JPS6370643A - Packet communication system - Google Patents

Abstract

PURPOSE:To prevent the drop-out of data in the lump corresponding to packet length even if packet abolition takes place by providing a block sequence restoration information adding section and a block sequence restoration section or the like. CONSTITUTION:An analog signal inputted from an analog input terminal 1 is converted into a digital signal by a coder 2, formed into a packet via a packet forming section 3 and a block division section 4 and each packet is split into four packets and sent. Then each block is replaced among four packets by an inter-packet block replacement section 5 and sent to the block sequence restoration information adding section 6 in the order of replacement and sent to a transmission line 7 as a packet with restored information. In case of reception, the packet with restored information is inputted to a block sequence restoration section 9 via the line 8, the sequence of the block is restored and sent to a decoder 10, where the signal is converted into an analog signal and outputted. Thus, the drop-out of digital signals by the length of packet in the lump is prevented.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention packetizes a digital signal sequence on the transmitting side of a communication device and sends it out to a transmission path, and on the receiving side, the received packet is converted into an original Direckle signal. This relates to a packet communication system that returns to the queue.

[Conventional technology]

In a method of packetizing and transmitting a digital signal sequence, generally in a packet communication network, the time (bucket) is the time from when it is transmitted from the transmitting device until it is received by the receiving device.
(hereinafter referred to as packet transfer delay time) differs from packet to packet. When communicating audio, 2 images, etc. using such a packet communication network, the deviation (delay fluctuation) in the packet transfer delay time for each packet is absorbed and the packets are transferred to the original digital signal at the same time interval as on the sending side. Need to get back in line.

Therefore, delay fluctuations are absorbed by multiplying packets that arrive at the receiving side with delay fluctuations in a buffer memory on the receiving side for a certain period of time (T)M, and the packets are converted to the original digital signal at the same time interval as on the transmitting side. A back-to-line method is used. In addition, when transmitting and receiving analog signals, the receiving side digitizes the analog signals using an encoder, etc.
Then, on the receiving side, it is necessary to convert the restored digital signal sequence into an analog signal using a decoder or the like.

At this time, the accumulation time T is the maximum value of the packet transfer delay time (
The difference (Tdu
+ax-Tdmin), it is possible to return all received packets to digital signal sequences at the same time intervals as on the transmitting side.

[Problem that the invention seeks to solve]

However, in this case, the storage time T becomes long, and the time from when a packet is sent from the transmitting side to when the packet is returned to the original digital signal sequence at the receiving side becomes long, which is not desirable in terms of quality for real-time communication. It becomes something that does not exist. therefore,
Generally, the accumulation time T is set to an appropriate value according to the required quality, and even when the packets stored in the buffer memory are exhausted and it is time to return the packets to the original digital signal string, If the next packet has not arrived at the receiving device, a method is used in which the packet is discarded and the packet is not returned to the original digital signal sequence.

In conventional packet communication systems, digital signal sequences are sequentially packetized on the transmitting side and sent out to the transmission line in that order. Therefore, if bucket 1- is discarded for the above reason, a packet of digital signals corresponding to the packet length will be lost, resulting in a drawback that, for example, the quality of audio and images will be significantly degraded.

The quality deterioration of B becomes more and more noticeable as the packet length increases.

The purpose of the present invention is to solve the problem that, in a packet communication system in which a digital signal string is packetized and transferred, when a packet is discarded, a group of digital signals corresponding to the packet length are lost in the restored digital signal string. There is a party that provides a packet communication method that solves the problem.

[Means for solving problems]

In the packet communication method according to the invention of Party B, on the transmitting side, a digital signal string is divided into fixed packet lengths in a packetizing section, and the digital signals divided into packet lengths are divided into a plurality of packets. After the inter-packet block swapping unit in Party B swaps blocks between packets, the block order restoration information addition unit adds information necessary for the receiving side to restore the block order to each packet. This is sent to the transmission path as a packet with restoration information, and on the receiving side, the block order restoration unit extracts the packet from the restoration information material bucket 1- received from the transmission path and restores the block order. The block order is restored between packets based on the information, and the original digital signal sequence is restored.

[Effect]

In B's invention, on the transmitting side, blocks are exchanged between multiple buckets 1 before sending the packet to the transmission path, so even if a packet is discarded, the packet length is equivalent to the packet length of the discarded packet. Data will not be lost in bulk.

〔Example〕

FIG. 1 is an overall configuration diagram of a first embodiment of this invention,
This is an example in which an analog signal is digitized, and the digital signal is packetized and transferred. In this figure, 1 is an analog signal input terminal, 2 is an encoder, 3 is a packetization unit, 4 is a block division unit, 5 is an inter-packet block replacement unit, 6 is a block order restoration information addition unit, and 7 is a transmitter for transmission. 8 is a receiving transmission line, 9 is a block order restoring unit, 10 is a decoder, and 11 is an analog signal output terminal.

