JPH11154935A - Method for absorbing re-transmission delay jitter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce a delay in a video frame due to re-transmission even in an environment having a large network delay by reducing the occurrence of re-transmission while absorbing retransmission delay jitter of the video frames. SOLUTION: In the case that data are sent in two-way by using frames of a fixed length having a frame number and an error check code and an error is detected in a received frame, a reject command of a frame whose error is detected is returned by using a transmission frame whose data area is reduced as prescribed. In the case that a transmission error rate or re-transmission frequency exceeds a threshold level even when it is not required to return the reject command, the data area is reduced as prescribed at all times and a frame number of transmission frames and a frame member having requested a reject command precedingly are sent by using the reduced data area.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a retransmission delay jitter absorbing method, and more particularly to a multimedia communication system for multiplexing and transmitting highly efficient compressed audio and video data and computer data.
The present invention relates to a method for absorbing jitter generated by retransmission control.

[0002]

2. Description of the Related Art In a videophone or videoconferencing system, a video signal having an enormous amount of information is generally encoded with high efficiency and then transmitted to a transmission destination. An example of such high-efficiency coding is H.264, which is an international standard of ITU-T. 261 and H.E. 263. All of these compression coding principles are motion compensation inter-frame prediction coding methods and variable length coding, and are very sensitive to transmission errors. That is, even a one-bit error propagates to a large area of the screen, and also propagates in the time axis direction.
The image quality is significantly deteriorated. For this reason, in an analog telephone line in which a transmission error is likely to occur, it is general to perform image correction after performing error correction by retransmission control.

[0003] On the other hand, voice data is also used in G.264 recommended by the ITU-T. As in 723, the data is transmitted after being compressed to an extremely low bit rate by high-efficiency coding. However, transmission errors are less noticeable than video data, but are sensitive to transmission delay. Therefore, error correction by retransmission is not used, and when an error is detected, noise is prevented from being caused by muting. G. FIG. The transmission rate specified by 723 is 5.3 Kbps or 6.3 Kbps, and generates a voice frame of about 20 bytes every 30 ms. When this is transmitted by a PHS having a transmission speed of 32 Kbps, one frame is generated in a cycle of 120 bytes.

[0004] As a method for multiplexing transmission of such video data and audio data, H.264 is also recommended by ITU-T. 223 have been issued. H. H.223 was initially considered for wired communication, but now H.223 22
H.3 is further extended to wireless communication,
223 / AnnexA (November 29, 1996) has been proposed. This H. H.223 / Annex A considers the characteristics of a wireless transmission path in which errors are large and errors occur in a burst manner. Several changes have been made since 223.

[0005] FIG. 3 shows a first multiplexed frame format of a 223 / AnnexA multiplexing layer. As shown in FIG. 20, in the first multiplexed frame format, after a synchronization flag, a fixed-length header and an information field in which video, audio, and data are multiplexed follow. The size of the synchronization flag can be selected, but a 4-byte synchronization flag is used in the initial state.

FIG. 2 shows a second multiplexed frame format of the multiplexing layer of H.223 / AnnexA. The multiplexed frame format in the initial state is the same as that in FIG. By the procedure of H.245, it is possible to change to the multiplexed frame format of FIG. In the second multiplexed frame format, the header is arranged before and after to improve the error resistance against burst errors, but the overhead of the header is doubled accordingly.

FIG. 3 shows a first format of a header in H.223 / AnnexA. FIG.
, A header includes a multiplex information section indicating what kind of data is multiplexed in the information field and how, an error detection code section as a CRC attached to the multiplex information section, and an error correction code section. including. The frame length of the multiplexed frame and the multiplexing structure of the multiplexed frame are not determined, and are notified from the transmitting side as configuration information during communication. Regarding the multiplexing structure, a maximum of 16 types of multiplexing structures are defined and a table corresponding to the header information in one-to-one correspondence is transferred to the receiving side in advance, and the receiving side receives the The header is analyzed in order to know what data is stored in which area of the information field.

H. In H.223 / AnnexA, a command and a sequence for changing the frame length of the multiplexed frame and the format of the header are specified.
As shown in FIG. 23, when the transmitting side sends a command to change the frame length and / or the format of the header, the receiving side that has received the command returns an acknowledgment.
After receiving the acknowledgment, the transmitting side sends out a multiplexed frame in which the frame length and / or the format of the header have been changed. Before making the change, it indicates from which multiplexed frame the change is made. A multiplexed frame including a header into which a counter is inserted is transmitted.

FIGS. 24A and 24B respectively show H.264 and H.264. 3 shows second and third formats of a header in H.223 / AnnexA. As shown in FIG. 24A, the second header includes a multiplex information section (MC).
And an error detection code unit, which is a CRC attached to the multiplex information unit, and an error correction code unit. On the other hand, as shown in FIG. 24B, the third header includes a multiplex information section (MC).
, A counter (HC), an error detection code unit which is a CRC attached to the multiplexed information unit and the counter, and an error correction code unit. Hereinafter, this counter will be described.

After receiving the acknowledgment, the transmitting side inserts a counter (HC) into the header as shown in FIG. 24B after the multiplex frame to be transmitted next, and as shown in FIG. A value indicating whether to change from the frame is set in the counter (HC). This value is decremented for each subsequent multiplex frame. When the value of the counter (HC) becomes 0, the transmitting side changes from the next multiplexed frame. After the changed multiplexed frame, FIG.
As shown in (a), the counter no longer exists.

In order to keep the header size constant, the length of the error correction code section differs between when there is a counter and when there is no counter, as shown in FIG. Since the receiving side does not know from which multiplexed frame the counter is inserted, the multiplexed frame received after returning the acknowledgment shall first correct the error assuming that there is no counter, If an error is detected by the detection, the error correction and error detection are performed assuming that there is a counter. If no error is detected, the counter is extracted. In the subsequent multiplexed frames, the counter is extracted assuming that there is a counter, and extraction is performed until the count value of the counter becomes zero.

On the other hand, PIAFS (PHS) is a conventional transmission control method for transmitting computer data using a PHS wireless line having a transmission speed of 32 kbps, in which transmission errors are more likely to occur than analog telephone lines.
Internet Access Forum Sta
ndard) has been proposed in Japan. PI
ARQ frames in AFS have a fixed length. FIG.
FIG. 5 shows the structure of the PIAFS ARQ frame. FIG.
In the frame type identification area, the frame type is
The data length of the significant data in the user data area is displayed in the user data length display area, the user data is displayed in the user data area, the transmission frame number is displayed in the frame number display area, and the oldest newest data is displayed in the request frame number display area. The received frame number stores an error detection code for detecting an error of the entire ARQ frame in the error detection code area. Frame synchronization is established prior to communication using a synchronization frame including a synchronization flag, and the data frame must have a synchronization flag by setting the frame length of the data frame to be the same as that of the synchronization frame. And not.

That is, in a PHS network which is a digital network with clock synchronization, bit synchronization is guaranteed.
In PIAFS, the synchronization flag “0” is set prior to data communication.
1010000111011110010100110
After the frame synchronization is established using the synchronization frame including "010011", data communication is performed while maintaining the frame synchronization by using a fixed-length data frame having the same frame length as the synchronization frame. No synchronization flag is required for this data frame. For example, assuming that a 4-byte synchronization flag is always added to an 80-byte frame, the transmission efficiency is reduced by 5%.

After establishing synchronization with the PIAFS synchronization frame, the present applicant uses H.264 by using a fixed-length multiplexed frame having the same frame length as the synchronization frame. 2
A method of removing the synchronization flag from the multiplexed frame format of the multiplexing layer of 23 / Annex A and thereby increasing the transmission efficiency has been already proposed (Japanese Patent Application No. Hei 8-1871).
No. 53).

In the above proposal, as shown in FIG.
80 by including the area for storing the audio frame of the frame
A mode is considered in which data is transmitted in a pattern in which a first fixed-length multiplexed frame of te and a second fixed-length multiplexed frame of 80 bytes that do not multiplex audio frames appear at a ratio of 2: 1. In this way, one frame of audio data can be transmitted in 120 bytes in an 80-byte multiplexed frame. Further, in this case, when no voice frame is generated in the silent section, the second
By using the fixed-length multiplexed frame, the video frame can be extended to a long frame, and the transmission band can be increased.
Even when a sound is generated during retransmission when an error occurs in the long frame, retransmission can be performed in the fixed length slot of the second fixed length multiplexed frame.

[0016]

SUMMARY OF THE INVENTION ITU-T H.264 22
The retransmission control method for video data specified in 3 / Annex A has the following two problems.

A first problem is that when a transmission error is detected in video data on the receiving side, an area for storing a reject signal to be transmitted to the transmitting side is secured only when necessary within a fixed-length video frame. It is caused by being done. H. 2
FIG. 27 shows the header format of the video retransmission control frame of 23 / AnnexA. In FIG. 27, For
The ward Control Field is forward control information, and includes a frame number SNf and an error detection code C.
RCf and Backward ControlField
d indicating the presence or absence of d. Backward
The Control Field includes an SM, which is a control command for rejection, etc., a frame number SNb requesting retransmission by rejection, and an error detection code CRCb. Note that BCH is an error correction code.

Here, Backward Contro
l Field occurs only when it is necessary to request retransmission by rejection, and Forward Control
The presence of 1Field PT indicates its presence. The advantage of this is that the Backward Control Fi
When it is not necessary to send an eld, the information area of the video is expanded correspondingly, and the transmission efficiency is increased. However, an error occurred in the transmitted video data, retransmission was requested, and an error occurred in the received video data,
Returning a reject is a completely independent event. That is, when the state of the transmission path is poor, the Backward Control Field is used during the first transmission.
A large video frame was transmitted without d, but a transmission error occurred in that frame and retransmission was requested by the other party, and an attempt was made to retransmit, and an error was detected in the video frame received at that time. The event of sending a reject to the frame occurs at the same time.
It can happen frequently. However, retransmission of a video frame and transmission of a reject cannot be performed at the same time, so that either one of the transmissions must be postponed, causing delay jitter in retransmission. FIG. 28 is a timing chart showing this state. In FIG. 28, the terminal 2 transmits the video frame 1 without the Backward Control Field, but since a transmission error has occurred, the terminal 1
d Reject the video 1 reject in the Control Field. The terminal 2 receives the reject of the video 1 and attempts to retransmit the video 1, but a retransmission wait occurs due to a difference in the multiplexing structure. During this retransmission wait, an error occurs in the video frame 7 transmitted by the terminal 1, and the terminal 2
Detects this. At this time, in the last multiplexed frame of the transmission of the terminal 2, retransmission of the video frame 1 and transmission of the reject for the video frame 7 are mixed, and the reject cannot be returned.

