JPH0683169B2 - Method for establishing resynchronization during voice compression transmission - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention utilizes a high-speed digital relay line to efficiently transmit voice information between multiplex converters, resulting in voice generated due to line error or the like. The present invention relates to a method for establishing resynchronization during voice compression transmission that minimizes the turbulence time.

[Conventional technology]

Up to now, as described in Japanese Patent Laid-Open No. 60-87538, when out-of-frame synchronization is detected on the receiving station side, the fact is notified to the transmitting station side and synchronization is restored from the transmitting station side. The operation is designed to take place.

[Problems to be solved by the invention]

However, so far, no consideration has been given to the delay in resynchronization establishment due to the delay time on the transmission path, especially when satellite communication is involved in the middle of an international line, etc.
Since about 250 ms is required in one direction of the transmission path, in the case of voice communication, there is a problem that the disturbance time of voice, that is, the time until resynchronization is established becomes long.

An object of the present invention is to provide, in the case where line-corresponding voice information is out of synchronization due to a line error, etc., when the line-corresponding voice information is transmitted and received as a multi-frame via a high speed digital relay line. , A method for establishing resynchronization during voice compression transmission, in which frame synchronization is reestablished early by a unique resynchronization establishment process for each line on the receiving side.

[Means for solving problems]

The above-mentioned purpose is to extract voice parameters from line-corresponding voice information by waveform analysis of the voice information, encode and compress them, and then multiplex them with other line-corresponding frames as frames to which voice / silence identification bits are added. A method for establishing resynchronization when transmitted as multiframes in a converted state to each other through a high-speed digital relay line, in which voice information corresponding to the line is continuously transmitted in each of the multiplex converters. When transmitted as a silent frame over two or more frames, a silence frame in which the encoded compressed voice parameter is replaced with the synchronization information as a specific bit pattern for each non-leading silence frame following the leading silence frame. While being transmitted as a multi-frame, The corresponding frame is separated and extracted from the multi-frame as a line-corresponding frame, and then a sync bit is detected corresponding to the line, and if it is a voice frame or a head silence frame, the intra-frame encoded compressed voice If the parameter is a non-leading silence frame, the frame synchronization is performed when each of the encoded compressed speech parameters, which is temporarily stored in advance and set inside the leading silence frame for the non-leading silence frame, is decoded. When the disconnection is detected, the resynchronization is established based on the synchronization information set in the non-head silent frame.

[Action]

When a voice is coded and compressed by linear prediction or the like and a voice frame is formed with a frame length of about 20 ms, the voice frame is composed of voiced / unvoiced identification information, voice parameter information, and the like. Now, in the silent section, the silent information continues for several frames. However, when the receiving side holds the first (first) frame information and repeats the decoding process for several frames, With regard to the silent information itself, the second and subsequent frames can be processed without being sent. Therefore, if the synchronization information is always set from the transmission side to the second and subsequent frames of the silent information, the reception side can reestablish the frame synchronization early by detecting the synchronization information when the frame synchronization is lost. However, it is something.

By the way, as described above, the synchronization information may be transmitted to the receiving side when the input voice is not present, that is, when the silence is generated, but may not be transmitted when the silence is not present. Regarding this, it is generally considered that the silence rate during the call is about 50%, and since there are considerable silence frames during the actual conversation, it is considered that there is no particular problem, If the receiving side does not detect the synchronization information for a certain period of time or more, the transmitting side may be notified of the fact as in the case so far. In addition,
The silent state mentioned above includes not only the silent state during a call but also the case where the call is not in progress. This is because when the call is not in progress, it is equivalent to the silent state during a call. Therefore, when the communication is not in progress, silent frames are continuously transmitted during that period, and synchronization information is inserted into each of these silent frames.

〔Example〕

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

First, a multiplexer (multiplex converter) according to the present invention will be described. FIG. 1 shows a block configuration as an example thereof. The multiplexer 100 is connected to the exchange 110 via a plurality of lines, and the voice from each line is analyzed by the voice encoding unit 102 in the line corresponding unit (dot chain line display) via the trunk interface unit 101,
The audio is compressed and then framed as audio information with a frame length of 20 ms. Second frame configuration
As shown in the figure, the frame has a frame sync bit (1
Bit), followed by a voiced / non-voiced identification bit, and further after that, a voice parameter is inserted. While this frame is input to the synchronization information insertion unit 103, the control unit 107 determines whether the frame is a voiced frame or a silence frame, and the determination result and 1
The relationship with that before the frame is determined, and the voiced frame and the silent first frame are multiplexed with the frames from other line corresponding units by the demultiplexing unit 104 in the same format as they are. For a silent frame, synchronization information (synchronization pattern information as a specific bit pattern consisting of a plurality of bits) is replaced and inserted at the position of a voice parameter, and then multiplexed and demultiplexed by another demultiplexing unit 104. It has become so. Synchronization information insertion unit 10
An example of the frame format processed in 3 is shown in FIG. The voiced frame and the silent first frame are kept in the same format, but with respect to the silent frames after the silent second frame, the synchronization information is inserted at the position of the voice parameter. From the demultiplexing unit 104, the transmission data of a plurality of lines are multiplexed into a multi-frame as much as possible, and then transmitted to the opposing multiplexer (having the same configuration as the multiplexer 100) via the relay line.