Next, using Figure 1, if the packet length is 256 bytes,
An example of operation will be described taking as an example a case where blocks are exchanged between two packets. First, the operation on the transmitting side will be described.

An analog game signal inputted from an analog signal input terminal 1 is converted into a digital signal by an encoder 2. The digital signal B is divided into packets of a fixed length by the packet/sorter 3 and sent to the block divider 4 in units of packets. The block dividing unit 4 divides each packet into two blocks of equal length.

FIG. 2 is a diagram showing the structure of a packet divided into blocks and the structure of the blocks.

The four packets sent from the port conversion unit 3 are converted into packet 1. Packet 2. Packet 3. The bucket is set to 4. FIG. 2 (,) shows a state in which each packet is divided into four. In FIG. 2(a), B1. .

B.2. B.3. Bl is a divided block of packet 1 (same for packets 2 to 4). There are two main ways to configure blocks: Note that each byte from the 1st byte to the 256th byte of packet 1 is expressed as bIp b2p b3p '''p bQ, 6.

The first method of configuring blocks is bl in B11.

b2. b , , , b64 to B12 b 6
6p b G6p b 07y”'p b12g to 1 B13 b 129ν b 130p b 13
1p ”'pb 192 to 1 B14b 193pb
IL141 b 195F ”'j b '25G
This is the method of assigning. The configuration of each block of the container 1 in this method is shown in FIG.

The same applies to packets 2 to 4.

The second method of configuring the blocks is Bolt) Iyb,
ri b 9p ′ ri b 253) B
,,l? -b 2. bIp t)10p"p
bZ64) to B13 b3. b7y bxlp
-p bzss to B14 b a, b s,
b□'AP ``'j b25G is allocated, and each block is composed of 4 pie 1-B and by 1-of the original packet.In this method, the composition of each block of packet 1 is Shown in Figure 2 (C).Bucket 1-2~
The same applies to 4.

The block dividing unit 4 divides each packet into four blocks using any of the methods described above and sends the divided blocks to the inter-packet block switching unit 5. The inter-soto block replacement unit 5 replaces each block among four packets.

FIG. 3 is a diagram showing an example of the results of block replacement by the inter-packet block replacement section 5. In FIG.

(R) In Figure 3, packet l-1', packet 2', packet 3'
, packet 4' is a packet after the blocks have been replaced, and the block order restoration information addition unit 6 sequentially starts from packet 1'.
sent to. The block order restoration information addition unit 6 selects information for restoring the block order on the receiving side (hereinafter referred to as restoration information 1) and the block configuration as shown in FIG. 2(b) or FIG. 2(C). Information indicating this (hereinafter referred to as restoration information 2) is added to each packet, and the packet is sent to the transmission line 7 as a packet with restoration information. As restoration information 1, each packet after block replacement is given a sequence number starting from 1, and the sequence numbers between 1.2 and 3.4 indicate that the blocks have been replaced.

Sequence numbers are 5, 6, 7, 8 and 9, 10, 11゜12
similarly indicates that blocks were exchanged between each of the four packets. The same applies to other packets.

Next, the operation on the receiving side will be explained.

The packet with restoration information received from the reception transmission path 8 is input to the block order restoration section 9 . The pro Lo+ pick order restoring unit 9 separates the packet from the restoring information 1 and the restoring information 2, and the order number of the restoring information 1 is 1゜2.3,4.
The order of blocks is restored between the four packets. Similarly, for buckets 1- with order numbers 5, 6, 7.8 and 9°10.11, 12, the block order is restored between each of the four packets. The same applies to other packets.

The packets with sequence numbers 1, 2, 3.degree.4 among the received packets are designated as bucket l-i', buckets 1-2', packet 3', and packet 4' in FIG. The result of restoring the order of blocks among these four packets is similar to that shown in FIG. 2(a). The block configuration of restoration information 2 is shown in Figure 2 (b
), the block order restoring unit 9
sends to the decoder 10 a packet with the block order restored, that is, a packet similar to that shown in FIG. 2(,). If the restoration information 2 indicates that the block order is as shown in FIG. b 2+ bL+ b
4r b 9.・.

After restoring in the order of b256, the packet is sent to the decoder 10. In the decoder 10 , the packets sent from the block order restoring section 9 are converted into analog signals and output to the analog signal output terminal 11 .