A second problem is that if the rejection of the same frame is repeated, the modulo makes one revolution.
FIG. 29 shows a timing chart when the return delay time (RTD) is 8 frame times. In FIG. 29, the modulo is 16. In the case of FIG. 29, 2 for frame number 3
The modulo of the transmission frame number rotates once before the reject reaches the transmission side, and a new frame having the same frame number 3 as the oldest retransmission waiting frame is transmitted on the reception side and received on the reception side. . In this case, the receiving side cannot determine whether the frame of frame number 3 is a new frame or a retransmission frame. for that reason,
Although the received frame is a new frame, the data of the new frame of frame number 3 is added to the head of the received frames 4 to 14 stored in the buffer and transferred to the user, As a result, the data sequence in which the order of the data blocks is changed is passed to the user. In this case, it is necessary to detect an abnormality on the application side and restart the transmission procedure, so that a large video delay occurs. If the identification bits of the old and new frames are added or the number of modulo bits is increased to avoid this, there is a problem that the overhead of the frame increases.

According to the multiplexing method proposed by the present applicant, the following two problems need to be solved.
The first problem arises because the transmission frame of the video data is composed of frames of two types as shown in FIG. The size of each frame can be determined only by the multiplex structure corresponding to the multiplex information stored in the header of the multiplex frame, and the other frame cannot be transmitted in both multiplex areas. Therefore, even if a transmission error occurs in a video frame and a retransmission request is returned from the receiving side, the retransmission is delayed until a multiplexed frame capable of storing the corresponding frame is transmitted. As a result, delay jitter occurs in the retransmission frame. This situation is shown in the timing chart of FIG. In FIG. 30,
A transmission error occurs in video frame 1 transmitted by terminal 1 multiplexed with an audio frame. After receiving this, terminal 2 detects the error and returns a rejection of video frame 1. In FIG. 30, the numbers marked on the video frames indicate frame numbers, and 1 + R indicates rejection of frame number 1. This reject is received by the terminal 1 while transmitting the video frame 5 in the case of FIG. Terminal 1
Should perform retransmission immediately, but since the multiplexing structure of the multiplexed frame storing the next video frame is different from the multiplexing structure of the multiplexed frame that transmitted video frame 1, video frame 1 Cannot be resent. Therefore, retransmission of video frame 1 is delayed until the video area of the next multiplexed frame.

When a plurality of audio frames are multiplexed on one multiplexed frame, of the plurality of frames,
Since the size of the multiplexed video frame changes depending on the number of silence frames, the delay jitter caused by waiting for the same multiplexed frame becomes larger.

The second problem is that when the header is arranged before and after the multiplexed frame, the redundancy becomes high. Conventionally, the header has an error correction code part which is a redundant bit for error correction. However, when the header is arranged before and after the multiplexed frame, the number of bits occupied by the header is simply doubled, and the number of redundant bits is increased. Is also doubled.

In wireless communication in which errors occur in a burst, the effect of arranging two headers before and after a multiplexed frame is very large. For example, when an 80-byte multiplexed frame is transmitted by PHS, the error rate of the header when two 15-bit headers are arranged before and after the multiplexed frame is the error rate when only one 15-bit header is arranged. It becomes 1/10 or less. On the other hand, when only one 31-bit header is arranged, the error rate of the header is:
This is only about 1/2 of the case where only one 15-bit header is arranged.

As described above, in wireless communication in which errors occur in bursts, the effect of arranging two headers before and after a multiplexed frame is very large, but the redundancy is doubled. The problem arises.

As described above, FIG. 23 shows the sequence when the header is changed to the format arranged before and after the frame. However, in a situation where the counter (HC) may be inserted, the header may be changed. At a minimum, it is necessary to have a size that can store the multiplexed information section, the counter, and the error detection code section, so that 2 bytes are required as shown in FIG. However, if a 2-byte header is always used without a counter (HC), the redundancy increases.

By the way, H. In a video frame of “223 Dec. 1996 Draft” of H.223 / AnnexA, if the video packet requested to be retransmitted is small and does not reach the size of the video frame provided from the multiplexing layer, dummy bits are stuffed in byte units. Staff area is provided. This staff area will be described below.

The presence or absence of the stuff area is determined by the Forward Cont of the ARQ header of the video frame shown in FIG.
roll field stuffing type ST (H.223 / Annex A, “6 Dec.
27, the PT shown in FIG. 27 disappears, and the Backward Cont.
roll Field). As shown in FIG. 32, the stuff area immediately follows the ARQ header and has at least one byte. The stuff area is composed of two types of stuffing messages, CBS and LBS. CBS “00000000” indicates that the stuff area continues in the next byte, and LBS “1111111”.
1 "indicates that the stuff area ends with this byte. The receiving side discards the stuffed CBS and LBS messages.

In addition, H.S. In a video frame of H.223 / Annex A, a divided packet transmission message for dividing and transmitting a video packet when the retransmission-requested video packet is large and cannot fit in the size of the video frame given from the multiplexing layer is included. Is provided. This divided packet transmission message will be described below.

Backward Control Field of the ARQ header of the video frame shown in FIG.
31 is shown in FIG. N (R) in the reject message is a frame number for which retransmission is requested by rejection. N (S) in the fragmented packet transmission message is a subframe number assigned to each divided into a plurality of packets.

By using the above two functions, a large-sized video packet which could not be transmitted due to multiplexed audio is divided into two or more as shown in FIG. , And multiplexing with voice.

FIG. 34 is a flowchart showing the processing operation of multiplex transmission using the divided packets. This figure 3
4 will be described with reference to FIG. The frame requested to be retransmitted is placed in a retransmission waiting queue. It is determined whether there is an audio frame to be multiplexed with the video frame to be transmitted next (step S).
4201), if there is an audio frame, the size of the video frame becomes small. If there is no retransmission video frame in the retransmission waiting queue (step S4220), the video frame having the small size of the initial transmission is transmitted (step S4211). If there is a retransmission video frame in the retransmission waiting queue, the size of the video frame is determined (S422).
1) If it enters a video frame, it is transmitted (step S4204). If it is not large, it is divided into two and the first retransmitted video frame (1) is transmitted (step S4207). If there is no audio frame to be multiplexed, the size of the video frame becomes large. If there is no retransmission video frame in the retransmission waiting queue (step S4222), the large video frame of the first transmission is transmitted (step S4219). If there is a retransmission video frame in the retransmission waiting queue, the size of the video frame is determined (step S4223), and if it is the same size as the video frame, it is transmitted (step S42).
14) If it is smaller, the bit is transmitted as stuff (step S4216).

FIG. 35 is a timing chart showing the processing operation of multiplex transmission using the above-mentioned divided packets. In FIG. 35, a transmission error occurs in the video frame 3 transmitted by the terminal 1 multiplexed with the audio frame. After receiving the transmission error, the terminal 2 detects the error and returns the rejection of the video frame 3. In FIG. 35, Video3 is a video frame of frame number 3, and R3 is a reject for frame number 3. In the case of the timing shown in FIG.
Is received. Terminal 1 multiplexes Video3 with Video3 (1) and Video3 (1) because there are audio frames to be multiplexed.
3 (2), first transmit Video3 (1),
Video3 (2) is returned to the head of the retransmission waiting queue. Then, in the next video frame, bits are stuffed and transmitted to Video3 (2).

However, when the division is performed as shown in FIGS. 33 and 35, a new problem arises that the transmission efficiency is deteriorated by the bit stuffing.

Accordingly, an object of the present invention is to provide a retransmission delay jitter absorbing method capable of effectively absorbing jitter due to retransmission delay.

[0035]

Means for Solving the Problems and Effects of the Invention The first invention is a method for transmitting data bidirectionally using a fixed-length frame to which a frame number and an error detection code are added. When an error is detected, a retransmission control method for requesting retransmission by returning a frame number and a reject command of a frame in which an error is detected using a specific area secured in a data area of a transmission frame. Is characterized in that the frame number of the transmission frame is transmitted using the specific area even when it is not detected that there is an error in the reception frame.

As described above, according to the first aspect, even when no error is detected in the received frame and it is not necessary to return the reject command, the specific area ( For example, Backward
By transmitting the frame number of the transmission frame using Control Field), resistance to transmission errors is increased, and quick rejection is possible. As a result, video delay due to retransmission can be reduced even in an environment with a large network delay.

According to a second aspect, in the first aspect, when the transmission error rate or the retransmission frequency is lower than a predetermined threshold value and the reception frame is not detected to be erroneous, The transmission of the frame number of the transmission frame performed using the specific area is stopped.

As described above, according to the second aspect, when the transmission error rate or the retransmission frequency is lower than a predetermined threshold value, it is not necessary to transmit the frame number of the transmission frame twice. Therefore, in such a case, the transmission of the frame number of the transmission frame using the specific area is stopped, so that the reduction of the transmission efficiency due to the double transmission of the transmission frame number is suppressed as much as possible.

According to a third aspect, in the first or second aspect, when transmitting the frame number of the transmission frame using the specific area, the transmission mode has the same code system as the reject command, and the frame number is A code indicating a number is transmitted together.

As described above, according to the third aspect, when the frame number of the transmission frame is transmitted twice, a code indicating that the frame number stored in the specific area is the transmission frame number is transmitted together. By doing so, the receiving side can easily recognize the transmission frame number.

According to a fourth aspect, in any one of the first to third aspects, the specific area is arranged at the end of the data area.

