JPS63314037A - System for establishing resynchronization in compressing and transmitting voice - Google Patents

Abstract

PURPOSE:To suppress the disturbing time of voice at a minimum level, by transmitting synchronization information from a transmission side to a voice nonaudio section separately from a frame synchronizing bit, and performing resynchronization independently based on the synchronization information at a reception side. CONSTITUTION:A voice frame is constituted in such a way that the frame synchronizing bit is positioned at the front, and an audio/non-audio identification bit is positioned following after the synchronizing bit, and a voice parameter is inserted following after the identification bit. At a synchronization information insertion part 103, the synchronization information is inserted in a nonaudio frame behind a second nonaudio frame, and after that, it is multiplexed on another frame from a line correspondence part at a multiplex separation part 104. At the reception side, frame synchronization is reestablished by detecting the synchronization information at the time of frame step out. In such a way, it is possible to perform the establishment of the resynchronization only by a processing at the reception side when the frame step out is generated, and to shorten a step out time and the disturbance of the voice.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is directed to minimizing the time for audio disturbances caused by line errors and the like when transmitting audio information efficiently using a high-speed digital line. This invention relates to a method for establishing resynchronization during compressed voice transmission.

[Conventional technology]

Until now, as described in Japanese Patent Application Laid-open No. 60-87538, when frame synchronization is detected on the receiving station side, the transmitting station is notified of this and synchronization is restored from the transmitting station. The action is about to take place.

[Problem that the invention seeks to solve]

However, so far, no consideration has been given to the delay in establishing resynchronization due to delay time on the transmission path, and when satellite communications such as international lines are interposed in the middle, it takes about 250 m5 in one direction of the transmission path. Therefore, in the case of voice communication, there is a problem that the voice disturbance time, that is, the time until resynchronization is established becomes long.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for establishing resynchronization during audio compression transmission, which makes it possible to minimize audio disturbance time.

[Means for solving problems]

The above object is achieved by transmitting synchronization information from the transmitting side in addition to frame synchronization bits during voice silent periods, and by independently attempting resynchronization on the receiving side based on the synchronization information.

[For production]

When audio is compressed by linear prediction or the like and converted into an audio frame with a frame length of about 20n+s, the audio frame consists of voice/silence identification information, audio parameter information, and the like. Now, in a silent section, silence information will continue over several frames, but if the receiving side retains the first frame information and processes this repeatedly for the number of frames, the silence information will be sent especially after the second frame. Therefore, if the transmitting side always sets synchronization information in frames after the second frame of silence information, the receiving side can detect the synchronization information when the frame is out of synchronization. This is where synchronization is re-established.

By the way, as described above, the synchronization information is transmitted to the receiving side when there is no input audio, that is, when there is no sound, but it may not be sent when there is no sound. Regarding this, it is generally reported that the silence rate during a call is about 50%, and since a considerable number of silent frames occur even during actual conversations, it is thought that no particular problem will occur. If synchronization information is not detected on the receiving side for a certain period of time or more, the sending side may be notified of this as in the previous case.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 4.

First, the multiplexing apparatus according to the present invention will be explained. FIG. 1 shows the block configuration of an example thereof.

The multiplexing device 100 is connected to an exchange 110 via a plurality of lines, and the voice from each line is passed through the trunk interface unit 101, and the voice waveform is analyzed by the voice encoding unit 102 in the line corresponding unit (shown by a dashed line). The audio is compressed and then framed as audio information with a frame length of 20 nis. The structure of a frame is shown in FIG. 2, where a frame synchronization bit is inserted at the beginning of the frame, followed by a voice/no-speech identification focus, and then an audio parameter. This frame is input to the synchronization information insertion section 103, while the control section 1
In 07, it is determined whether it is a voice frame or a silent frame, and the relationship between this determination result and the one frame before is determined, and the voice frame and the first silent frame remain in the same format. The demultiplexer 104 multiplexes the frame with frames from other line corresponding parts, but synchronization information (continuous synchronization pattern information) is inserted at the position of the audio parameter for silent frames after the second silent frame. Furthermore, the demultiplexing section 104 multiplexes the signals with those from other line corresponding sections. FIG. 3 shows an example of the frame format processed by the synchronization information insertion unit 103. The sound frame and the first silent frame are left in the same format, but synchronization information is inserted at the position of the audio parameter for the second silent frame and subsequent silent frames.