The above explanation has been given by taking as an example the case where blocks are exchanged between four packets, but it can be implemented in the same manner as in case B, where blocks are exchanged between two or more packets.

Further, although this embodiment shows a case where blocks are replaced after packetization, it is also possible to replace blocks at the same time as packetization.

FIG. 4 is a block diagram of the transmitting side of the second embodiment in which the processor is replaced at the same time as packetization. In the second embodiment, since the receiving side is the same as in the previous embodiment, only the transmitting side will be described.

In FIG. 4, 12 is an analog signal input terminal, 13 is an encoder, 14 is a switch, 15, 16, 17° 181.1
256-byte memory for each block, 19 a block order restoration information addition unit, and 2o a transmission line. , the operation on the transmitting side will be explained using FIG.

C11) The analog signal input from the analog signal input terminal 12 is converted into a digital signal by the encoder 13. This digital signal is divided into bytes in order from the beginning and is called a byte string, and in order: bIy b2t b3t
Let b4+b%. The byte string of the diretano signal output from the encoder 13 is sent to the switch 14. The switch 14 stores the byte string of the digital signal in the memory 15. Memory 16. Memory 17° Send to memory 18. There is method 2-) as a method of sending this.

The first method is to sequentially store 64 bytes of the digital signal in the memory 15. Memory 16. Memory 172.
Transfer to memory 18 1. When the transfer of 64 bytes to the memory 18 is completed, the byte string is inputted again in sequence from the memory 15 by 64 bytes, and this process is repeated.

Memory 15. Memory 16, memory 17. The contents of each memory when all of the memories 18 are full are the same as those shown in FIG. 31, and the contents of the memory 15 are as shown in FIG.
1' and the contents of memory 16 are packet 2' and the contents of memory 16.
7O] The contents are packet=i I・3', the contents of memory 18 are the same as the contents of packet 4', and bucket 1
-1' B1. ,B,□,B□3. B,, is shown in Figure 2 (b
) has a configuration ゛C. The same applies to packets 2' to 4'.

The second method of sending the byte string from the switch 14 to the memories 15 to 18 is to send the +- string one byte at a time to the memory 15, memory 16, memory 17, memory 18, Next, input the byte string one byte at a time to the memory 15 again in the above order, and repeat this process.When all of the memories 15 to 18 are full, the contents of each memory are sequentially shown in Figure 3. ) Same as packet 1' to packet 4', and 811. B,
□FBI3) 1'La has the configuration shown in Figure 2 (C). The same applies to packets 2' to 4'.

The block order restoration information adding unit 19 sequentially takes in a byte string of 256 bytes from the memory that has become full, adds restoration information 1 and restoration information 2, and sends it to the transmission transmission line 20 as a restoration information attached packet. The configuration and operation of the receiving side are the same as in the first embodiment.

Figure 5 shows that in the conventional packet communication system that does not exchange blocks between packets, 3 out of 4 packets
FIG. 4 is a diagram showing a situation when the th packet is discarded. FIG. 5(a) shows one bar sent on the transmitting side. FIG. 5(b) shows the digital signal string restored at the receiving side when the third packet is discarded. In FIG. 5(b), the broken line indicates the absence of data.

FIG. 6 is a diagram showing how data is lost when packets are discarded in the present invention. Figure 6 (a
) shows the result of dividing each packet into four blocks in the program dividing unit 4 of FIG. FIG. 6(b) shows the result of exchanging blocks between packets in the inter-packet block exchanging unit 5. FIG. 6(C) shows a case where the third packet is discarded on the receiving side. Figure 6(d) shows the received packets I-1'' and I-2''.
, as a result of restoring the block order from packet 4″,
Packet 1'', Packet 2'', Bucket I-3'', Bucket
1.4" is obtained. Each packet is missing the third proc.
The order of the data in the packet r-)l・1'' of FIG. 6(d) when using the block configuration of b) is shown. In this case, in the packet l-1'', the data of b 64 bytes are missing. This missing item is ``Tsuto 2'', Bucket)・3
The same applies to ``Pagasol I-4''.

FIG. 6(f) shows the result of restoring the data order of packet 1'' when using the block configuration of FIG. 2(c). In this case, b 3. b ? J b 11F ”
'Jb2filp b2'15 data is missing. Similar omissions apply to other Baga and Soto. According to this method, the loss occurs in byte units, which are even shorter than blocks, and is therefore highly effective in suppressing quality deterioration due to discarded packets.

As described above, according to the present invention, even if a packet is discarded, data corresponding to the packet length is not lost all at once, but the data corresponding to the packet length is shorter than the length of the packet as shown in FIG. 6(e). Since data is missing in units of programs or one byte out of every several bytes as shown in Figure 6(f), quality deterioration due to data loss is reduced, and a simple interpolation process can be applied to the missing data portion. By adding it, it is possible to perfect the quality.

