuit (# 2) 8 for converting the speed of a digital audio signal from the trunk line 101.
A bit steal insertion circuit (# 2) 10 for inserting a predetermined second frame pattern added by bit stealing to the speed conversion output, and outputs of the bit steal insertion circuits 9 and 10 And a selector 11 for selecting and outputting to the exchange 202.

On the encoding side, an encoder 1 for encoding the digital audio signal from the exchange 202, a speed conversion circuit (# 1) 2 for converting the speed of the audio signal, and a second A bit steal detection circuit (# 2) 3 for detecting a frame pattern, and a bit steal detection circuit (# 1) 4 for detecting a first frame pattern of the audio signal
And a selector 5 for selecting an output of the encoder 1 and the speed conversion circuit (# 1) 2 and leading the output to the trunk line 100.

The selector 5 has a bit steal detection circuit (#
2) The selector 11 is controlled in accordance with the detection output of 3, and the selector 11 is controlled in accordance with the output of the OR gate 6 which takes the OR of the detection outputs of the bit steal detection circuits 3 and 4.

Other high-efficiency speech coding devices 205 to 207
Has the same configuration.

In the present invention, different frame patterns predetermined by the two bit steal insertion circuits (# 1, # 2) 9, 10 are inserted.
2 shows an example of a data format having the frame pattern, respectively, in the case of μ-law PCM.

FIG. 2 shows a bit steal insertion circuit (# 1).
9 is a data format obtained by adding a PCM data frame pattern by bit stealing.
“S” explicitly indicates the insertion of bit steal data. In this example, the LSB bits of the sample are converted into PCM data once every six samples for the PCM signal output from the decoder 7. Is replaced with a frame pattern.

FIG. 3 shows a bit stealing insertion circuit (#
2) This is the data format of the encoded data obtained at 10 and subjected to speed conversion in which the encoded data frame pattern is added by bit stealing. “D” is the encoded data when the encoded data is 16 Kbps, and “F”
Indicates a frame pattern for encoded data in which encoded data is inserted into a sample (16 Kbps to 64 Kbps) subjected to speed conversion by the speed conversion circuit (# 2) 8.

As shown in FIG. 7, the encoded data "D" is assigned to the LSB side of the sample, and the other bits are set to "1". Even when these samples are input to the encoders 1 and 21 as PCM signals, there is an advantage that noise output from the decoders 40 and 42 can be reduced. Note that M
Since the SB bit is a polarity bit, only the polarity of the signal changes.

Hereinafter, the operation of the embodiment of the present invention will be described. In the normal state, the high-efficiency speech encoding device 20
4 and 205 and 206 and 207 are facing each other.
That is, the frame synchronization between the audio CODECs has been established between the encoder 1 and the decoder 20, and between the encoder 41 and the decoder 7. Similarly, the encoder 21 and the decoder 42,
The frame synchronization between the audio CODECs between the encoder 43 and the decoder 27 has been established.

Here, in the case of a call that is dropped by the own station (calls between telephones 209 and 208), the circuit switch 201 connects the path 106 to the high-efficiency speech encoder 205 by a connection signal from the telephone 209. .

By connecting this path, the circuit switch 201
Are input by the encoder 41 and output to the digital line 101. The encoded data is decoded by the decoder 7 and output to the trunk exchange 202 as a voice signal.

Further, the trunk line switch 202 is connected to the high-efficiency speech coding apparatus 204 through the path 10 by the connection signal described above.
2 is connected, and the voice signal input from the trunk line switch 202 is coded by the coder 1
Output to 00.

The coded data is decoded into a voice signal by the decoder 40 and output to the circuit switch 201, thereby establishing a telephone call between the telephones 209 and 208.
At this time, there is no particular problem since the high-efficiency speech encoding devices 204 and 205 are in the opposite state and the frame synchronization between the speech CODECs is in the established state. The same applies to telephone calls between telephones 210 and 208.