As a characteristic of PHS and wireless data communication, since errors tend to occur in bursts, even if frame numbers are adjacently transmitted twice, there is a high probability that an error will occur in both of them, and the frame numbers will be transmitted without errors. Can be transmitted only about twice. However, when the frame number is transmitted twice with an interval as in the fourth aspect, the probability that the frame number can be transmitted without error is dramatically improved as compared with the case of a single header. As a result, compared with the case where the frame numbers are transmitted adjacently and doubly, the error resilience of the information can be improved while the redundancy remains the same, and the probability that the reject can be returned quickly increases.

According to a fifth aspect of the present invention, when data is transmitted bidirectionally using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, the transmission frame In the retransmission control method for requesting retransmission, by using the specific area secured in the data area and returning the frame number and the reject command of the frame in which the error has been detected, there may be an error in the received frame. Even if it is not detected, the frame number of the frame for which retransmission was requested last time and the reject command are repeatedly retransmitted using the specific area.

In the case of the conventional method in which the reject is sent only once, if an error occurs in the reject, the reject is returned again after waiting for the timeout of the timer activated on the side that returned the reject, so that the retransmission delay is enormous. become. On the other hand, according to the fifth aspect, the same reject is continuously transmitted a plurality of times.
Even if an error occurs in rejection, if there is no error in the next rejection, a retransmission frame can be obtained earlier than waiting for a timeout. As a result, video delay due to retransmission can be reduced.

According to a sixth aspect, in the fifth aspect, when the transmission error rate or the retransmission frequency is lower than a predetermined threshold, and when no error is detected in the received frame, The retransmission of the reject command and the frame number of the frame requested to be retransmitted last time using the specific area are stopped.

As described above, according to the sixth aspect, when the transmission error rate or the retransmission frequency is lower than a predetermined threshold value, there is little possibility that an error occurs in the transmitted rejection. In such a case, by stopping the reject transmission using the specific area,
A reduction in transmission efficiency due to continuous reject transmission is minimized.

According to a seventh aspect of the present invention, in the fifth or sixth aspect, a code different from the first transmitted reject command is used as the retransmitted reject command.

As described above, according to the seventh aspect, a code different from the reject command transmitted first is used as the reject command to be retransmitted. It is possible to identify whether the command was transmitted first or retransmitted.

According to an eighth aspect of the present invention, when data is bidirectionally transmitted using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, In the retransmission control method for requesting retransmission, by using the specific area secured in the data area and returning the frame number and the reject command of the frame in which the error has been detected, there may be an error in the received frame. Even if it is not detected, the frame number of the oldest unarrived frame and the reject command are transmitted using the specific area.

As described above, according to the eighth aspect, even when no error is detected in the received frame and it is not necessary to return the reject command, the specific area ( For example, Backward
By transmitting the frame number of the oldest unarrived frame and the reject command using Control Field, it is possible to inform the other party whether there is a possibility that the modulo will make one revolution.

According to a ninth aspect, in the eighth aspect, when the frame number of the oldest unarrived frame is received together with the reject command, a transmittable frame number is calculated based on the received frame number. Features.

As described above, according to the ninth aspect, when the frame number of the oldest unarrived frame is received together with the reject command, the transmittable frame number is calculated based on the received frame number. This prevents the modulo from rotating once, and eliminates the delay in video that occurs when the frame order is out of order and restarts from initialization, and the retransmission wait that occurs due to differences in the multiplexing structure. .

In a tenth aspect based on the eighth or ninth aspect, when all of the frames requested to be retransmitted have arrived and no error is detected in the received frame, the specific area is set to the specified area. And transmitting a predetermined command other than the reject command.

As described above, according to the tenth aspect, all the frames for which retransmission has been requested have arrived.
If no error is detected in the received frame, a command other than the reject command is transmitted using the specific area, thereby eliminating unnecessary data transmission.

In an eleventh aspect based on the eighth or ninth aspect, when all of the frames requested to be retransmitted have arrived and no error is detected in the received frame, the specific area is set to the specified area. In this case, a transmission frame number and a code indicating that the frame number is a transmission frame number having the same code system as the reject command are transmitted.

As described above, according to the eleventh aspect, the case where all the retransmission-requested frames have arrived,
If no error is detected in the received frame, the frame number of the transmitted frame using the specific area,
Since a code indicating that the frame number is a transmission frame number is transmitted, resistance to transmission errors is increased, and quick rejection is possible. As a result, video delay due to retransmission can be reduced even in an environment with a large network delay.

According to a twelfth aspect, when data is bidirectionally transmitted using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, an error is detected. In the retransmission control method for requesting retransmission by returning the frame number of the detected frame and the reject command, if there is a frame to be transmitted, a specific area for storing predetermined information is set before and after the frame. And if no error is detected in the received frame, using a specific area,
When the frame number of the transmission frame is transmitted and an error is detected in the reception frame, the frame number of the transmission frame, the frame number of the frame in which the error was detected, and the reject command Is transmitted.

As described above, according to the twelfth aspect, the specific area for storing the predetermined information is doubly arranged at an interval before and after the transmission frame, so that the PHS or the wireless data communication can be performed. , It is possible to transmit information strongly against a burst error.

According to a thirteenth aspect, in the twelfth aspect,
A redundant code for error correction is added to the frame number of the transmission frame stored in the specific area, and when an error is detected in the received frame, the redundant code for error correction is shortened to the specific area. Is stored with a frame number of a frame in which an error has been detected and a reject command in a vacant portion.

As described above, according to the thirteenth aspect, the frame number of the frame in which the error has been detected and the reject command are stored in the vacant portion of the specific area formed by shortening the redundant code for error correction. Because it is stored
Transmission frame overhead can be reduced.

According to a fourteenth aspect of the present invention, when n and m are both positive integers and n <m is satisfied, n multiplexed frames out of a basic multiplexed frame sequence composed of m multiplexed frames are provided. When a voice frame is multiplexed and transmitted in a multiplexed frame, or when n voice frames are multiplexed and transmitted in one multiplexed frame, the area other than the voice frame in the multiplexed frame is stored in an area. A video frame of a semi-fixed length with at least a frame number and an error detection code is stored, and when an error is detected in the received frame, the video frame stored in the received frame in which the error is detected is retransmitted. In such a multiplex transmission system, different retransmission queues are prepared for different frame lengths of different video frames according to the type and number of audio frames to be multiplexed. When a video frame is retransmitted, if there is a retransmission wait in the retransmission queue corresponding to the video frame length in the multiplexed frame used for retransmission, the retransmission waiting frame is retransmitted. And transmitting a new video frame of the same length as.

As described above, according to the fourteenth aspect, it is possible to prevent a decrease in transmission efficiency caused by stuffing bits when a video packet is divided.

According to a fifteenth aspect, in the fourteenth aspect,
A threshold value is set in advance for the number of retransmission waiting frames accumulated in the retransmission queue, and when the number of retransmission waiting frames accumulated in the retransmission queue exceeds the threshold value, If the frame length is smaller than the video frame length in the multiplexed frame used for retransmission, the retransmission waiting frame is transmitted by stuffing the lacking portion, and the frame length of the retransmission waiting frame is within the multiplexed frame used for retransmission. If the video frame length is larger than the video frame length, the frame is transmitted by dividing the frame to be retransmitted.

As described above, according to the fifteenth aspect, it is possible to prevent a reduction in transmission efficiency caused by indiscriminately dividing video packets and stuffing bits, and to retransmit to a specific retransmission wait queue. It is possible to prevent retransmission delay jitter caused by accumulation of video frames.

According to a sixteenth aspect, in the fourteenth aspect,
A threshold value is set in advance for the number of retransmission waiting frames accumulated in the retransmission queue, and when the number of retransmission waiting frames exceeds the threshold value, the frame length of the retransmission waiting frame is used for retransmission. When the frame length is smaller than the video frame length in the multiplexed frame, the lacking portion is defined as a data frame area or a dummy area and the other type of data is stored to transmit the retransmission waiting frame, and the retransmission waiting frame is transmitted. When the frame length is longer than the video frame length in the multiplex frame used for retransmission, the retransmission waiting frame is transmitted by dividing the retransmission waiting frame.

As described above, according to the sixteenth aspect, it is possible to prevent a reduction in transmission efficiency caused by indiscriminately dividing video packets and stuffing bits, and to retransmit to a specific retransmission wait queue. It is possible to prevent retransmission delay jitter caused by accumulation of video frames.

According to a seventeenth aspect, when the first to k-th data frames (k is an integer of 2 or more) to which a frame number and an error detection code are added are multiplexed and transmitted by a multiplexed frame, In the multiplex transmission system that retransmits the data frame when the frame number received without error in the first data frame is detected without error in the multiplex information indicating the multiplex structure of the multiplex frame, The first header to which a code is added is arranged before and after the multiplexed frame.

As described above, according to the seventeenth aspect, since the first header is arranged before and after the multiplexed frame, it is possible to transmit information strong against burst errors in PHS and wireless data communication. Become.

According to an eighteenth aspect, in the seventeenth aspect,
In the initial state, an error detection code is added to the multiplexed information,
The second of these has redundant bits added for error correction.
Is placed only at the beginning of the multiplexed frame, a counter is added to the multiplexed information to notify the change timing of the setting of the multiplexed frame, and an error detection code is added to the multiplexed information and the counter. Redundant bits for error correction are added, and a third header having the same number of bits as the second header is arranged only at the beginning of the multiplexed frame. It is characterized by being arranged in.

Normally, it is often sufficient to change the setting of the multiplexed frame only at the time of initialization. Therefore, in the twenty-first aspect, setting is performed with a header with error correction redundancy only at the time of initial setting, and after setting is completed, headers without error correction redundancy are arranged before and after the multiplexed frame,
The error rate of the header is reduced without increasing the overhead. As described above, according to the eighteenth aspect, by reducing the error rate of the header, it is possible to reduce the retransmission delay caused by discarding the transmission data without increasing the overhead.

According to a nineteenth aspect, when the first to k-th data frames (k is an integer of 2 or more) to which a frame number and an error detection code are added are multiplexed and transmitted by a multiplexed frame, In the multiplex transmission system that retransmits the data frame when the frame number received without error in the first data frame is detected without error in the multiplex information indicating the multiplex structure of the multiplex frame, One of the first header to which a code has been added and the second header to which an error detection code has been added to the multiplexed information and redundant bits for error correction have been added thereto. It is characterized in that it is arranged before a frame and one of them is arranged after a multiplexed frame.