The above explanation is based on an example in which blocks are exchanged between four packets, but the same effect can be obtained by exchanging blocks between two or more packets.
It is highly effective to exchange blocks between three or more packets. Further, although we will explain the case where the packet is discarded at the receiving terminal as an example, the same effect can be obtained when the packet is discarded within the network due to congestion within the network.

Note that in the packet communication system of the present invention, since data is exchanged between packets, a scrambling effect occurs.

〔Effect of the invention〕

As explained above, this invention exchanges blocks among multiple packets and then sends the packets to the transmission path]
7. After receiving multiple bano r-yl on the receiving side, the block order is restored by exchanging blocks between packets and performing the reverse operation, so the bucket) is not discarded at the receiving device. When multiplied, data corresponding to the packet length of the discarded bucket 1- is not lost all at once, but data is lost in units of blocks shorter than the packet length, or one byte of data is lost in several bytes. As a result, there is no significant quality deterioration even if one part is discarded, and the length of the missing part can be shortened by consolidating the data, so it is possible to easily fill in the missing part with interpolation. This process has advantages such as the ability to improve quality.

[Brief explanation of the drawing]

Fig. 1 is an overall configuration diagram of the first embodiment of the present invention, Fig. 2 is a diagram showing the structure of a packet divided into blocks and the structure of the blocks, and Fig. 3 is a diagram showing a packet after block replacement. Fig. 4 is a diagram showing the configuration of the transmitting side of the second embodiment of the invention of B, and Fig. 5 shows the packets on the transmitting side and reception when packets are discarded in the conventional packet communication system. FIG. 6 is a diagram of a packet on the transmitting side and a digital signal sequence restored on the receiving side when a packet is discarded in the present invention. In the figure, 1 is an analog signal input terminal, 2 is an encoder, 3 is a packetization unit, 4 is a block division unit, 5 is an inter-packet block replacement unit, 6 is a block order restoration information addition unit, and 7 is a transmission line for transmission. , 8 is a reception transmission path, 9 is a block order restoring unit, 10 is a decoder, 11 is an analog signal output terminal, 1
2 is an analog signal input terminal, 13 is an encoder, 14 is a switch, 15゜16, 17, 18 is a 256-byte memory,
19 is a block order restoration information adding section, and 20 is a transmission channel for transmission. Procedural amendment (iI rudder dated January 12, 1988 11 Director's Palace 1, case description Japanese Patent Application No. 61-214008 2, title of invention Packet communication system 3, person making the amendment Patent application related to the case) Address: 1-1-6 Uchisaiwai-cho, Chiyoda-ku, Tokyo Name (4
22) Nippon Telegraph and Telephone Corporation Representative Tsune Shinfuji 4, Agent Address: 5th floor, 6th floor, Hironobu Building, 31-16 Sakuragaoka-cho, Shibuya-ku, Tokyo 150 Japan Column for detailed explanation of the invention in the specification subject to amendment and drawings 6. Contents of amendment (1) "1 Receiving side" on page 3, line 7 of the specification is corrected to "Transmitting side." (2) Correctly correct 1-2 on page 7, line 5 to ``4 j.'' (3) Similarly, ``banoryl'' on page 14, lines 18-19.
・Correct "1", packet 2", packet 4" to [packet 1', packet) 1.2', packet 71.4'. (4) Similarly, page 16, line 3, 1-complete”, [-Improvement-
Correct it to 1. (5) Correct Figure 2 (al, (cl) and Figure 6 (d) as shown in the attached sheet. Above 2nd F 6th

Claims (1)

[Claims] In a packet communication system in which a digital signal stream is packetized on the transmitting side of a communication device and sent to a transmission path, and the received packet is returned to the original digital signal stream on the receiving side, a packetization unit and an inter-packet block switching unit are used. , using a communication device equipped with a block order restoration information adding section and a block order restoration section, on the transmitting side, the digital signal string is divided into fixed packet lengths in the packetization section, and the digital signals divided into packet lengths are divided into a plurality of packet lengths. The packet is divided into blocks and sent to the inter-packet block swapping unit, and after the inter-packet block swapping unit swaps blocks between packets, the receiving side restores the order of the blocks in the block order restoration information addition unit. The necessary information is added to each packet and sent as a packet with restoration information to the transmission path. On the receiving side, a block order restoration unit extracts the packet from the packet with restoration information received from the transmission path and restores the packet. A packet communication method characterized by restoring the order of blocks between packets based on information for restoring the order of blocks and returning the original digital signal sequence.