Next, consider the case of a call for relay exchange. As described above, the normal high-efficiency speech encoders 204, 205, and 2
06 and 207 are in opposition. Here, the telephone 209
Considering the telephone conversation between the telephone switch 209 and the telephone set 209, the circuit switch 201 receives the connection signal from the telephone
5 and the connection signal is transmitted to the trunk exchange 202.

When the trunk circuit switch 202 determines from the received connection signal that the call is not a drop-off call, it transfers the connection signal to the corresponding circuit switch 203 and, at the same time, passes the path 10 of the high-efficiency speech encoders 204 and 206.
4, 105 are connected. Then, the circuit switch 203 establishes a path 10 between the high-efficiency speech encoder 207 and the telephone 210.
7, the communication path between the telephones 209 and 210 is established.

However, by switching to the relay exchange,
It is necessary to perform resynchronization pull-in of the encoded data frame between the encoder 43 and the decoder 40 and between the encoder 41 and the decoder 42, and there is a problem of deterioration of voice quality due to the processing. . Further, the deterioration of voice quality is also increased by the time difference of the path connection in the relay exchange, but this time difference cannot be detected by the conventional technology.

Therefore, in the present invention, as in the prior art, means for constantly inserting a frame pattern into the output data to the trunk circuit switch 202 by bit stealing, and the above-described bit stealing from the data input from the trunk circuit switch. Means for detecting the frame pattern
It is possible to discriminate between a call dropped by the local station and a call exchanged for relay.

Then, as shown in FIGS. 2 and 3, two types of frame patterns for PCM data and coded data are defined to detect the timing of switching the path of the trunk exchange at the time of trunk exchange. It is possible to minimize the deterioration of the voice quality.

The operation at the time of relay exchange will be described below in detail. When the trunk exchange 202 detects that the call is a transit exchange, the trunk exchange 202 connects the paths 104 and 105 for the transit exchange, but consider the case where the 105 is connected first and the 104 is connected with a delay. . At this time, since the bit steal detection circuits (# 1) 4, 24 and the bit steal detection circuits (# 2) 3, 23 are in the non-detection state, the A inputs are selected for the selectors 5, 11, 25, and 31. .

First, since the path 105 is connected, the input 12 to the high-efficiency speech encoder 204 is applied to the PCM data output from the decoder 27 by the bit steal insertion circuit 1 (29). The PCM data into which the frame pattern has been inserted is input, and the data is detected by the bit steal detection circuit (# 1) 4 as a frame pattern.

Based on the detection result, the selector 11 selects the B input.
2) The encoded data in which the encoded data frame pattern is inserted by the bit steal insertion circuit (# 2) 10 into the encoded data whose speed has been converted by 8 is output. Since the bit steal detection circuit (# 2) 3 is in the non-detection state, the selector 5 keeps selecting the A input.

On the other hand, normal PCM data is input to the input 32 to the high-efficiency speech encoder 206 because the path 104 is not connected.

Next, consider the case where the path 104 is connected. Since the path 104 is connected, the input 32 to the high-efficiency speech encoder 206 has the encoded data frame pattern, which is the data of the output 15 of the high-efficiency speech encoder 204, into which the encoded data frame pattern is inserted. Is entered.
With this input data, the bit steal detection circuit (#
2) 23 detects the frame pattern, and the selectors 25 and 31 select the B input. By this. Output 33
, The encoded data from the encoder 41 is output, and the output 35 outputs the encoded data in which the encoded data frame pattern is inserted.

Since the encoded data is input to the input 12 of the high-efficiency audio encoding device 204, the bit steal detection circuit (# 2) 3 is in a detection time chart, and the selector 5 selects the B input. The output 13 outputs encoded data from the encoder 43.

FIG. 4 shows a time chart for the above explanation.

As described above, by defining the frame patterns for PCM data and coded data, it is possible to detect the completion of the transit exchange processing inside the transit exchange, and reduce the deterioration of voice quality. It becomes possible.

As a frame pattern, X.X. 50 frame patterns for encoded data. It is conceivable to define a 50R frame pattern.