According to a twentieth aspect, in the nineteenth aspect,
When changing the setting of the multiplex frame, the second header is
A counter for notifying the change timing of the setting of the multiplexed frame is added to the multiplexed information, an error detection code is added to the multiplexed information and the counter, and a redundant bit for error correction is added thereto. Is replaced with a third header having the same number of bits as the header.

As described above, according to the nineteenth and twentieth aspects, by reducing the error rate of the header, the setting can be changed while reducing the retransmission delay caused by discarding the transmission data, and the overhead can be reduced. Can be reduced.

According to a twenty-first aspect, in the nineteenth aspect,
The second header includes information effective only when the multiplexed frame is located before or after the multiplexed frame.

According to a twenty-second aspect, when the first to k-th data frames (k is an integer of 2 or more) to which a frame number and an error detection code are added are multiplex-transmitted by a multiplexed frame, In the multiplex transmission system that retransmits the data frame when the frame number received without error in the first data frame is detected without error in the multiplex information indicating the multiplex structure of the multiplex frame, The first header to which the code is added is disposed before or before or after the multiplexed frame, and a counter for notifying the change timing of the setting of the multiplexed frame is added to the multiplexed information. A third header having a different number of bits from the first header in which an error detection code is added to the first header and redundant bits for error correction are added to the multiplexed frame. Characterized in that prior to or placed before and after the.

According to a twenty-third invention, in the twenty-second invention,
When the setting of the multiplexed frame is changed in a state where the first header is arranged before and after the multiplexed frame, one of the first headers is replaced with a third header.

According to a twenty-fourth aspect, when at least a video frame and a data frame are multiplexed and transmitted in a multiplexed frame, a frame number of the video frame and / or a supervisory message of the video frame are multiplexed and transmitted as a data frame. It is characterized by the following.

As described above, according to the twenty-fourth aspect, the frame number of the video frame and / or the supervisory message of the video frame are multiplexed in the data frame and are transmitted twice, thereby making the data frame effective. It can be utilized, and the error tolerance of the header of the video frame and the monitoring frame can be enhanced.

According to a twenty-fifth aspect, in the twenty-fourth aspect,
A frame number of a video frame received as a data frame and / or a monitoring message of a video frame are used for processing a video frame multiplexed in the same multiplexed frame or the monitoring message thereof.

According to a twenty-sixth aspect, when at least a video frame and a control frame are multiplexed on a multiplexed frame and transmitted, a frame number of the video frame and / or a supervisory message of the video frame are multiplexed and transmitted as a control frame. It is characterized by the following.

As described above, according to the twenty-sixth aspect, the frame number of the video frame and / or the supervisory message of the video frame are multiplexed in the control frame and are transmitted twice, thereby making the control frame effective. It can be utilized, and the error tolerance of the header of the video frame and the monitoring frame can be enhanced.

According to a twenty-seventh aspect, in the twenty-sixth aspect,
The frame number of the video frame received as the control frame and / or the monitoring message of the video frame are used for processing the video frame multiplexed in the same multiplexed frame or the monitoring message thereof.

[0083]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 is a timing chart for explaining a retransmission delay jitter absorbing method according to a first embodiment of the present invention. In the first embodiment, data is bidirectionally transmitted in a fixed-length frame with a frame number and an error detection code, and when an error is detected in a received frame, the data area of the transmitted frame is reduced by a certain amount. In the retransmission control method of returning the frame number and reject command of a frame in which an error is detected in the area where the error has been detected, even if it is not necessary to return the frame number and reject command, the transmission By transmitting this frame number, retransmission of a transmitted video frame and rejection of a video frame in which an error is detected during reception can be simultaneously returned.

In FIG. 1, terminals 1 and 2 are Bac
In the area of the keyword Control Field, F
Like the older Control Field,
The video frame is transmitted by storing the forward control information of the video frame being transmitted. When a transmission error occurs in video frame 1 transmitted by terminal 2, terminal 1
Is the Backward Contr of video frame 6
ol Field rejects video 1 reject. Thus, when returning a reject, Bac
There is a kward Control Field.
The terminal 2 receives the reject of the video 1 and attempts to retransmit the video 1, but a retransmission wait occurs due to a difference in the multiplexing structure. If an error occurs in the video frame 7 transmitted by the terminal 1 during this retransmission waiting, the reception process of the terminal 2 detects this. Accordingly, terminal 2 can simultaneously return the retransmission of video frame 1 and the rejection of video frame 7 in which an error has been detected in the last multiplexed frame of the multiplexed frame sequence shown in the transmission processing of terminal 2.

As described above, Backward C
The fact that the area of the control field is left empty for transmitting information other than video frames has a negative effect on transmission efficiency, but the frame number is transmitted twice using this free area. 2 for transmission error
Demonstrate twice the strength. H. If the Forward Control Field also has an error detection code as in H.223 / AnnexA, even if an error is detected in the video data of the received video frame, if there is no error in the frame number, a reject can be returned immediately. . However, a frame in which an error is detected in the frame number is discarded, and when a next frame having no error in the frame number is received, a missing number is detected, and a reject is returned for the first time. In other words, having a double frame number means that the probability of quickly returning a reject is doubled.

As described above, in the first embodiment, even when there is no need to return the reject command,
By effectively using the Ward Control Field and suppressing the retransmission occurrence while absorbing the retransmission delay jitter of the video frame, it is possible to reduce the video delay due to the retransmission even in an environment with a large network delay.

In the present embodiment, when the number of errors is small, only the downside that the transmission efficiency is reduced appears. When the number of errors is small, Backward is used.
An area of the control field may be allocated for storing video frames. The error rate serving as a criterion in this case may be obtained from a communication device such as a PHS, may be obtained from the error frequency of a multiplexed frame or a video frame, or may be obtained from the frequency of a video frame retransmission request. Is also good.

Also, Backward Control
Forward Control Fi in Field
When the information of the field is transmitted twice, a code indicating that the information of the field is the information of the Forward Control Field may be transmitted in the SM (see FIG. 27) described in the related art.

Further, although the S-Rej system is used as the video retransmission system here, the same effect as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Second Embodiment) FIG. 2 shows a second embodiment of the present invention.
6 is a timing chart for explaining a retransmission delay jitter absorbing method according to the embodiment. In the second embodiment, the first embodiment described above differs from
Forward Control in control field
Instead of transmitting the ol Field information twice, the previously returned reject is continued to be sent in the Backward Control Field until the reject is updated. This makes it possible to simultaneously return the retransmitted video frame and the rejection of the video frame for which an error was detected during reception. In the case of the conventional method in which the reject is sent only once, when an error occurs in the reject, the reject is returned again after waiting for the timeout of the timer activated on the side that returned the reject, so that the retransmission delay becomes enormous. On the other hand, in the second embodiment, since the same reject is continuously transmitted a plurality of times, even if an error occurs in the reject, if there is no error in the next reject, it is faster than waiting for a timeout. A retransmission frame can be obtained.

As described above, in the second embodiment, even when there is no need to return the reject command,
By effectively using the Ward Control Field and suppressing the retransmission occurrence while absorbing the retransmission delay jitter of the video frame, it is possible to reduce the video delay due to the retransmission even in an environment with a large network delay.

In the second embodiment, when the number of errors is small, only the downside that the transmission efficiency is reduced appears.
An area of dControl Field may be allocated for storing video frames. The error rate serving as a criterion in this case may be obtained from a communication device such as a PHS, may be obtained from the error frequency of a multiplexed frame or a video frame, or may be obtained from the frequency of a video frame retransmission request. Is also good.

[0093] While the reject is not returned even once, the Backward Control Field is used.
In the area of Forward Control Field
May be transmitted twice.

[0094] Also, Backward Control
When transmitting the information of the previous reject in the field, the reject stored in the field of the SM described in the related art (see FIG. 27) is the information of the repeat of the previously transmitted reject. Is preferably transmitted.

Furthermore, although the S-Rej system is used here as the video retransmission system, the same effects as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Third Embodiment) FIG. 3 shows a third embodiment of the present invention.
6 is a timing chart for explaining a retransmission delay jitter absorbing method according to the embodiment. In the first embodiment described above, the Backward Control Field
d, the information of the Forward Control Field is transmitted twice. In the third embodiment, however, the Backward Control Field is used.
In ld, the frame number and reject of the oldest unarrived frame are continuously sent until the oldest unarrived frame is received, thereby preventing the inconvenience due to one rotation of the modulo. As described in the section of “Problems to be Solved by the Related Art and the Invention”, when waiting for the oldest unarrived frame, rejection of the same frame is repeated, and when the modulo makes one revolution, the frame order is changed. It goes wrong and you need to reinitialize the transmission and start over. In this case, a large delay occurs in the video.
In the present embodiment, when there is a frame that has not arrived on the receiving side, the frame number of the oldest frame is continuously sent to the transmitting side, and on the transmitting side, the frame number of the transmitting frame is (the frame number-1). , Stop transmitting new frames. As a result, it is possible to prevent the modulo from rotating once, and it is possible to eliminate a video delay that occurs due to a frame order being out of order and restarting from initialization, and a retransmission wait that occurs due to a difference in multiplexing structure.

Since the notification of the non-arriving frame may not reach the transmitting side due to a transmission error, it is always notified when there is no need to return a reject command for requesting retransmission of the received frame or other necessary commands. It is desirable to keep. When the oldest unarrived frame that has been notified so far is received, and when there is a subsequent unreceived frame, the frame number is continuously notified.

Also, in order to restart the transmission when the transmission of a new frame is stopped or to know on the transmission side that there is no need to stop the transmission, a frame in which no transmission error has occurred, or a frame in which a reject is returned Are all retransmitted without error, the Backward Control F
command other than reject in the area of
information of the old Control Field
A duplicate command is transmitted or a command indicating that there is no command to be transmitted is transmitted. However, if a command indicating that there is no command to be transmitted is transmitted, the efficiency is reduced. On the sending side,
When this is received, transmission is restarted, and there is no need to stop transmission until the oldest unarrived frame is notified again.

As described above, according to the third embodiment,
Ba also when it is not necessary to return the reject command
By effectively utilizing the ckward control field, it is possible to prevent a modulo rotation while suppressing a video delay due to retransmission even in an environment with a large network delay.