From the demultiplexing unit 104, the transmission data of multiple lines is multiplexed into multiple frames as much as possible, and then sent to the multiplexing bag facing each other via a relay line! (same configuration as multiplexer 100).

On the other hand, data in the multi-frame format received from the trunk line is demultiplexed by the demultiplexer 104 for each line, and is sent to the synchronization detector 105, the audio decoder 106, the trunk interface, and in the format shown in FIG. By sending the data to the exchange 110 via the section 101, voice calls can be made on each line.

Now, to explain the processing in the control unit that controls the synchronization detection unit provided for the line with reference to FIG. The number of bits of frames containing bits is counted, and information for one frame is reported to the control unit 107. Control unit 107
Then, a process 402 is performed to determine whether the received data is a frame synchronization bit pattern from the beginning, and if a frame synchronization bit is detected, an initial setting process (frame synchronization detection instruction process 403 to the hardware, and a process 403 unique to the receiving side) is performed. Timer reset processing 404 to give up on establishing synchronization
, after performing an out-of-synchronization counter reset process 405) to ignore momentary bit errors, a sound/non-speech determination process 406 is performed. Here, if there is sound, after initial setting processing (resetting processing 407 for the silence output when synchronization is lost), decoding processing 408 of the audio is performed from the audio parameters. In addition, in the case of silence, a process 409 for determining whether synchronization information is being detected indicating that the synchronization is already out of synchronization is performed, and if no synchronization information is being detected, a process 410 for determining whether the first frame is silent is performed. It has become. In the case of a silent first frame, a decoding process 412 is performed after an audio parameter saving process 411. In addition, if it is after the silent second frame, read processing 4 of the audio parameters of the saved silent first frame is performed.
13, decoding processing 412 is performed.

On the other hand, if synchronization has been lost in the frame synchronization determination process 402, a process 414 for determining whether synchronization information is being detected is performed, and if no synchronization information is being detected, a process 415 for determining whether synchronization has occurred consecutively n times is performed. It looks like this. If the occurrence occurs n times in a row, an instruction process 417 is performed for the synchronization detection unit 105 to detect synchronization information during silent periods as a resynchronization establishment process. Therefore, a monitoring timer setting process 418 is performed, and silence is output to the user by a silence output process 419 until the voice information is received after resynchronization is established. The synchronization detection unit 105 performs a process of detecting synchronization information, and when synchronization information is detected, the number of frame bits following the synchronization information is counted and reported to the control unit 107, but if no synchronization information is detected, it is reported as a timeout. , the determination process 401 causes the resynchronization establishment process to be performed by notifying the sending side of the out-of-synchronization, as in the case of the prior art.

〔Effect of the invention〕

As described above, according to the present invention, when frame synchronization occurs, resynchronization can be established only by processing on the receiving side, and by shortening the synchronization time, audio disturbance to the user can be shortened. This has the effect of improving service.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a block configuration of an example of a multiplexing device according to the present invention, FIG. 2 is a diagram showing a general audio frame configuration, and FIG. 3 is a diagram showing a configuration in which synchronization information according to the present invention is set. FIG. 4 is a diagram showing three frames of audio frames including audio frames, and FIG. 4 is a diagram showing the flow of resynchronization establishment processing when out-of-synchronization is detected on the receiving side. 100... Multiplexing device, 101... Trunk interface unit, 102... Audio encoding unit, 103...
- Synchronization information insertion section, 104... demultiplexing section, 105.
...Synchronization detection section, 106...Audio decoding section, 107.
...Control unit, 110...Switching machine.

Claims (1)

[Claims] 1. A method for establishing resynchronization when audio information compatible with a line is compressed and encoded, multiplexed with audio information compatible with other lines, and then transmitted to each other's multiplex converter via a high-speed relay line. If the audio information in a compression-encoded frame format compatible with the line is silent information when transmitting the audio information, synchronization information is transmitted instead of the audio information, while frame synchronization is not performed when receiving the frame-based audio information compatible with the line. A method for establishing resynchronization during audio compression transmission, characterized in that when a deviation is detected, resynchronization is established based on synchronization information set instead of audio information.