In the period from the time when the path 104 is connected to the time when the bit steal detecting circuit 23 detects the frame pattern, the encoded data input to the input 32 of the high-efficiency audio encoding device 206 is encoded. The encoding process is performed by the encoder 21. However, the data format shown in FIG. 3 does not cause a problem because the data is close to a small signal of PCM data. However, FIG. 3 shows μ-Law P
This shows the case of CM, and when the encoding side changes, the format is close to the small signal code on the encoding side.

In the above description, the connection of the path 104 is delayed from the connection of the path 105. The same applies to the case where the connection is reversed or the case where the connection is performed simultaneously.

The bit steal detection circuit (# 1) 9
Detects the "S" bit shown in FIG. 2, and the bit steal detection circuit (# 2) 10 detects the "F" bit shown in FIG.

Specifically, the bit steal detection circuit (#
1) Explaining 9, a pattern that does not occur in PCM data is determined in advance, and the transmitting side (in this example, the bit stealing insertion circuit side) checks the LSB bit of the sample in units of 6 samples to determine the frame. It is detected by checking the pattern.

[0067]

As described above, according to the present invention, by defining two types of frame patterns, it becomes possible to detect the path switching timing of the transit exchange and output only valid data. Therefore, there is an effect that the deterioration of the voice quality can be reduced.

[Brief description of the drawings]

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

FIG. 2 is a diagram showing a format example of PCM data used in an embodiment of the present invention.

FIG. 3 is a diagram showing a format example of encoded data used in the embodiment of the present invention.

FIG. 4 is a timing chart showing the operation of the embodiment of the present invention.

FIG. 5 is a block diagram of a conventional high-efficiency speech encoding device.

FIG. 6 is a timing chart showing the operation of the block in FIG. 6;

FIG. 7 is a general frame format diagram of encoded data.

[Explanation of symbols]

 1,21,41,43 Encoder 2,8,22,28 Speed conversion circuit 3,4,23,24 Bit steal detection circuit 5,11,25,31 Selector 6,26 OR circuit 7,27,40, 42 Decoder 9, 10, 29, 30 Bit stealth insertion circuit 100-111 Digital trunk line 201-203 Switch 204-207 High-efficiency speech encoder 208-210 Telephone

Claims (3)

[Claims] 1. An exchange for performing intra-station and inter-station relay exchange of digital voice signals at a first transmission rate, and a digitally encoded digital signal at a second transmission rate lower than the first transmission rate. A high-efficiency voice encoding device provided between a digital trunk line for transmitting a voice signal and performing high-efficiency encoding and decoding of the digital voice signal, wherein the digital voice signal from the exchange is transcoded. Coding means for converting the digital voice signal from the exchange into a second transmission rate; decoding means for decoding and outputting the digital voice signal from the trunk line Means, second rate conversion means for converting the speed of the digital audio signal from the trunk line to the first transmission rate, and a bit steal for the output of the decoding means. First bit stealing insertion means for inserting the added predetermined first frame pattern, and predetermined second frame added by bit stealing to the output of the second speed conversion means Second bit stealing insertion means for inserting a pattern, first detection means for detecting the first frame pattern from the exchange, and second detection for detecting the second frame pattern from the exchange. Means, first switching means for selecting the output of the first and second bit steal insertion means in accordance with the detection results of the first and second detection means, and leading the output to the exchange; and And a second switching unit for selecting the output of the encoding unit and the output of the first speed conversion unit in accordance with the detection result of the detection unit and extracting the output to the trunk line. Highly efficient speech coding device. 2. The method according to claim 1, wherein the first frame pattern is a predetermined frame pattern for PCM data, and the second frame pattern is a predetermined frame pattern for encoded data. The high-efficiency speech encoding device according to claim 1, wherein 3. An exchange for performing intra-station and inter-station relay exchange of a digital voice signal at a first transmission rate, and a highly efficient coded digital signal at a second transmission rate lower than the first transmission rate. 3. A high-efficiency speech coding apparatus according to claim 1, further comprising: a digital trunk line for transmitting a speech signal; and the high-efficiency speech encoding device according to claim 1 provided between each of said plurality of trunk lines and said exchange. Intermediate switching system.