In the third embodiment, when the number of errors is small, only the downside that the efficiency is reduced appears. Therefore, when the number of errors is small, Backward is used.
The control field area may be allocated to video frames. The error rate serving as a criterion in this case may be obtained from a communication device such as a PHS, may be obtained from the error frequency of a multiplexed frame or a video frame, or may be obtained from the frequency of a video frame retransmission request. Is also good.

[0101] Also, Backward Control
When transmitting the information of the previous reject in the field, it is preferable to transmit the code indicating that the information is the oldest unreceived frame in the SM (see FIG. 27) described in the related art.

Furthermore, although the S-Rej system is used here as the video retransmission system, the same effects as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Fourth Embodiment) FIG. 4 shows a fourth embodiment of the present invention.
FIG. 10 is a diagram showing an example of a configuration of a video frame header used in the retransmission delay jitter absorbing method according to the embodiment. In FIG. 4, in the present embodiment, Backward Co
The control field is placed at the end of the video frame, and if there is no Backward information, the forward control information of the transmitted video frame is stored in the Backward Control Field in the same manner as in the Forward Control Field. Send video frames.

As a characteristic of PHS and wireless data communication, errors are likely to occur in a burst manner. Therefore, even if frame numbers are arranged and transmitted twice as in the first embodiment,
There is a high probability that an error will occur in both, and the probability of transmitting a frame number without an error is only about twice as high. However, if the frame number is transmitted twice at intervals,
The probability that the frame number can be transmitted without error can be reduced to about the square as compared with the case of a single header. As a result, the redundancy is kept the same as that of the first embodiment, and the Forward Control when the information of the Backward Control Field does not exist.
The error resilience of the 1 Field information is enhanced, and the probability that the reject can be returned quickly increases.

As described above, according to the fourth embodiment,
If it is not necessary to return the reject command,
By making effective use of the acknowledgment control field, the retransmission delay jitter of a video frame is absorbed and the occurrence of retransmission is suppressed. As a result, even in an environment where the network delay is large, it is possible to suppress the video delay due to the retransmission to a small value.

In the fourth embodiment, when the number of errors is small, only the downside that the efficiency is reduced appears, and when the number of errors is small, Backward is used.
The control field area may be allocated to video frames. The error rate serving as a criterion in this case may be obtained from a communication device such as a PHS, may be obtained from the error frequency of a multiplexed frame or a video frame, or may be obtained from the frequency of a video frame retransmission request. Is also good.

[0107] Also, Backward Control
Forward Control Fi in Field
In the case of transmitting the information of the Eld twice, the Forward Contrast by the SM described in the related art (see FIG. 27) is used.
It is preferable to transmit a code indicating that the information is control field information.

Furthermore, although the S-Rej system is used here as the video retransmission system, the same effects as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Fifth Embodiment) FIG. 5 shows a fifth embodiment of the present invention.
FIG. 10 is a diagram showing an example of a configuration of a video frame header used in the retransmission delay jitter absorbing method according to the embodiment. In FIG. 5, in the fifth embodiment, Forward C
ontrol Field and Backward C
Reduce the BCH part of the control field
In the area that was the BCH of the backward control field, the backward control
ld and forward control fi
eld and Backward Control Field
The error rate of the ARQ header is reduced by arranging a pair of d at the beginning and end of the video frame. In the present embodiment, the error correction capability is 5 bits in the related art because the BCH is reduced.
However, since the same information is arranged at intervals, it is less susceptible to burst errors. For this reason, in the data communication in wireless communication where burst errors frequently occur, the probability that the header information is erroneous is lower with the same overhead as in the first embodiment. Further, in the case of the present embodiment, not only the transmission frame number but also the frame number of the rejected frame is improved in error resilience, and the probability that the reject information is erroneous and a delay due to waiting for a timeout can be reduced can be reduced.

When it is not necessary to transmit a reject, the Backward Control Field
At d, the transmission frame number may be notified as in the first embodiment, or the oldest unreceived frame number may be notified as in the third embodiment.

As described above, according to the fifth embodiment,
By suppressing the occurrence of retransmission while absorbing the retransmission delay jitter of the video frame, it is possible to reduce the video delay due to retransmission even in an environment with a large network delay.

In the fifth embodiment, when the number of errors is small, only the downside that the efficiency is reduced appears, and when the number of errors is small, Backward is used.
The control field area may be allocated to video frames. The error rate serving as a criterion in this case may be obtained from a communication device such as a PHS, may be obtained from the error frequency of a multiplexed frame or a video frame, or may be obtained from the frequency of a video frame retransmission request. Is also good.

[0113] Also, Backward Control
Forward Control Fi in Field
In the case of transmitting the information of the Eld twice, the Forward Contrast by the SM described in the related art (see FIG. 27) is used.
It is preferable to transmit a code indicating that the information is control field information.

Further, as the video retransmission system, the S-Rej system is used here, but the same effects as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

Further, in the fifth embodiment, the BCH portion is reduced, but as shown in FIG.
ontrol Field and Backward con
control field as it is (ie, BC
Even if they are arranged twice before and after the frame (without reducing the H portion), strong resistance to burst errors can be obtained. In this case, since the BCH portion is intact, the error correction capability is 5 bits, but the overhead of the ARQ header is doubled.

In the fifth embodiment, the BCH is 4B
Although it is attached, it is not necessary to attach BCH to reduce the processing. In this case, although the error correction ability is lost,
In an environment where a burst error is dominant, an error reduction due to the duplication can be expected, and the effect of reducing redundancy is exhibited.

When the BCH is not added, as shown in FIG. 7, the Forward Control Field
d is 1.5 bytes, Backward Control
The field may be 1.5 bytes, and the size of each ARQ header may be 3 bytes in total to reduce overhead. In this case, S in the Backward Control Field in the portion where two bits are left
M may be extended, and Forward Control
Field and Backward Control F
The sequence number of field may be extended one bit at a time. Similarly, in FIG.
The SM in the control field may be extended by one bit, or the sequence numbers of the Forward Control Field and the Backward Control Field may be extended by one bit.

(Sixth Embodiment) In the sixth embodiment,
In order to prevent the transmission efficiency from being reduced by dividing the video packet, two retransmission queues are provided for each size. That is, in the present embodiment, there are two retransmission waiting queues for retransmission video frames (small) where audio frames are multiplexed, and retransmission waiting queues (large) for retransmission video frames where audio frames are not multiplexed. Prepare

FIG. 8 shows a flowchart of the processing operation in the sixth embodiment. The frame requested to be retransmitted is placed in a retransmission queue corresponding to the size of the frame. Thereafter, it is determined whether there is an audio frame to be multiplexed with the next video frame to be transmitted (step S440).
1) If there is an audio frame to be multiplexed, the size of the video frame becomes small, so it is determined whether or not there is a retransmission video frame in the retransmission waiting queue (small) (step S4402). If there is no retransmission video frame in the retransmission waiting queue (small), a small video frame for initial transmission is transmitted (step S4411), and if there is a retransmission video frame in the retransmission waiting queue (small), it is transmitted. (Step S4404). Step S above
If there is no audio frame to be multiplexed in 4401, the size of the video frame becomes large, so it is determined whether or not there is a retransmission video frame in the retransmission waiting queue (large) (step S4412). Retransmission queue (large)
If there is no retransmitted video frame, the video frame having the large size of the initial transmission is transmitted (step S4419),
If there is a retransmission video frame in the retransmission waiting queue (large), it is transmitted (step S4414).

FIG. 9 shows a timing chart of the processing operation in the sixth embodiment. In FIG. 9, a transmission error occurs in the video frame 3 transmitted by the terminal 1 multiplexed with the audio frame. After receiving the transmission error, the terminal 2 detects the error and returns the rejection of the video frame 3. In FIG. 9, Video3 is a video frame of frame number 3,
R3 indicates rejection for frame number 3. This reject is received by the terminal 1 during the transmission of the video frame 7 at the timing of FIG. Video3 for which retransmission has been requested has a large size because voice has not been multiplexed, and is thus placed in a retransmission waiting queue (large). Since the video frame transmitted next by the terminal 1 includes an audio frame to be multiplexed, Video3 is not retransmitted, and since there is no retransmitted video frame in the retransmission waiting queue (small), the first transmission Video8 is transmitted. in this case,
Video3 is retransmitted when there is no audio to be multiplexed next.

As described above, according to the sixth embodiment,
It is possible to prevent a decrease in transmission efficiency caused by stuffing bits when a video packet is divided.

Here, as an example of multiplexing frames in a sound section, audio was multiplexed at a rate of 2 frames to 3 frames. However, audio was multiplexed at a rate of 1 frame to 2 frames. The present invention is also effective for multiplexed frames in the same manner as described above. In addition, even when there are audio frames of different sizes, such as when a background sound is transmitted in a silent section, the present invention provides the above-described method by using a different retransmission queue for each video frame length corresponding to the audio frames. It has the same effect as.

The present invention is similarly effective when a multiplexed frame in which a dummy or data is defined is used instead of the stuffing.

Furthermore, although the S-Rej system is used here as the video retransmission system, the same effects as described above can be obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Seventh Embodiment) According to the sixth embodiment described above, after a video frame in which no sound is multiplexed continues in a silent section, it enters a voiced section, and a plurality of retransmission requests during a silent section are made. When a video frame having a large size must be retransmitted in a sound section, as shown in FIG. 10, retransmission video frames are accumulated in a retransmission waiting queue (large), and a large retransmission delay jitter occurs. The seventh embodiment solves the problem generated in the sixth embodiment.

In the seventh embodiment, similar to the sixth embodiment, a retransmission wait queue (small) for retransmission video frames in which audio frames have been multiplexed is used as a retransmission wait queue.
And a retransmission wait queue (large) for retransmission video frames in which audio frames have not been multiplexed. First and second thresholds are set for the respective queue waiting numbers. deep.