On the other hand, the data in the multi-frame format received from the relay line is separated by the demultiplexing unit 104 for each line, and the synchronization detecting unit 105, the voice decoding unit 106 and the trunk interface are formed in the form as shown in FIG. By being sent to the exchange 110 via the unit 101, a voice call for each line is possible.

Now, with reference to FIG. 4, a description will be given of the processing in the control unit which controls the synchronization detection unit provided for the line, and the hardware of the synchronization detection unit 105 is set to the frame synchronization detection instruction at the time of initial setting. The number of bits of a frame including bits is counted and information for one frame is reported to the control unit 107. The control unit 107 performs a determination process 402 from the beginning of the received data whether it is a frame synchronization bit pattern, and if a frame synchronization bit is detected, an initialization process (frame synchronization detection instruction process 403 for hardware, reception side After the reset processing 404 of the timer for giving up the original re-synchronization establishment and the reset processing 405 of the out-of-sync counter for ignoring the momentary bit error, the judgment processing 406 of the presence or absence of sound is performed. ing. If there is sound here, initial setting processing (silent output reset processing 407 that is output when synchronization is lost)
After that, a voice decoding process 408 is performed from the voice parameter. Further, in the case of no sound, it is determined whether the synchronization information is already being detected, which indicates that the synchronization is out of sync. It has become. In the case of the silent first frame, the decoding process 412 is performed after the voice parameter saving process 411. In the silent second frame and thereafter, the read processing 413 post-decoding processing 412 of the voice parameter of the saved silent first frame is performed.

On the other hand, if the synchronization is lost in the frame synchronization determination processing 402, after the determination processing 414 whether the synchronization information is being detected, if it is not being detected, it is determined whether the synchronization has been lost n times in succession.
415 is to be performed. When it occurs n times consecutively, the instruction processing 417 is performed as the resynchronization establishment processing to detect the synchronization information when there is no sound, but in addition to this, when the synchronization information cannot be detected. Therefore, a monitoring timer setting process 418 is performed, and a silence output process 419 outputs silence to the user until reception of voiced information after resynchronization is established. The synchronization detection unit 105 performs the synchronization information detection process, and when the synchronization information is detected, the number of frame bits subsequent to the synchronization information is counted and reported to the control unit 107, but when the synchronization information is not detected, it is reported as a timeout. ,Determination process
As in the case of the conventional technique, the 401 performs resynchronization establishment processing by notifying the transmitting side of the loss of synchronization.

〔The invention's effect〕

As described above, according to the present invention, when out-of-frame synchronization occurs, resynchronization can be established only by processing on the receiving side, and the out-of-synchronization time can be shortened to shorten the audio disturbance to the user. Therefore, there is an effect that the service is improved.

[Brief description of drawings]

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 configuration of a general voice frame, and FIG. 3 is a diagram showing synchronization information according to the present invention set. FIG. 4 is a diagram showing a voice frame including a voice frame for three frames, and FIG. 4 is a diagram showing a flow of a resynchronization establishment process at the time of detecting out-of-sync on the receiving side. 100 ... Multiplexing device, 101 ... Trunk interface unit, 102 ... Voice coding unit, 103 ... Synchronization information insertion unit, 10
4 ... demultiplexing unit, 105 ... synchronization detecting unit, 106 ... voice decoding unit, 107 ... control unit, 110 ... switchboard.

Claims (1)

[Claims] 1. A line-corresponding frame as a frame to which voice parameters are extracted, coded and compressed from a line-corresponding voice information by waveform analysis of the voice information, and a voice / silence identification bit is added. This is a method for establishing resynchronization when they are transmitted as multi-frames in a multiplexed state to each other through a high-speed digital relay line to each other. When transmitted as a silent frame over two or more frames, for each non-head silence frame following the head silence frame, the silence condition in which the encoded compressed voice parameter is replaced with the synchronization information as the specific bit pattern. While being transmitted as a frame, the number of times it is received as a multi-frame from the high-speed digital repeater line. The corresponding frame is separated and extracted from the multi-frame as a line-corresponding frame, and then a sync bit is detected corresponding to the line, and if it is a voice frame or a head silence frame, the intra-frame encoded compressed voice If the parameter is a non-leading silence frame, the frame synchronization is performed when each of the encoded compressed speech parameters, which is temporarily stored in advance and set inside the leading silence frame for the non-leading silence frame, is decoded. A method for establishing resynchronization during voice compression transmission, which enables resynchronization to be established based on the synchronization information set in the non-leading silence frame when disconnection is detected.