FIG. 11 shows a flowchart of the processing operation in the seventh embodiment. The retransmission requested frame is
It is placed in a retransmission waiting queue according to its size. Thereafter, it is determined whether there is an audio frame to be multiplexed with the next video frame to be transmitted (step S470).
1) If there is an audio frame to be multiplexed, it is determined whether or not there is a retransmission video frame in the retransmission waiting queue (small) (step S4702). If there is no retransmission video frame in the retransmission waiting queue (small), it is determined whether there is a retransmission video frame in the retransmission waiting queue (large) (step S4708). If there is a retransmission video frame in the retransmission waiting queue (large), it is determined whether or not the number of retransmission video frames (large) is larger than a predetermined second threshold value (step S4709). If the number of retransmitted video frames (large) is larger than the second threshold value, the retransmitted video frame at the head of the retransmission wait queue (large) is divided into two retransmitted video frames, and the first retransmitted video frame is divided. (1) is transmitted, and the second retransmitted video frame (2) is returned to the head of the retransmission wait queue (small) (step S470).
7).

On the other hand, in step S4702,
When there is a retransmission video frame in the retransmission waiting queue (small), it is determined whether or not the number of the retransmission video frames (small) is larger than a predetermined first threshold (step S4703). Determines whether the number of retransmitted video frames (large) is larger than a predetermined second threshold value (step S4705). If the number of retransmitted video frames (large) is larger than the second threshold, the retransmitted video frame at the head of the retransmission wait queue (large) is also divided into two retransmitted video frames, and the first retransmitted video frame is divided. The frame (1) is transmitted, and the second retransmission video frame (2) is returned to the head of the retransmission waiting queue (small) (step S4707). Step S4703 above
In step S47, if the number of retransmitted video frames (small) is larger than the first threshold value, or
At 05, if the number of retransmitted video frames (large) is smaller than the second threshold value, retransmitted video frames (small) are transmitted (step S4704). If there are no retransmission video frames in both the retransmission waiting queue (large) and (small), the first transmission video frame (small) is transmitted (step S4711).

If there is no audio frame to be multiplexed in step S4701, it is determined whether there is a retransmission video frame in the retransmission waiting queue (large) (step S4).
712). If there is no retransmission video frame in the retransmission waiting queue (large), it is determined whether there is a retransmission video frame in the retransmission waiting queue (small) (step S4712). If there are resending video frames in the resending queue (small),
It is determined whether the number of retransmitted video frames (small) is larger than a predetermined first threshold value (step S471).
8). If the number of retransmitted video frames (small) is larger than the first threshold, bits are stuffed and transmitted to the retransmitted video frames (small) (step S4716).
Also, in step S4712, when there is a retransmission video frame in the retransmission waiting queue (large), it is determined whether the number of retransmission video frames (large) is larger than a second predetermined threshold value (step S4712). S4713). If the number of retransmitted video frames (large) is smaller than the second threshold, it is determined whether the number of retransmitted video frames (small) is larger than the first threshold (step S471).
5). If the number of retransmitted video frames (small) is larger than the first threshold value, the bits are also stuffed and transmitted to the retransmitted video frames (small) (step S471).
6). In step S4713, when the number of retransmitted video frames (large) is larger than the second threshold,
Alternatively, if the number of retransmitted video frames (small) is smaller than the first threshold value in step S4715, retransmitted video frames (large) are transmitted (step S4714). If there are no retransmission video frames in both the retransmission waiting queue (large) and (small), the first transmission video frame (large) is transmitted (step S4719).

FIG. 12 is a timing chart of the processing operation in the seventh embodiment. In FIG. 12, the first and second threshold values in the flowchart of FIG.

In FIG. 12, a transmission error occurs in video frames 1, 2, and 3 (video frames in which audio frames are not multiplexed) transmitted by terminal 1, and after receiving this, terminal 2 detects the error and detects an error. Video frames 1, 2 and 3 are rejected. In the case of the timing shown in FIG.
eo5 is received by the terminal 1 during transmission. Video1 for which retransmission has been requested is large in size because voice has not been multiplexed, and is therefore placed in a retransmission waiting queue (large).
Since the video frame transmitted next by the terminal 1 has no audio frame to be multiplexed, the video 1 is immediately retransmitted. Next, the rejection for Video2 is Video
eo1 is received by terminal 1 during retransmission. Video2 for which retransmission has been requested has a large size because voice has not been multiplexed, and is thus placed in a retransmission waiting queue (large).
Since the video frame transmitted next by the terminal 1 has an audio frame to be multiplexed, Video2 is not retransmitted, and since there is no retransmission video frame in the retransmission waiting queue (small), Video6 of the first transmission is transmitted. . In addition, Vi
The reject for video3 is received by the terminal 1 while retransmitting Video2. Video3 requested for retransmission
Is placed in a retransmission waiting queue (large) because its size is large because the voice is not multiplexed. At this time, two retransmission video frames are accumulated in the retransmission waiting queue (large), and the number exceeds the threshold value. Therefore, the next video frame has audio frames to be multiplexed and is small in size. Video2 (1) and Video
2 (2), and first, Video2 (1) is transmitted in the next video frame, and Video2 (2) is returned to the head of the retransmission waiting queue (small). In the next video frame, since there is no audio to be multiplexed and neither of the retransmission waiting queues has exceeded the threshold value, Video3 having the same size is retransmitted. Then, in the next video frame, Video2
The bit is stuffed and transmitted in (2).

As described above, according to the seventh embodiment,
It is possible to prevent a reduction in transmission efficiency caused by unnecessarily dividing a video packet and stuffing bits, and prevent a retransmission delay jitter caused by accumulation of retransmission video frames in a specific retransmission wait queue.

Here, as an example of multiplexing frames in a sound section, audio was multiplexed at a rate of 2 frames to 3 frames. However, audio was multiplexed at a rate of 1 frame to 2 frames. The present invention is also effective for multiplexed frames in the same manner as described above. In addition, even when there are audio frames of different sizes, such as when a background sound is transmitted in a silent section, the present invention provides the above-described method by using a different retransmission queue for each video frame length corresponding to the audio frames. It has the same effect as.

Also, the present invention has the same effect as described above even if a multiplexed frame in which a dummy or data is defined is used instead of the stuffing.

Although the first and second threshold values are both set to 2 in this embodiment, the respective threshold values are:
What is necessary is just to set individually according to the structure of a multiplexing frame, or the quality of radio | wireless.

Furthermore, although the S-Rej system is used here as the video retransmission system, the same effects as described above are also obtained when the present invention is applied to the hybrid ARQ system combined with FEC.

(Eighth Embodiment) FIG. 13 is a diagram showing an example of the configuration of a video frame header used in a retransmission delay jitter absorbing method according to an eighth embodiment of the present invention.
In FIG. 13, in the present embodiment, the first header from which the BCH portion of the header of the multiplexed frame has been deleted is double-placed at the beginning and end of the multiplexed frame to reduce the header error rate. In the present embodiment, the error correction capability is lost because the BCH is deleted. However, since the same information is arranged at intervals, it is less susceptible to burst errors. For this reason, the probability of erroneous header information is lower in wireless data communication in which burst errors frequently occur, despite the same overhead as in FIG. 20 of the conventional example.

However, in the structure shown in FIG. 13, since a counter (HC) required for changing the setting cannot be arranged, the changing procedure shown in FIG. 14 is employed.

In the procedure shown in FIG. 14, in the initial state, an error detection code is added to the multiplexed information, and a second header to which redundant bits for error correction are added is added to the head of the multiplexed frame. Just have arranged. When the header is changed, a counter (HC) for notifying the change timing is added to the multiplexed information, and an error detection code is added to the header information and the counter (HC). A third multiplex header having the same number of bits as the second header to which the bits are added is arranged only at the head of the multiplex frame. Then, the first header is arranged before and after the multiplexed frame at the same time as the setting is changed.

In the prior art, when a counter (HC) for setting the change timing of the setting is not required, the header length is kept constant by allocating the corresponding bits to redundant bits for error correction. in this way,
Increasing the number of redundant bits for error correction when the number of headers is one can reduce the probability of erroneous headers and reduce retransmission delay caused by discarding transmission data. However, as in the present embodiment, the headers are arranged twice. In this case, the effect of reducing the probability of erroneous headers due to the use of doubles is remarkably large, so that redundant portions for error correction become extra.

Normally, it is often sufficient to change the setting only at the time of initialization. Therefore, in the present embodiment, the setting is performed with the header with the error correction redundancy only at the time of the initial setting, and after the setting is completed, the header without the error correction redundancy is arranged before and after the multiplexed frame to reduce the overhead. The error rate of the header is reduced without increasing.

As described above, according to the eighth embodiment,
By reducing the error rate of the header, it is possible to reduce retransmission delay due to discarding transmission data without increasing overhead.

Although the header is changed in the eighth embodiment, the same applies to the case where the multiplexed frame length, the synchronization flag length, other settings, or any combination thereof are changed.

(Ninth Embodiment) In the eighth embodiment,
After changing to the multiplexed frame of FIG. 13, the setting cannot be changed. It is possible that the setting needs to be changed even after the completion of the initial setting. In such a case, a method as shown in FIG. 15 is adopted.

In the procedure shown in FIG. 15, a first header in which an error detection code is added to multiplexed information is arranged at the rear of the multiplexed frame, and an error detection code and redundant bits for error correction are added to the multiplexed information. The second header thus arranged is arranged in front of the multiplexed frame. Then, the second header is replaced with a third header when the header is changed. The third header is configured by adding a counter for notifying the change timing to the multiplexed information, adding an error detection code to the header information and the counter, and further adding a redundant bit for error correction thereto. , Has the same number of bits as the second header.

In FIG. 15, after the initial setting, the front header includes a redundant portion for error correction, and when the setting is changed, a counter (HC) for taking a change timing can be arranged. , The redundant portion for error correction is deleted. In this method, the overhead is increased by 1 byte compared to the eighth embodiment, but the setting can be changed even after the initial setting, and the overhead is higher than that of the conventional method in which redundancy for error correction is added before and after. Is getting smaller.

As described above, according to the ninth embodiment,
By reducing the error rate of the header, it is possible to change the settings while reducing the retransmission delay caused by discarding the transmission data,
In addition, overhead can be reduced as compared with the related art.

In the ninth embodiment, the first header is arranged at the rear and the second or third header is arranged at the front. However, the first header is arranged at the front, and the second or third header is arranged at the front. The third header may be arranged at the rear.

In the ninth embodiment, the header is changed. However, the same applies to the case where the multiplexed frame length, the synchronization flag length, other settings, or any combination thereof are changed.

Further, in the ninth embodiment, information other than the purpose of taking the change timing is stored in the first header, but this information is stored in the case where the information is located before or after the multiplexed frame. Only valid information may be included. In any case, such header information is protected by an error correction code in the same header, and is protected by duplication in the entire multiplexed frame.

(Tenth Embodiment) In the eighth and ninth embodiments, it has been considered that the size of the header is always kept constant. Although the processing is somewhat complicated, a method as shown in FIG. 16 is also conceivable.

The procedure of FIG. 16 is to arrange the first header in which an error detection code is added to the multiplex information before and after the multiplex frame, and to notify the change timing when changing the setting of the multiplex frame length. After adding a counter to the multiplexed information and adding an error detection code to the header information and the counter, a third header (having a different number of bits from the first header) obtained by adding redundant bits for error correction thereto is added thereto. It is placed before the multiplex frame.

When this method is used, the processing is complicated to cut out headers of different sizes, but the setting can be changed any number of times while maintaining the same redundancy and error tolerance as in the eighth embodiment. is there.

As described above, according to the tenth embodiment, by reducing the error rate of the header, it is possible to change the setting while reducing the retransmission delay caused by discarding the transmission data, and to reduce the overhead compared to the related art. Can be reduced.

In the tenth embodiment, the case where the size changes only at the front of the double header is described. However, the size may change with one header. May change.

In the tenth embodiment, the multiplex frame length is changed. However, the same applies to the case where the synchronization flag length or other settings or any combination thereof is changed.

(Eleventh Embodiment) FIG. 17 is a diagram showing an example of multiplexing of multiplexed frames according to the eleventh embodiment of the present invention. An audio frame, a video frame, and a data frame are multiplexed in one fixed-length multiplex frame. 17, MC is a header of a multiplexed frame, A is an audio frame, V1 and V2 are video frames, D is a data frame, rej is a monitoring frame,
F1 and F2 are Forward Control F
field. Here, the header (MC) is 4
Bytes, audio frame (A) is 25 bytes, video frame (V1) is 31 bytes, video frame (V2) is 56 bytes, data frame (D) is 4 bytes or 2 bytes or 0 bytes, and monitoring frame (rej) is 2 bytes, Forward Control Field
d (F1) and (F2) are each 2 bytes.

Video Frame Forward Con
FIG. 18 shows the configuration of the control field. In FIG. 18, SN is a sequence number of video data, FEC
Is an error detection code for detecting an SN error (or an error correction code for correcting an SN error, or a combination of these codes). Here, the sequence number (SN) is 6 bits, and the error detection code (FE)
C) has 9 bits.

As described above, in the eleventh embodiment, information obtained by duplicating the header of a video frame multiplexed in a multiplexed frame and information obtained by duplicating a monitoring frame multiplexed in a multiplexed frame are stored in the multiplexed frame. Multiplexed in the data frame. As a result, the data frame (D) can be effectively used, and the error tolerance of the header of the video frame and the monitoring frame can be enhanced.

FIG. 19 shows the flow of information between the modules on the receiving side. In FIG. 19, the received multiplexed frame is separated into a data frame, a voice frame, and a video frame in the multiplexing layer 1. The separated audio frame is subjected to error control by the audio error control unit 2 and then passed to the audio information processing unit 4. The separated video frame is passed to the video information processing unit 5 after error control is performed by the video error control unit 3. here,
It is assumed that data error control is entrusted to the data processing unit 6, and in this example, the data processing unit 6 does not perform data error control. Further, the data processing unit 6 transfers the information passed from the multiplexing unit 1 to the video error control unit 3, as indicated by the thick arrow in the figure. The video error control unit 3 performs error detection (or error detection and error correction) on the information passed from the data processing unit 6, and when there is no error, the Forward Con
It is used as a control field or a monitoring frame.

When there is no reject message to return, as shown in FIG.
The information obtained by duplicating the control field (F1) and the Forward Control Field (F1)
The information obtained by duplicating 2) is created one by one, and those arranged are multiplexed as a data frame (D). In this case, the data frame (D) is passed to the video error control unit 223 by a thick arrow in FIG. 19, and the Forward Control Field of the video frame is
Error control.

When transmitting only one monitoring frame,
As shown in FIG. 17B, one 2-byte monitoring frame (rej) is multiplexed, and the frame length of the data frame (D) is reduced by the frame length of the monitoring frame (rej), 2 (= 4-2). ) Bytes. Furthermore, information in which the monitoring frame (rej) is duplicated is created and multiplexed as a data frame (D). In this case, the data frame (D) is passed to the video error control unit 3 by a thick arrow in FIG. 19, and is used for error control of the monitoring frame (rej).

In the case where two monitoring frames are transmitted, FIG.
As shown in FIG. 7C, the monitoring frames (rej1) and (rej2) are multiplexed, and the data frame (D) is reduced to two (2) monitoring frames by 0 (= 4-2-2) bytes.

In a more general multiplexed frame, information for duplication transmitted in the data frame (D) area is based on the Forward Control of a video frame multiplexed in the multiplexed frame.
The field and the monitoring frame may be arranged in this order. For example, in FIG. 17A, in order of F1 and F2, FIG.
7 (b), rej is the data frame (D)
Is stored in the area.

As described above, according to the eleventh embodiment, even in a multiplex transmission system which does not specify the Forward Control Field or the multiplexing of the monitoring frame, the data processing unit uses the present invention to duplicate them. Error tolerance can be enhanced. Further, the monitoring frame can be transmitted only when necessary without transmitting useless dummy data.

[Brief description of the drawings]

FIG. 1 is a timing chart for explaining a retransmission delay jitter absorbing method according to a first embodiment of the present invention.

FIG. 2 is a timing chart for explaining a retransmission delay jitter absorbing method according to a second embodiment of the present invention.

FIG. 3 is a timing chart for explaining a retransmission delay jitter absorbing method according to a third embodiment of the present invention.

FIG. 4 is a diagram illustrating an example of a configuration of a video frame header used in a retransmission delay jitter absorption method according to a fourth embodiment of the present invention.

FIG. 5 is a diagram showing an example of a configuration of a video frame header used in a retransmission delay jitter absorbing method according to a fifth embodiment of the present invention.

FIG. 6 is a diagram illustrating another example of the configuration of a video frame header used in the retransmission delay jitter absorption method according to the fifth embodiment of the present invention.

FIG. 7 is a diagram showing still another example of the configuration of a video frame header used in the retransmission delay jitter absorbing method according to the fifth embodiment of the present invention.

FIG. 8 is a flowchart showing a processing operation of a retransmission delay jitter absorbing method according to a sixth embodiment of the present invention.

FIG. 9 is a timing chart showing a processing operation of a retransmission delay jitter absorbing method according to a sixth embodiment of the present invention.

FIG. 10 is a timing chart for explaining a problem of the retransmission delay jitter absorbing method according to the sixth embodiment of the present invention.

FIG. 11 is a flowchart showing a processing operation of a retransmission delay jitter absorbing method according to a seventh embodiment of the present invention.

FIG. 12 is a timing chart showing a processing operation of a retransmission delay jitter absorbing method according to a seventh embodiment of the present invention.

FIG. 13 is a diagram illustrating an example of a configuration of a video frame header used in a retransmission delay jitter absorption method according to an eighth embodiment of the present invention.

FIG. 14 is a sequence chart showing a procedure at the time of changing the setting of a multiplexed frame in the retransmission delay jitter absorbing method according to the eighth embodiment of the present invention.

FIG. 15 is a sequence chart showing a procedure for changing the setting of a multiplexed frame in the retransmission delay jitter absorbing method according to the ninth embodiment of the present invention.

FIG. 16 is a sequence chart showing a procedure for changing the setting of a multiplexed frame in the retransmission delay jitter absorbing method according to the tenth embodiment of the present invention.

FIG. 17 is a diagram illustrating an example of multiplexing of multiplexed frames used in a twelfth embodiment of the present invention.

FIG. 18 is a diagram illustrating Forward C of a video frame in a multiplexed frame used in the twelfth embodiment of the present invention.
It is a figure which shows the structure of an control field.

FIG. 19 is a diagram showing a flow of information between modules on the receiving side of the multiplex transmission system according to the twelfth embodiment of the present invention.

FIG. First of multiplex layer of H.223 / AnnexA
FIG. 3 is a diagram showing a multiplexed frame format.

FIG. H.223 / AnnexA multiplexing layer second
FIG. 3 is a diagram showing a multiplexed frame format.

FIG. It is a figure showing the 1st format of the header in H.223 / AnnexA.

FIG. 23 is a sequence chart showing a procedure for changing the setting of a conventional multiplexed frame.

FIG. It is a figure which shows the 2nd and 3rd format of the header in H.223 / AnnexA.

FIG. 25 is a diagram showing a frame configuration of a conventional PIAFS.

FIG. 26 is a timing chart showing data transmission using a fixed-length multiplex frame proposed by the present applicant.

FIG. It is a figure which shows the header format of the video retransmission control frame in H.223 / AnnexA.

FIG. 5 is a timing chart for explaining video retransmission delay jitter occurring in H.223 / AnnexA.

FIG. FIG. 22 is a timing chart for explaining a modulo rotation at the time of video retransmission that occurs in H.223 / AnnexA.

FIG. 5 is a timing chart for explaining video retransmission delay jitter occurring in H.223 / AnnexA.

FIG. It is a figure which shows the structure of the ARQ header of the video frame in H.223 / AnnexA.

FIG. FIG. 3 is a diagram for explaining a stuff area of a video frame in H.223 / AnnexA.

FIG. It is a figure for demonstrating the packet division | segmentation transmission of the video frame using H.223 / AnnexA.

FIG. 34 is a flowchart showing a processing operation in packet division transmission.

FIG. 5 is a timing chart for explaining video retransmission delay jitter occurring in H.223 / AnnexA.

[Explanation of symbols]

MC: multiplexed information section CRC: error detection code section BCH: error correction code HC: counter SM: control command A: voice frame V1, V2: video frame D: data frame rej: monitoring frame F1, F2: Forward Control Fie
ld SN: sequence number of video data FEC: error detection code for detecting an error in SN 1: multiplexing layer 2: audio error control unit 3: video error control unit 4: audio information processing unit 5: video information processing unit 6 Data processing unit

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tsuyoshi Kuroda 1006 Kadoma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (27)

[Claims] When transmitting data bidirectionally using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, the data is included in the data area of the transmitted frame. Using the secured specific area,
In the retransmission control method for requesting retransmission by returning the frame number and reject command of the frame in which the error was detected, even if no error is detected in the received frame, the specific area is used. And transmitting a frame number of a transmission frame. 2. A transmission frame performed using the specific area when a transmission error rate or a retransmission frequency is lower than a predetermined threshold value and when no error is detected in a received frame. 2. The retransmission delay jitter absorbing method according to claim 1, wherein the transmission of the frame number is stopped. 3. When transmitting a frame number of a transmission frame using the specific area, a code having the same code system as the reject command and transmitting a code indicating that the frame number is a transmission frame number is transmitted together. The retransmission delay jitter absorbing method according to claim 1 or 2, wherein 4. The retransmission delay jitter absorbing method according to claim 1, wherein said specific area is arranged at the end of said data area. 5. When bidirectionally transmitting data using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, the data is included in the data area of the transmitted frame. Using the secured specific area,
In the retransmission control method for requesting retransmission by returning the frame number and reject command of the frame in which the error was detected, even if no error is detected in the received frame, the specific area is used. A frame number of a frame for which retransmission was previously requested and a reject command are repeatedly retransmitted. 6. When the transmission error rate or the retransmission frequency is lower than a predetermined threshold value and no error is detected in the received frame, the previous retransmission using the specific area is performed. The retransmission delay jitter absorbing method according to claim 5, wherein the retransmission of the frame number and the reject command of the frame for which the request has been made is stopped. 7. The retransmission delay jitter absorbing method according to claim 5, wherein a code different from a retransmission command transmitted first is used as the retransmission command to be retransmitted. 8. When data is bidirectionally transmitted using a fixed-length frame to which a frame number and an error detection code are added, if an error is detected in a received frame, the error is included in the data area of the transmitted frame. Using the secured specific area,
In the retransmission control method for requesting retransmission by returning the frame number and reject command of the frame in which the error was detected, even if no error is detected in the received frame, the specific area is used. Transmitting the frame number of the oldest unarrived frame and a reject command. 9. The apparatus according to claim 8, wherein, when a frame number of the oldest unarrived frame is received together with the reject command, a transmittable frame number is calculated based on the received frame number. The retransmission delay jitter absorption method as described in the above. 10. When all frames requested to be retransmitted have arrived and no error is detected in a received frame, a predetermined command other than the reject command is transmitted using the specific area. The retransmission delay jitter absorption method according to claim 8, wherein the retransmission delay jitter is transmitted. 11. When all the frames requested to be retransmitted have arrived and no error is detected in the received frame, the frame number of the transmission frame and the reject 10. The retransmission delay jitter absorbing method according to claim 8, wherein a code having the same code system as the command and a code indicating that the frame number is a transmission frame number is transmitted. 12. When data is bidirectionally transmitted using a fixed-length frame to which a frame number and an error detection code are added, when an error is detected in a received frame, the frame in which the error is detected. In the retransmission control method of requesting retransmission by returning the frame number and reject command of the frame, if there is a frame to be transmitted, a specific area for storing predetermined information is doubled before and after the frame. If no error is detected in the received frame,
The frame number of the transmission frame is transmitted using the specific region, and when it is detected that the reception frame has an error, the frame number of the transmission frame is transmitted using the specific region,
A retransmission delay jitter absorbing method, comprising transmitting a frame number of a frame in which an error has been detected and the reject command. 13. A redundant code for error correction is added to a frame number of a transmission frame stored in the specific area, and when an error is detected in a received frame, the redundant code for error correction is replaced with the redundant code for error correction. In the vacant part of the specific area created by shortening,
13. The retransmission delay jitter absorbing method according to claim 12, wherein a frame number of a frame in which the error is detected and the reject command are stored. 14. Both n and m are positive integers, and n <
When the relationship of m is established, a voice frame is multiplexed and transmitted to n multiplexed frames out of a basic multiplexed frame sequence composed of m multiplexed frames, or n multiplexed frames are transmitted to one multiplexed frame. When multiplexing and transmitting the audio frames of the multiplexed frame, at least a semi-fixed length video frame with a frame number and an error detection code is stored in an area other than the area where the audio frame is stored in the multiplexed frame. When an error is detected in a received frame, in a multiplex transmission system in which a video frame stored in the received frame in which the error is detected is retransmitted, depending on the type of audio frame to be multiplexed and the number of frames, Different retransmission queues are prepared for each frame length of different video frames, and when retransmitting video frames, If there is a retransmission wait in the retransmission queue corresponding to the video frame length, the retransmitting frame is retransmitted. If there is no retransmission waiting, a new video frame having the same length as the video frame length is transmitted. , Retransmission delay jitter absorption method. 15. A threshold value is set in advance for the number of retransmission waiting frames accumulated in the retransmission queue, and when the number of retransmission waiting frames accumulated in the retransmission queue exceeds the threshold value. If the frame length of the retransmission waiting frame is smaller than the video frame length in the multiplexed frame used for retransmission, the retransmission waiting frame is transmitted by stuffing the insufficiency and the frame length of the retransmission waiting frame is transmitted. 15. The retransmission delay jitter absorption according to claim 14, wherein, when is larger than the video frame length in the multiplexed frame used for retransmission, the retransmission waiting frame is transmitted by dividing the retransmission waiting frame. Method. 16. A threshold value is set in advance for the number of retransmission waiting frames accumulated in the retransmission queue, and when the number of retransmission waiting frames exceeds the threshold value, a retransmission waiting frame is set. If the frame length is smaller than the video frame length in the multiplexed frame used for retransmission, the lacking portion is defined as a data frame area or a dummy area, and another type of data is stored to transmit the retransmission waiting frame. If the frame length of the frame waiting for retransmission is longer than the video frame length in the multiplexed frame used for retransmission, the frame for retransmission is transmitted by dividing the frame for retransmission. Item 15. The retransmission delay jitter absorbing method according to Item 14. 17. When multiplexing and transmitting a first to k-th data frame (k is an integer of 2 or more) to which a frame number and an error detection code are added by a multiplexed frame, In a multiplex transmission system that retransmits a data frame when a frame number received without error in a data frame is detected, an error detection code is added to multiplex information indicating a multiplex structure of the multiplex frame. A method of absorbing retransmission delay jitter, wherein the first header is arranged before and after the multiplexed frame. 18. In an initial state, an error detection code is added to the multiplexed information, and a second header to which redundant bits for error correction are added is arranged only at the beginning of a multiplexed frame. A counter for notifying the change timing of the setting of the multiplexed frame is added to the multiplexed information, an error detection code is added to the multiplexed information and the counter, and a redundant bit for error correction is added to these. A third header having the same number of bits as that of the first header is arranged only at the beginning of the multiplexed frame, and the first header is arranged before and after the multiplexed frame at the same time as the setting is changed. Item 18. The retransmission delay jitter absorbing method according to Item 17. 19. When the first to k-th data frames (k is an integer of 2 or more) to which a frame number and an error detection code are added are multiplex-transmitted by a multiplex frame, the first data frame is received on the receiving side. In a multiplex transmission system that retransmits a data frame when a frame number received without error in a data frame is detected, an error detection code is added to multiplex information indicating a multiplex structure of the multiplex frame. One of the first header and the second header in which an error detection code has been added to the multiplexed information and redundant bits for error correction have been added to the multiplexed information. Wherein the other is arranged after the multiplexed frame. 20. When changing the setting of a multiplexed frame,
The second header is added to the multiplexed information with a counter for notifying the change timing of the setting of the multiplexed frame, an error detection code is added to the multiplexed information and the counter, and an error correction code is added to these. A third bit having the same number of bits as the second header with redundant bits added
20. The retransmission delay jitter absorbing method according to claim 19, wherein the retransmission delay jitter is replaced with a header of a retransmission delay. 21. The retransmission delay jitter absorbing method according to claim 19, wherein the second header contains information effective only when the frame is located before or after a multiplexed frame. 22. When the first to k-th data frames (k is an integer of 2 or more) to which a frame number and an error detection code are added are multiplexed and transmitted by a multiplexed frame, a first frame is transmitted to the receiving side. In a multiplex transmission system that retransmits a data frame when a frame number received without error in a data frame is detected, an error detection code is added to multiplex information indicating a multiplex structure of the multiplex frame. The first header is placed before or after the multiplexed frame, and a counter for notifying the change timing of the setting of the multiplexed frame is added to the multiplexed information. A third header having a bit number different from that of the first header to which a detection code is added and to which redundant bits for error correction are added is multiplexed. Characterized in that prior to or placed before and after the retransmission delay jitter method. 23. When the setting of the multiplexed frame is changed in a state where the first header is arranged before and after the multiplexed frame, one of the first headers is added to the third header. 23. The method according to claim 22, wherein
2. The retransmission delay jitter absorbing method according to 1. 24. When multiplexing at least a video frame and a data frame on a multiplexed frame and transmitting the multiplexed frame, a frame number of the video frame and / or a supervisory message of the video frame are multiplexed and transmitted as the data frame. A retransmission delay jitter absorbing method characterized by the above-mentioned. 25. A frame number of the video frame and / or a monitoring message of the video frame received as the data frame are used for processing a video frame multiplexed in the same multiplexed frame or the monitoring message thereof. 25. The retransmission delay jitter absorbing method according to claim 24, wherein: 26. When multiplexing at least a video frame and a control frame in a multiplexed frame and transmitting the multiplexed frame, a frame number of the video frame and / or a supervisory message of the video frame are multiplexed and transmitted as the control frame. A retransmission delay jitter absorbing method characterized by the above-mentioned. 27. A frame number of the video frame and / or a monitor message of the video frame received as the control frame are used for processing a video frame multiplexed on the same multiplexed frame or the monitor message thereof. The retransmission delay jitter absorbing method according to claim 26, characterized in that: