JP2001313673A - Packet transmitting method, packet repeater and data terminal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a packet transmitting method by which a header compression packet to be restoration impossible due to a partial loss in packet is reduced even when a header in a non-compression packet to be transmitted is compressed. SOLUTION: At the side of transmission, a non-compression header comprised in each non-compression packet is held as a transmission side reference header and the non-compression packet to be transmitted is converted into the header compression packet through the use of >=2 transmission side reference headers. In the meantime, at the side of reception, the non-compression header obtained from the packet which is received from the transmission side is held as a reception side reference header and also the received header compression packet is restored to be the non-compression packet through the use of the reception side reference header corresponding to at least one transmission side reference header among the >=2 transmission side reference headers which are used for conversion into the header compression packet at the transmission side.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a packet transmission method, and a packet relay device and a data terminal used for the method.

[0002]

2. Description of the Related Art For example, a packet exchanged between a plurality of data terminals usually includes, in addition to data (payload) to be transmitted, some header having information for specifying a transmission destination. However, if the ratio of the data amount of the header to one packet (overhead) is large, there is a problem that the communication efficiency is reduced. In order to solve such a problem, conventionally, various techniques for compressing a header included in each packet have been proposed. For example, the IETF (Internet
RFC (Request Fo) issued by the Engineering Task Force
r Comment) 2508 includes Internet Protocol (IP) header, User Datagram Protocol (UDP) header, and RTP
(Realtime Transport Protocol) An RTP compression protocol for compressing a header is disclosed. Below, this R
An outline of a header compression procedure according to the TP compression protocol will be described.

[0003] This RTP header compression protocol is used, for example, when a packet is exchanged between two data terminals via a network, the packet is included in the network and relays the packet exchanged between the data terminals. It is applied to communication performed between two node devices (for example, a router). Hereinafter, as shown in FIG. 11, one of these node devices is called a transmitting node, and the other is called a receiving node.

[0004] First, a transmitting node receives an uncompressed packet (hereinafter, referred to as an "uncompressed packet") a from a transmitting data terminal (or a node device preceding the transmitting node). , Converts the uncompressed packet into a full header packet, and transmits the full header packet to the receiving node (in FIG. 11). The full header packet refers to a packet that can restore an uncompressed packet only from the content of the full header packet without referring to another packet. The receiving node that has received the full header packet a restores the full header packet a into an uncompressed packet a and transmits it to the next node device (or the receiving data terminal).

Further, the transmitting node compresses a header included in the next received uncompressed packet b into a compressed header based on the content of the previously received uncompressed packet a, and compresses the header containing the compressed header. The packet b is transmitted to the receiving node (in FIG. 11). On the other hand, the receiving node receives the header compressed packet b, but the header compressed packet cannot be restored to an uncompressed packet only from the content of the header compressed packet. Therefore,
The receiving node restores the compressed header included in the header compressed packet b into an uncompressed header based on the contents of the previously received and restored uncompressed packet a, and converts the uncompressed packet b including the uncompressed header into an uncompressed packet b. It is transmitted to another node device (or a data terminal on the receiving side). Further, the transmitting node transmits a header compressed packet c obtained by compressing the uncompressed packet c to the receiving node in the same manner as described above, while the receiving node transmits a compressed header in the received header compressed packet c. Is restored to an uncompressed header based on the content of the uncompressed packet b previously received and restored, and an uncompressed packet c including the uncompressed header is transmitted. As described above, in the related art, the uncompressed packet received by the transmitting node is compressed based on the content of one previously received packet, while the header compressed packet received by the receiving node is compressed. Has been decompressed into an uncompressed packet based on the contents of one previously received packet.

[0006]

However, in the above-mentioned prior art, the following problems may occur. Here, FIG. 11 illustrates a case where the header compressed packet d transmitted from the transmitting node is lost for some reason before reaching the receiving node (FIG. 1).
In 1). In this case, the receiving node does not receive the header compression packet d but receives the next header compression packet e.
Will be received. However, since the header compressed packet e can be restored only based on the contents of the uncompressed packet d obtained by restoring the header compressed packet d, as shown in FIG.
Cannot be restored if lost. Therefore,
The header compressed packet e must be discarded even though it is normally received. Furthermore, as a result of the header compressed packet e not being restored in this way, the next received header compressed packet f cannot be restored, and is therefore discarded. Eventually, due to the loss of one packet d, the header compressed packets e to h received until the next full header packet i is received are discarded (in FIG. 11).

As described above, in the prior art, the header compression of an uncompressed packet or the restoration from a header-compressed packet to an uncompressed packet is performed by referring to only one packet. Therefore, when one referenced packet is lost for some reason, a number of subsequent packets are discarded in a chain, causing a problem that communication efficiency is significantly reduced. When a wireless section is included between the transmitting node and the receiving node, the above problem is particularly prominent because packets are easily lost in the section.

[0008] The present invention has been made in view of the above-described circumstances, and even when a header in an uncompressed packet to be transmitted is compressed, the header is restored due to loss of a part of the packet. It is an object of the present invention to provide a packet transmission method, a packet relay device, and a data terminal capable of reducing uncompressed header compressed packets.

[0009]

In order to solve the above-mentioned problem, the present invention provides a network including a transmitting side and a receiving side, wherein the transmitting side transmits an uncompressed packet to be transmitted to a full header having a full header. A packet transmission method for converting a full header packet or a header compressed packet received from the transmitting side into an uncompressed packet while converting the packet into a packet or a header compressed packet having a compressed header and transmitting the packet to the receiving side. The transmitting side holds an uncompressed header included in each of a plurality of uncompressed packets as a transmitting side reference header, and uses two or more transmitting side reference headers to transmit an uncompressed packet to be transmitted. While converting the data into a header compressed packet, the receiving side transmits the full header packet or header compressed packet received from the transmitting side. The uncompressed header in the uncompressed header obtained by restoring the packet is retained as a receiving-side reference header, and at least one of the two or more transmitting-side reference headers used for conversion to a header-compressed packet on the transmitting side is used. Using the reception-side reference header corresponding to the transmission-side reference header, the header compression packet is restored to an uncompressed packet.

[0010] The present invention can also be implemented as a packet relay device that operates as a transmitting side or a receiving side in the above packet transmission method. That is, the packet relay device on the transmission side in the above packet transmission method,
A packet relay device interposed between a plurality of data terminals and relaying uncompressed packets exchanged between the data terminals, comprising: a receiving unit for receiving an uncompressed packet; and a non-compressed packet included in the received uncompressed packet. Holding means for holding a compressed header as a transmission-side reference header, and conversion for converting a part of the received uncompressed packet into a header compression packet having a compression header using the two or more transmission-side reference headers Means for transmitting the header compressed packet. On the other hand, the packet relay device on the receiving side in the packet transmission method is a packet relay device that relays a plurality of uncompressed packets exchanged between a plurality of data terminals, and a part of the plurality of uncompressed packets. A packet relay device capable of communicating with a transmitting side that converts an uncompressed packet into a header compressed packet using an uncompressed header included in two or more other uncompressed packets and transmits the packet, and transmits the packet from the transmitting side. Receiving means for receiving the received packet, holding means for holding an uncompressed header obtained from the received packet as a receiving-side reference header, and when the header compressed packet is received from the transmitting side, the transmitting side At least one of the two or more uncompressed headers used for the conversion into the header compressed packet corresponds to the uncompressed header. Using the recipient reference header, characterized by comprising a restoring means for restoring the header-compressed packet to the uncompressed packet, and transmitting means for transmitting the restored uncompressed packets.

Further, the present invention can be embodied as a data terminal capable of realizing the above-described packet transmission method.
That is, the data terminal on the transmitting side in the packet transmission method is a data terminal that can exchange packets with another data terminal, and is included in a plurality of uncompressed packets to be transmitted to the other data terminal. Holding means for holding an uncompressed header to be transmitted as a transmission-side reference header, and using two or more transmission-side reference headers to convert a part of the plurality of uncompressed packets to be transmitted into a header compressed packet having a compressed header. And a transmitting means for transmitting the header compressed packet. On the other hand, the data terminal that receives data in the above packet transmission method is a data terminal that can exchange packets with another data terminal via a packet relay device, and a part of a plurality of uncompressed packets. A data terminal communicable with a packet relay device that converts an uncompressed packet of (1) into a header compressed packet using an uncompressed header included in two or more other uncompressed packets and transmits the packet. Receiving means for receiving a packet transmitted from, holding means for holding an uncompressed header obtained from the received packet as a receiving side reference header, and when the header compressed packet is received from the packet relay device, Of the two or more uncompressed headers used in the conversion to the header compressed packet in the packet relay device, at least Even using the recipient reference header corresponding to one of the non-compressed header, characterized by comprising a restoring means for restoring the header-compressed packet to the uncompressed packet.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. Such an embodiment shows one aspect of the present invention, and does not limit the present invention, and can be arbitrarily changed within the scope of the present invention.

A: First Embodiment A-1: Configuration of First Embodiment FIG. 1 is a block diagram schematically illustrating the configuration of a communication system to which a packet transmission method according to the present invention can be applied. As shown in the figure, in this communication system,
For example, a transmitting data terminal 1 such as a personal computer
And the receiving data terminal 2 can exchange packets via a network 3 including the Internet or the like. In the following, a description will be given of a case where the transmitting data terminal 1 transmits a packet to the receiving data terminal 2 as an example.

[0014] As shown in FIG.
The transmitting node 3a and the receiving node 3b are included. The transmitting node 3a and the receiving node 3b
It has a role of relaying a packet exchanged between the transmission side data terminal 1 and the reception side data terminal 2, and is, for example, a router (path control device). Note that, in FIG. 2, in order to prevent the drawing from being complicated,
Although only one receiving node 3b is shown in the figure, it goes without saying that the network 3 actually includes a larger number of nodes.

In the configuration shown in FIGS. 1 and 2, the transmitting data terminal 1 sends a packet to the receiving data terminal 2 as a destination (a packet that has not been subjected to any compression.
“Uncompressed packets”) are sequentially transmitted to the network 3. On the other hand, the transmitting node 3a receives the uncompressed packet transmitted from the transmitting data terminal 1 from the transmitting data terminal 1 or via another node device, and transmits the uncompressed packet including the full header. The packet is converted into a header packet or a header compressed packet including a compressed header and transmitted to the receiving node 3b. on the other hand,
The receiving node 3b restores the full header packet or header compressed packet received from the transmitting node 3a into an uncompressed packet, and transmits the uncompressed packet to the next node device or the receiving data terminal 2. Note that the transmitting node 3a
A packet transmitted from the transmitting data terminal 1 may be directly received from the transmitting data terminal 1 or a packet may be received from the transmitting data terminal 1 via another node device. For convenience of explanation, it is assumed that a packet is directly received from the transmitting data terminal 1. Similarly, the receiving node 3b may transmit the uncompressed packet directly to the receiving data terminal 2 or may transmit the uncompressed packet to another node device. It is assumed that the packet is transmitted directly.

As shown in FIG. 2, the transmitting node 3a includes a receiving unit 31a, a transmitting unit 32a, a control unit 33a, and a storage unit 34a. The receiving unit 31a
This is a means for receiving an uncompressed packet from the transmitting data terminal 1 and outputting it to the control unit 33a. Also, the transmitting unit 3
2a is means for transmitting data output from the control unit 33a to the receiving node 3b via a communication line. The control unit 33a is configured to include a CPU and the like, and is a unit that controls each unit in the transmission-side node 3a by executing a main program stored in the storage unit 34a. The storage unit 34a stores, in addition to the main program, a header compression program for converting an uncompressed packet received from the transmitting data terminal 1 into a full header packet or a header compressed packet.

The receiving node 3b has the same configuration as the transmitting node 3a, and includes a receiving unit 31b, a transmitting unit 32b, a control unit 33b, and a storage unit 34b. However, the control unit 33b in the receiving node 3b stores the storage unit 34
By executing the header decompression program stored in b, it plays a role of decompressing a full header packet or header compressed packet received from the receiving node 3b into an uncompressed packet.

Next, the structure of an uncompressed packet, a full header packet or a header compressed packet will be described with reference to FIGS. First, FIG. 3 is a diagram schematically illustrating the configuration of an uncompressed packet. As shown in the drawing, an uncompressed packet in the present embodiment is configured to include an RTP payload and an uncompressed header. The RTP payload includes data to be transmitted, for example, image data and audio data. On the other hand, uncompressed headers are RTP, U
It consists of an RTP header, a UDP header and an IP header added to the RTP payload according to DP and IP.

The transmitting side node 3a is connected to the transmitting side data terminal 1
Uncompressed packets received from
Alternatively, the packet is converted into one of the header compressed packets shown in FIG. 4A and transmitted to the receiving node 3b. Specifically, it is as follows.

First, the transmitting node 3a transmits at least the first uncompressed packet among the uncompressed packets to be transmitted to the receiving node 3b as a full header packet. In the full header packet, the length included in the IP header and the UDP header of the uncompressed packet shown in FIG.
T_ID ”and a link sequence number (hereinafter“ link_seq ”). Here, CONTEXT_ID is information for identifying the full header packet, and link_seq is a transmission order of the full header packet or the header compression packet transmitted from the transmitting node 3a to the receiving node 3b. It is a serial number to be added. Upon receiving such a full header packet, the receiving node 3b receives
The CONTEXT_ID and the link_seq are extracted, and the uncompressed header is generated by restoring the value of length to be included in the IP header and the UDP header according to the lower layer. That is, the full header packet is the full header included in the full header packet (that is, the length is CONTEXT_
It is a packet that can restore an uncompressed packet including an uncompressed header based only on the contents of the ID and the uncompressed header replaced with link_seq and the like.

On the other hand, a header compressed packet is a packet from which an uncompressed packet cannot be restored only from its own contents. Hereinafter, this header compression packet will be described.

First, the data of the hatched portion in the uncompressed header shown in FIG. 3, that is, the version number (V) in the IP header and the payload type (PT; RTP payload of the RTP payload) in the RTP header (E.g., including image data or audio data, information on an encoding method, etc.) is expected to be constant among uncompressed packets transmitted from the transmitting data terminal 1. Data (hereinafter, these data are collectively referred to as “constant value fields”). Since the data of this constant value field has already been notified to the receiving node 3b by the previously transmitted full header packet, the constant value field is commonly used for the subsequent packets unless the contents are changed. be able to. That is, there is no need to duplicately notify the receiving node 3b of all packets. From this point of view, FIG.
As shown in (1), the header compression packet does not include the data in the fixed value field. The receiving node 3b holds at least the uncompressed header restored from the full header packet received first, and uses the constant value field in the uncompressed header to convert the compressed header in the header compressed packet received thereafter. It decompresses the uncompressed header.

However, the data in the constant value field does not always have a constant value in all packets, and may change. When such a change is made, the transmitting node 3a transmits the full header packet to the receiving node 3b without compressing the header in the packet.

Next, among the uncompressed headers shown in FIG. 3, the data of the sparsely hatched parts, that is, for example, the length in the IP header, the length in the UDP header, etc.
This is information that can be obtained based on information added according to the lower layer (that is, data of a part other than the uncompressed header) and the like. That is, even when the receiving node 3b receives a packet that does not include such information, the receiving node 3b can acquire the information based on information of another layer. Therefore, as shown in FIG. 4A, the above information is not included in the above-described header compression packet.

Next, the data of the unhatched portion in the uncompressed header shown in FIG. 3, that is, the ID in the IP header and the sequence number (S
N; Sequence Number) and time stamp (TS; Time)
Stamp) is data whose value can continuously change (by a predetermined difference value) between two uncompressed packets to be transmitted, and the difference value (change amount) between each uncompressed packet.
Is data that is expected to be substantially constant (hereinafter, these data are collectively referred to as “variable fields”).
Here, assuming that there is no change in the difference value, the receiving node 3b holds the first uncompressed packet and the difference value of the variable field between the uncompressed packet and the next uncompressed packet. If so, the variable field of the subsequent packet can be restored by sequentially adding the held difference value to the value of the variable field in the first uncompressed packet. Therefore, FIG.
As shown in (a), the value of the variable field itself is not included in the compressed header in principle.

However, the difference value between the uncompressed packets in the variable field is not always constant among all the uncompressed packets, and may change. In such a case, if the changed difference value is notified to the receiving node 3b, the receiving node 3b calculates the difference between the content of the uncompressed header restored immediately before and the newly notified difference value. Thereafter, the content of the compressed header in the received header compressed packet can be restored. Hereinafter, a procedure for notifying the receiving node 3b of the difference value in the present embodiment will be described.

In the present embodiment, the transmitting node 3a
A plurality of uncompressed headers (sender-side reference headers) of the uncompressed packets received from the transmitting data terminal 1 and already transmitted to the receiving node 3b as full header packets or header compressed packets are stored. I have. Then, a difference value of the variable field is obtained between the uncompressed header of the uncompressed packet to be transmitted as the header compressed packet and each of the held two or more uncompressed headers, and the combination of these two or more difference values is determined. Receiving node 3b
Is to be notified. In other words, by referring to the contents of two or more uncompressed headers, the uncompressed header of the uncompressed packet to be transmitted to the receiving node 3b is converted to a compressed header.
It should be noted that two or more uncompressed headers that are referred to when converting an uncompressed packet into a header compressed packet can be arbitrarily selected, but for convenience of description, a receiving node will be described immediately before an uncompressed packet to be converted. 2b, the uncompressed header of the uncompressed packet transmitted to the receiving node 3b and the uncompressed header of the uncompressed packet transmitted to the receiving node 3b four times before the uncompressed packet to be converted. Shall be referred to. Specifically, it is as follows. Here, FIG. 5 shows that the transmitting node 3a sequentially receives the uncompressed packets a to p from the transmitting data terminal 1 and a part of these uncompressed packets as a header compressed packet to the receiving node 3b. It is assumed that transmission is performed. Now, assuming that the transmitting node 3a receives the uncompressed packet f from the transmitting data terminal 1 and transmits the uncompressed packet f to the receiving node 3b as a header compressed packet, the receiving node 3b is notified of this transmission. The combination of the difference values is the difference value Δ of the variable field between the uncompressed packet f and the immediately preceding uncompressed packet e.
X, and the combination of the difference value ΔY of the variable field between the uncompressed packet f and the uncompressed packet b immediately before the uncompressed packet f.

On the other hand, when the receiving node 3b is notified of the combination of the difference values by receiving the header compression packet,
Among the uncompressed headers (reception-side reference headers) of the uncompressed packets that have already been received and restored, the uncompressed header used for compression of the header compressed packet, that is, the uncompressed header received and restored immediately before the header compressed packet One of the difference values reported this time is added to one of the uncompressed header of the uncompressed packet and the uncompressed header of the uncompressed packet received and restored four times before the header compressed packet. Then, the value of the variable field in the uncompressed header to be restored from the header compression packet received this time is restored. Specifically, referring to FIG. 5 again,
When receiving the header compressed packet f, the receiving node 3b receives the uncompressed packet e immediately before the header compressed packet f or the uncompressed packet e received and restored four times before the header compressed packet f. One of the compressed packets b is selected. Here, when the uncompressed header e is selected, ΔX of the difference value notified this time is selected, and when the uncompressed header b is selected, the difference value ΔY is set in the variable field in the uncompressed header. , The value of the variable field of the uncompressed header to be restored from the header compression packet received this time is restored. Here, in the above example, the full header packet or the header compressed packet before the decompression of the uncompressed packet b is lost between the transmitting node 3a and the receiving node 3b for some reason, and actually, It is conceivable that the uncompressed header b cannot be used for restoring the header compressed packet f. However, even in such a case, if the uncompressed packet e has been normally received and restored, the variable field in the uncompressed header to be restored from the header compressed packet f can be restored. That is, in the present embodiment, since the contents of the plurality of uncompressed headers are referred to and the uncompressed packets are converted into the header compressed packets, at least one of the reference source uncompressed headers is converted. If held in the receiving node, the contents of the header compressed packet can be restored. As a result,
It is possible to solve the problem of the related art in which a large number of packets are discarded in a chain due to the loss of a certain packet.

By the way, two difference values ΔX are added to the compressed header.
And ΔY, the data amount of the header may increase. From this point of view, the header compression packet in the present embodiment does not include the difference values ΔX and ΔY themselves, as shown by the broken line in FIG.
A difference value code corresponding to the difference values ΔX and ΔY is included. Hereinafter, the difference value code will be described in detail.

The amount of change in the value of the variable field is
Generally, it can be generally predicted according to the characteristics of data to be transmitted (audio data or image data). For example, a time stamp TS in an uncompressed packet including audio data having a sampling frequency of 80 kHz and a frame length of 20 ms
(Variable field) is “160 (= 8000 × 0.0)
2) It is expected to change every time. That is, the minimum unit amount of change in the value of the time stamp TS is “160”. Also, for example, the sampling frequency 90 kHz, the time stamp TS of image data of 30 frames per second is expected to change every “3000 (= 90000 × 1/30)”. That is, the minimum unit amount of change in the value of the time stamp TS is “3000”.

In view of such circumstances, the difference values ΔX and ΔX of the value of the time stamp TS between each uncompressed packet are considered.
Y is expected to be a multiple of the above minimum unit quantity. The same can be said for the values of the other variable fields (that is, the ID in the IP header and the sequence number SN in the RTP header). Therefore, in this case, if the minimum unit amount is notified to the receiving node in advance, for subsequent uncompressed packets, the difference values ΔX and ΔY to be notified each correspond to the multiple of the minimum unit amount. If the information (hereinafter, referred to as “magnification information D1” and “magnification information D2”) indicating this is notified to the receiving node 3b, the receiving node 3b will be notified of the magnification notified to the minimum unit amount. ΔX and Δ
Y can be obtained. That is, the minimum unit amount of the change in the value of the variable field is referred to as a “common factor”.
In other words, there is a relationship of (difference value ΔX) = (magnification information D1) × (common factor) (difference value ΔY) = (magnification information D2) × (common factor) Then, the magnification information D1 and D2 have a smaller data amount than the difference values ΔX and ΔY.

In view of such circumstances, the difference value code of the header compression packet in the present embodiment is shown in FIG.
As shown in the figure, the magnification information D1 and D2 are included in place of the difference values ΔX and ΔY. In this case, “0” is set to the flag F in the difference value code.

On the other hand, two difference values ΔX and ΔY
May not be a multiple of the common factor that has already been notified to the receiving node 3b. In such a case,
If there is a new common factor (in other words, the greatest common divisor of the two difference values) corresponding to the difference values ΔX and ΔY obtained this time, the flag F is set to “1” and Magnification information D1 and D2 corresponding to the new common factor and a difference value code including the new common factor (region A surrounded by a broken line in the figure) are inserted into the header compression packet and transmitted to the receiving node. It will be. Furthermore, a case where the two difference values ΔX and ΔY do not have a common factor may be considered. In this case, "1" is set to all of the flags F, D1, and D2, and the value of the variable field included in the uncompressed packet to be converted to the header compressed packet is directly used as the difference value code. (Region B surrounded by a broken line in the figure). The above is the details of the difference value code in the present embodiment.

As described above, in the present embodiment, FIG.
As shown in (a), the difference code corresponding to the combination of the difference values ΔX and ΔY obtained for the uncompressed packet to be transmitted (or including the value of the variable field itself) is converted into each variable field (ID, SN and TS) are included in the compressed header in the header compressed packet. Here, in the present embodiment, the difference value code obtained based on the content of the uncompressed packet to be transmitted is the difference code obtained for each of the two uncompressed headers referred to when the uncompressed packet is compressed. When the value code is the same as the value code, the difference value code is not included in the compressed header obtained by compressing the uncompressed header of the uncompressed packet to be transmitted. For example, in FIG. 5, assuming that the current compression target is an uncompressed packet k, the difference value code obtained for the uncompressed packet k is referred to as the uncompressed packet g referred to when compressing the uncompressed packet. Is the same as the difference value code of the uncompressed packet j,
Is not contained in the header compression packet k obtained by compressing If any one of these uncompressed packets g and j is held together with the difference value code in the receiving node 3b, this difference value code can be used as the difference value code of the uncompressed packet k. is there. In addition, as is clear from the above,
“The difference value codes are the same” means that both the difference values ΔX and ΔY are the same.

In order to indicate which variable field the difference value code is included in, the compressed header contains a flag S, as shown in FIG.
T and I are included. Here, the flags S and T are
Sequence number SN and time stamp T in RTP header
The flag corresponds to S, and the flag I corresponds to the ID in the IP header. Specifically, for example, as indicated by a broken line in FIG. 4A, the difference value code (IP_ID) obtained according to the value of the ID in the IP header is notified to the receiving node 3b immediately before. If it is different from the difference value code, the new difference value code (IP_ID) is added, and a header compression packet including a compression header in which the flag I is set to “1” is transmitted to the reception-side node 3b. Because Similarly, the sequence number S in the RTP header
If the difference value code (RTP_SN) obtained according to the value of N is different from the difference value code notified to the receiving node 3b immediately before, the new difference value code (RTP_SN) is added, and , A header compressed packet including a compressed header with the flag S set to “1” is transmitted to the receiving node 3b.

A-2: Operation of First Embodiment Next, a specific example of the operation of the present embodiment will be described. Prior to the operation described below, a common factor determined in advance according to the characteristics of data to be transmitted / received (such as audio data or image data) is transmitted from the transmitting data terminal 3a to the receiving data terminal 3b. It shall be notified. In the following, description will be made separately for the operation of the transmitting node 3a and the operation of the receiving node 3b.

A-2-1: Operation of transmitting node 3a First, the control unit 33a in the transmitting node 3a receives the first uncompressed packet a from the transmitting data terminal 1 via the receiving unit 31a. The uncompressed packet a is converted into a full header packet a and transmitted to the receiving node 3b.
Specifically, the length included in the uncompressed header of the uncompressed packet a is replaced with CONTEXT_ID, link_seq, and the like, and the full header packet a is generated and transmitted. Thereafter, similarly, when the second to fourth uncompressed packets b to d are received, they are converted into full header packets and transmitted to the receiving node 3b (see FIG. 8). The reason why the first to fourth uncompressed packets are transmitted as full header packets is as follows. That is, as described above, in the present embodiment, the uncompressed packet to be transmitted is compressed into a header compressed packet with reference to the immediately preceding uncompressed packet and the immediately preceding uncompressed packet. This is because the first to fourth uncompressed packets among the uncompressed packets transmitted from the transmission-side data terminal 1 do not have the fourth uncompressed packet to be referred to.

Thereafter, the transmitting node 3a sequentially converts the uncompressed packet received from the transmitting data terminal 1 into a header compressed packet and transmits it to the receiving node 3b. The details are as follows.

FIG. 6 is a block diagram showing a functional configuration for header compression in the transmitting node 3a. As shown in the figure, the transmission-side node 3a includes a division unit 301, a shift register 302, subtraction units 303 and 304, an encoding unit 305, a compressed header generation unit 306, and a header compressed packet generation unit 307. . FIG.
The dividing unit 301, the subtracting units 303 and 304, the encoding unit 305, the compressed header generating unit 306, and the header compressed packet generating unit 307 shown in FIG. 2 are arranged such that the control unit 33a in the transmitting node 3a shown in FIG. This is realized by executing the stored header compression program.

First, when an uncompressed packet e to be transmitted to the receiving node 3b as a header compressed packet is received from the transmitting data terminal 1, the dividing unit 301 divides the uncompressed packet e into the uncompressed header e and the RTP payload e. And output. Specifically, the uncompressed header e is output to the shift register 302, the subtracters 303 and 304, and the compressed header generator 306, while the RTP payload e is output to the header compressed packet generator 307.

The shift register 302 stores the uncompressed header e supplied from the division unit 301. In particular,
As shown in FIG. 7A, the uncompressed headers of the uncompressed packets a to d already transmitted as full header packets are sequentially stored in the shift register 302 each time they are received. When the uncompressed header e of the uncompressed packet e to be transmitted as the header compressed packet is supplied, the shift register 302 shifts the already stored uncompressed headers a to d one by one, and The uncompressed header e supplied this time is stored in the area vacated by the shift (the 0th stage represented by “0” in FIG. 6). As a result, the contents stored in the shift register 302 are as shown in FIG. The uncompressed header stored in the shift register 302 in this way corresponds to a “transmission-side reference header” in the claims.

Next, when the writing of the uncompressed header e is performed, the subtraction unit 303
The value of the variable field of the reference header d stored in the row (represented by “1” in FIG. 6) is read. Then, as shown in FIG. 7B, the subtraction unit 303 calculates a difference value ΔX between the value of the read variable field and the value of the variable field in the uncompressed header e received this time, and Output to 305. Similarly, the subtraction unit 30
4 reads the value of the variable field of the uncompressed header a stored in the fourth row (represented by “4” in FIG. 6), and reads this value and the value of the uncompressed header e received this time. , And outputs the difference value ΔY to the encoding unit 305.

On the other hand, encoding section 305 generates information to be included in the above-described difference value code from difference values ΔX and ΔY notified from subtraction sections 303 and 304. Specifically, when the difference values ΔX and ΔY are multiples of the common factor that has already been notified to the receiving node 3b,
The respective magnifications are output to the compression header generator 306 as magnification information D1 and D2. Further, the difference values ΔX and ΔY
However, if the common factor is not a multiple of the common factor, the greatest common divisor of these values is used as the common factor, and this common factor is output to the compressed header generating unit 306 together with the magnification information D1 and D2 representing the respective magnifications. . Furthermore, the difference values ΔX and Δ
If Y does not have the greatest common divisor, it notifies the compressed header generator 306 that the value of the variable field should be included as it is.

Next, the compressed header generator 306 generates a compressed header according to the information notified from the encoder 305. Specifically, it is as follows. (1) When only magnification information D1 and D2 are supplied,
The compressed header generation unit 306 generates a difference value code (portion indicated by a solid line in FIG. 4B) including the notified magnification information D1 and D2. (2) When the magnification information D1 and D2 and a new common factor are supplied, the compressed header generation unit 306 includes not only the notified magnification information D1 and D2 but also the common factor (the broken line in FIG. 5B). Part) and a difference value code in which the flag F is set to “1”. (3) When the encoding unit 305 notifies that the value of the variable field should be included as it is, the compressed header generation unit 3
06 extracts each value of the variable field from the uncompressed header e supplied from the dividing unit 301, includes the value of the variable field, and sets “1” to all of the flags F and D1 and D2. Generate a difference value code.

Further, the compressed header generation unit 306 determines whether or not to include the obtained difference value code in the compressed header. Here, the uncompressed header e is the first packet transmitted as a header compressed packet, and has not been notified of any difference value code to the receiving side node 3b, and thus is obtained as described above. A compressed header including all of the difference value codes is generated, and the difference value codes are stored in the storage unit 34a.

Then, the compressed header generating section 306 sets the above-mentioned difference value code and a flag (FIG. 4A) in which "1" is set in accordance with which variable field the difference value code relates to. Flag S, T or I)
And a compressed header e including CONTEXT_ID and link_seq
Is generated and output to the header compression packet generation unit 307. On the other hand, the header compressed packet generator 307 generates a header compressed packet e by adding the compressed header e supplied from the compressed header generator 306 to the RTP payload e supplied from the divider 301. To the receiving node 3b.

Thereafter, the same operation is performed when an uncompressed packet f to be transmitted as a header compressed packet is received. That is, first, the uncompressed header f of the uncompressed packet f is stored in the shift register 302. Here, the uncompressed header a is no longer referred to when compressing packets after the uncompressed packet f. Therefore, as shown in FIG. 7C, the uncompressed header a is discarded in the shift for storing the uncompressed header f.

Further, for the uncompressed header f,
The difference values ΔX and ΔY for each variable field are calculated between the immediately preceding uncompressed header e and the immediately preceding uncompressed header b (see FIG. 7C), and the difference values ΔX and ΔY are calculated. Is generated for each variable field, and the header compression packet f including this difference value code is transmitted to the receiving node 3b.

The same operation is performed for the subsequent uncompressed packets. As a result, as shown in FIG. 8, the uncompressed packets a to d are transmitted to the receiving node 3b as full header packets, while Uncompressed packet e ~
k is transmitted to the receiving node 3b as a header compression packet. Here, for the packets after the uncompressed packet i, each of the two uncompressed headers referred to when compressing the uncompressed packet is transmitted to the receiving node 3b as a header compressed packet including a difference value code. (See FIG. 5). For example, as shown in FIG. 5, the uncompressed packet i is converted into a header compressed packet by referring to the uncompressed headers h and e, and these uncompressed headers h and e include the difference value code. It is converted to a compressed header, and has already been transmitted to the receiving node 3b as a header compressed packet including this. Therefore, the uncompressed packets after the uncompressed packet i,
The obtained difference value code is the same as both of the difference value codes of the other two uncompressed headers (uncompressed headers h and e in the above example) used for compressing the uncompressed packet. For uncompressed packets,
The difference value code is transmitted to the receiving side node 3b as a header compressed packet from which the difference value code is omitted. When the difference value code can be omitted for only some of the variable values among the difference value codes for each variable field (ID, SN, and TS), the difference value codes for other variable fields are included in the compressed header. At the same time, "1" is set in the corresponding flag (see FIG. 4A).
On the other hand, upon receiving the header compressed packet from which the difference value code has been omitted, the receiving node 3b uses the difference value code held together with one of the two uncompressed headers that are the reference sources of the header compressed packet. hand,
The header compressed packet received this time is restored to an uncompressed packet.

A-2-2: Operation of Receiving Node 3b Next, a specific example of the operation of the receiving node 3b receiving the full header packet or the header compressed packet transmitted from the transmitting node 3a will be described.

First, the control unit 33b in the receiving node 3b
Receives the full header packet a from the transmitting node 3a via the receiving unit 31b, converts the full header packet a into an uncompressed packet, and transmits the packet to the receiving data terminal 2. As described above, the non-compressed packets b to d are also transmitted as full header packets from the transmitting node 3a, but the receiving node 3b also similarly transmits the uncompressed packets b to The data is converted to d and transmitted to the receiving data terminal 2.

Thereafter, the receiving node 3b receives the header compressed packets sequentially transmitted from the transmitting node 3a,
Each header compressed packet is restored to an uncompressed packet and transmitted to the receiving data terminal 2. Hereinafter, the restoration operation in the receiving node 3b will be described with reference to FIG. FIG.
FIG. 14 is a block diagram illustrating a functional configuration related to a restoration operation of a receiving node 3b. As shown in FIG.
Is divided into a division unit 310, a loss detection unit 311, a decoding unit 312,
The configuration includes a header selection unit 313, an uncompressed header generation unit 314, an uncompressed packet generation unit 315, and a shift register 316. Note that the control unit 33b in the reception-side node 3b illustrated in FIG.
This is realized by executing the header restoration program stored in b.

First, upon receiving the header compressed packet e from the transmitting node 3a, the dividing section 310 divides the header compressed packet e into a compressed header e and an RTP payload e, and outputs them. Specifically, the compressed header e is sent to the loss detection unit 311, the decoding unit 312, and the uncompressed header generation unit 31.
4, while the RTP payload e is output to the uncompressed packet generator 315.

The loss detecting unit 311 stores the link_seq included in the compressed header e supplied from the dividing unit 310.
It should be noted that the loss detection unit 311 uses the link in the compressed header e.
Not only _seq but also in the already received full header packet
The link_seq is also stored at each reception, and if link_seq is not continuous between full header packets or header compression packets sequentially received from the transmitting node 3a, packet loss Is determined to have occurred. Further, the loss detection unit 31
1 detects a packet loss and shifts the shift register 31
6 is shifted, but since no packet loss has occurred between the full header packet a and the header compression packet e, this shift operation will be described later.

On the other hand, the decoding unit 312 calculates the difference value code included in the compressed header e supplied from the division unit 310 from
The difference values ΔX and ΔY of the variable field are calculated. Specifically, it is as follows. (1) When only the magnification information D1 and D2 are included in the difference value code, the decoding unit 312 multiplies the common factor notified in advance by the magnification represented by each of the magnification information D1 and D2 to obtain the difference values ΔX and ΔY is calculated and output to the uncompressed header generation unit 314. (2) When the difference value code includes the magnification information D1 and D2 and the common factor, the decoding unit 312 multiplies the common factor by the magnification represented by each of the magnification information D1 and D2 to obtain the difference value ΔX and the difference value ΔX. ΔY is calculated and output to the uncompressed header generation unit 314. (3) When "1" is set to all of F, D1, and D2 in the difference value code, and when the value of the variable field itself is included in the difference value code, the value of the variable field is uncompressed. Output to the generation unit 314.

On the other hand, the shift register 316 sequentially stores the uncompressed packets that have been received by the receiving node 3b as full header packets or header compressed packets and that have been restored. That is, the uncompressed header stored in the shift register 306 corresponds to the “reception-side reference header” in the claims. Specifically, when the header compressed packet e is restored, as shown in FIG. 10A, the uncompressed headers a to d in the uncompressed packets restored from the full header packets a to d are shifted. It is stored in the first to fourth stages of the register 316. In such a situation, the header selection unit 313
The uncompressed header d stored in the first stage of the shift register 316 and the uncompressed header a stored in the fourth stage are read, and one of the read uncompressed headers is selected to perform uncompressed operation. Output to the header generation unit 314. In other words, the header compressed packet e
One of the uncompressed headers referred to when the compressed header e is generated in the transmission-side node 3a is selected by the header selection unit 313, and the uncompressed header generation unit 31
4 is output. For example, as shown in FIG. 7B, in the transmitting node 3a, the uncompressed header e
The uncompressed headers a and d are referred to at the time of compression. On the other hand, at the receiving side node 3b, when the compressed header e is decompressed, the uncompressed header a
One of d and d is referred to.
As a result, even if either the full header packet a or d is lost between the transmitting node 3a and the receiving node 3b, only one of the uncompressed headers a or d is retained in the receiving node 3b. In this case, the compressed header e can be restored.

The uncompressed header generator 314 uses the difference values ΔX and ΔY (or the variable field itself) supplied from the decoder 312 and the uncompressed header supplied from the header selector 313 to generate The compressed header e of the received header compressed packet e is restored to the uncompressed header e. The details are as follows.

The uncompressed header generation unit 314 includes the decoding unit 31
2, the difference value ΔX or ΔY is supplied from the header selection unit 313, one of the difference value ΔX or ΔY is added to the variable field of the uncompressed header, and the variable field of the compressed header e is added. Restore the value of.
Specifically, when the uncompressed header d stored in the first stage of the shift register 316 (that is, the uncompressed header immediately before the packet e) is supplied from the header selection unit 313, While ΔX is added to the value of the variable field of the compressed header d, the header selection unit 313 sends the uncompressed header a stored in the fourth stage of the shift register 316 (that is, the uncompressed header four packets before the packet e). )
Is supplied, ΔY is added to the value of the variable field of the uncompressed header a. Further, the uncompressed header generation unit 314 generates an uncompressed header e including the variable field thus obtained and a constant value field in the uncompressed header supplied from the header selection unit 313, and generates the uncompressed header e. Is output to the shift register 316 and the uncompressed packet generator 314. Shift register 316
When the uncompressed header e is supplied, the storage contents are shifted by one stage, and the uncompressed header e supplied this time is stored in an empty area. The uncompressed header a, which is no longer likely to be referred to when restoring the compressed header thereafter, is discarded with the shift.

On the other hand, the uncompressed packet generator 315 generates an uncompressed packet e by adding the uncompressed header e supplied from the uncompressed header generator 314 to the RTP payload e supplied from the divider 310. The uncompressed packet e is transmitted to the receiving data terminal 2 via the transmitting unit 32a.

Here, in FIG. 8, the transmitting node 3
The example illustrates a case where the header compression packet f transmitted from a is lost for some reason. The operation of the receiving node 3b in this case is as follows.

As described above, when the header compressed packet f is lost and the header compressed packet g is subsequently received, the loss detecting unit 311 of the receiving node 3b determines the link_seq of the compressed header e received immediately before, and Since link_seq of the compressed header g received this time is not continuous, it should have been received between these packets.
One packet (header compressed packet f) is detected as being lost. In this case, loss detection section 311 shifts shift registers 316 by the number of stages corresponding to the number of lost packets.
Shift the stored contents of. As a result, as shown in FIG. 10B, an area in the shift register 316 in which the packet f which should have been originally received and restored but has been lost is to be stored is an empty area.

Here, as described above, when the compressed header is restored, the first stage and the fourth stage of the shift register 316 are used.
Any one of the uncompressed headers stored in the stage is referred to, but when restoring the compressed header g, the first stage of the shift register 316 is an empty area. Therefore, the header selection unit 313 reads the uncompressed header b stored in the fourth stage of the shift register 316 and outputs it to the uncompressed header generation unit 314. As a result, the uncompressed header generation unit 314 can restore the header compressed packet g even though the header compressed packet f has been lost.

The same operation is performed when a subsequent header compression packet is received. For example, in FIG.
It is assumed that a header compressed packet k is normally received after three header compressed packets h, i, and j are lost. In this case, the loss detection unit 311 refers to the link_seq of the header compression packet g received immediately before and the link_seq of the header compression packet k received this time, so that three packets are lost between these packets. Is detected, the shift register 316 is shifted by three stages. As a result, when restoring the header compressed packet k, the contents stored in the shift register 316 are as shown in FIG.
The state shown in FIG. In this case, the header selection unit 31
3, the uncompressed header g stored in the fourth stage of the shift register 316 is read. Then, the uncompressed header generation unit 314 calculates the contents of the uncompressed header g,
The uncompressed header k is restored using the difference value code included in the compressed header k.

As described above, in the present embodiment, the uncompressed header to be transmitted from the transmitting node 3a to the receiving node 3b is compressed into a compressed header by referring to two uncompressed headers. On the other hand, in the receiving node 3b, one of the uncompressed headers referred to in the compression is referred to, and the compressed header is restored to the uncompressed header. Therefore, packet loss occurs between the transmitting node 3a and the receiving node 3b, and as a result, one of the two uncompressed headers used for header compression is restored in the receiving node 3b. If not,
The compressed header can be restored using the other uncompressed header. Therefore, as compared with the above-described conventional technique, there is an advantage that the number of packets discarded in a chain due to the loss of a certain packet can be significantly reduced.

B: Second Embodiment In the first embodiment, the transmitting node 3a has a function of converting an uncompressed packet into a header compressed packet or a full header packet (hereinafter, referred to as a “compression function”). The receiving node 3b has a function of converting a header compressed packet or a full header packet into an uncompressed packet (hereinafter, referred to as a "decompression function"). The transmitting data terminal 1 may have a compression function, or the receiving data terminal 2 may have a decompression function.

Specifically, the transmitting data terminal 1 converts an uncompressed packet to be transmitted into a full header packet or a header compressed packet and transmits the packet (ie, FIG. 6).
On the other hand, the node device in the network 3 that receives these converts the full header packet or header compressed packet into an uncompressed packet and transmits it to the next node device or the receiving data terminal 2. You may make it. Similarly, a node device located in front of the receiving data terminal 2 converts the received uncompressed packet into a full header packet or a header compressed packet and transmits the packet to the receiving data terminal 2. The side data terminal 2 converts the full header packet or header compressed packet into an uncompressed packet (that is, performs the decompression function illustrated in FIG. 9), and displays an image or sounds according to the data included in the uncompressed packet. May be performed.

As described above, the packet transmission method according to the present invention can be applied not only to the packet relay device shown as the transmitting node 3a or the receiving node 3b in the first embodiment, but also to the source of the packet. Alternatively, the present invention can be applied to a data terminal as a transmission destination. In other words, the concept of “transmission side” and “reception side” in the claims includes not only a packet relay device that relays a packet exchanged between data terminals but also a data terminal that is a source or destination of a packet. It is.

C: Modifications Although one embodiment of the present invention has been described above, the above embodiment is merely an example, and various modifications may be made to the above embodiment without departing from the spirit of the present invention. be able to. For example, the following modifications can be considered. In each of the modified examples described below, those having the function of compressing the uncompressed header, that is, the transmitting node 3a in the first embodiment and the transmitting data terminal 1 in the second embodiment are referred to as “transmitting side”. The receiving node 3b in the first embodiment and the receiving data terminal 2 in the second embodiment are referred to as a "receiving side".

<Modification 1> In the above embodiment, when the uncompressed header is compressed on the transmitting side, the immediately preceding uncompressed header and the fourth uncompressed header are referred to, while the compressed header is restored on the receiving side. At this time, one of the immediately preceding uncompressed header and the immediately preceding uncompressed header is referred to, but the uncompressed header to be referred to in the compression and decompression is not limited to this. , Can be arbitrarily selected. Also,
The number of uncompressed headers referred to when the compressed header is generated on the transmitting side, or the number of uncompressed packets that are candidates for reference in restoring the compressed header on the receiving side (that is, read out by the header selection unit 313 shown in FIG. 9). The number of uncompressed headers) is not limited to two, but may be more than two.

<Modification 2> In the above embodiment, in order to reduce the data amount of the compressed header, the difference values ΔX and Δ
Although magnification information D1 and D2 are included in the compressed header instead of Y, the following may be performed to further reduce the data amount of the compressed header. That is, predetermined conditions for the difference values ΔX and ΔY are held in advance on the transmission side and the reception side, and the difference values ΔX and ΔY obtained based on the uncompressed header to be compressed satisfy the predetermined conditions. The transmitting side transmits the compressed header in which the difference values ΔX and ΔY are omitted (ie, FIG.
The header compression packet including the compressed header not including the difference value code shown in (a) is transmitted to the receiving side. On the other hand, when the receiving side receives the header compressed packet from which the difference value code is omitted, the difference values ΔX and ΔY are restored based on the predetermined condition. Specifically, the following embodiments can be considered.

A. First Mode (Setting of Default Value) For example, among variable fields, the sequence number (SN) in the RTP header often increases by “1” between each packet. As a result, the difference value ΔX of the variable field between the uncompressed header to be compressed on the transmitting side and the immediately preceding uncompressed header is “1”, while the uncompressed header to be compressed and its four It is considered that the difference value ΔY of the variable field from the previous uncompressed header is often “4”. In consideration of such circumstances, default values of ΔX = “1” and ΔY = “4” are held on the transmission side and the reception side, and the transmission side obtains the default values based on the uncompressed header to be compressed. When the difference value ΔX is “1” and the difference value ΔY is “4”, a compressed header in which the difference value code is omitted is generated.
On the other hand, when the receiving side receives a header compressed packet including the compressed header in which the difference value code is omitted, the difference value ΔX is “1” and the difference value ΔY is “1” with reference to the held default value. 4 ".

Similarly, when voice data is to be transmitted / received, it is considered that the time stamp value in the RTP header often increases by a constant value for each packet. Therefore, when referring to the immediately preceding uncompressed header and the four uncompressed headers before compressing the uncompressed header, the difference value ΔX = “constant numerical value”, ΔY = “(constant numerical value) × If the default value of "4" is held, the difference value or difference value code to be included in the compressed header can be omitted in many cases. In general, when referring to the uncompressed header n times before the uncompressed header to be compressed and the uncompressed header m times before, the difference value ΔX = “(constant numerical value) × n”, A default value of ΔY = “(constant numerical value) × m” is stored.

As described above, the values that are likely to be the difference values ΔX and ΔY are previously stored in the transmitting node and the receiving side as default values, and the actually obtained difference values ΔX and ΔY are used as the default values. If they are the same, the difference values ΔX and ΔY or the difference value code may not be included in the compressed header. By doing so, it is possible to further reduce the data amount of the header compressed packet transmitted from the transmission side to the reception side. Note that the obtained difference values ΔX and ΔY
Is different from the predetermined default value, a difference value code is generated from the difference values ΔX and ΔY as described in the above embodiment, and a compressed header including the code is included in the header compressed packet. Alternatively, a compressed header that directly includes the value of the variable field in the difference value code may be included in the header compressed packet.

B. Second Mode (Setting of Conditional Expressions) Predetermined conditional expressions regarding the difference values ΔX and ΔY are held in advance on the transmitting side and the receiving side, and on the transmitting side, the obtained difference values satisfy the above conditional expressions. In this case, a compressed header omitting the difference value is generated, and on the receiving side, when the compressed header omitting the difference value is received, the difference value ΔX is calculated based on the above conditional expression.
And ΔY may be restored. For example, it is as follows.

(1) The difference value ΔX of the variable field is converted into an uncompressed header N to be compressed on the transmission side and an uncompressed header Nn referred to at the time of the compression (in each of the above embodiments, the immediately preceding uncompressed header Nn). Since it is a header, "n =
1 "), while the difference value ΔX (N, N-n)
The difference value ΔY is calculated based on the uncompressed header N to be compressed on the transmission side and the uncompressed header N-
m (in the above embodiments, “m = 4” because the header is the fourth uncompressed header), and the difference value ΔY (N, N−
m), the conditional expressions of ΔX (N, N−n) = ΔX (N−n, N−2n) and ΔY (N, N−m) = ΔY (N−m, N−2m) The difference values ΔX (N, N−n) and ΔY (N, N) obtained based on the current compression target are stored in the transmitting side and the receiving side in advance.
If N−n) satisfies the conditional expression, a compressed header in which the difference value or the difference value code corresponding to the difference value is omitted is generated. The case shown in each of the above embodiments, that is, the case where header compression is performed with reference to the immediately preceding uncompressed header (n = 1) and the immediately preceding uncompressed header (m = 4) will be described with reference to FIG. To explain, for example, a difference value ΔX (j, i) between the uncompressed header j and the uncompressed header i and a difference value Δ between the uncompressed header j and the uncompressed header f
Y (j, f) is a difference value ΔX (f, e) between the uncompressed header f and the uncompressed header e and a difference value ΔY (f, b) between the uncompressed header f and the uncompressed header b. ), The difference values ΔX (j, i) and ΔY (j,
A compressed header omitting f) is generated, and a header compressed packet including this compressed header is transmitted to the receiving side.

On the other hand, on the receiving side, the uncompressed header f
And ΔX (f, e) and ΔY (f, b) are stored in the shift register 316, and when the header compressed packet j from which the difference code or the like is omitted is received later, the above conditional expression is satisfied. And when the compressed header is decompressed, the difference value ΔX (j,
i) and ΔY (j, f), a difference value ΔX (f,
e) and ΔY (f, b) are used. This has the advantage that the data amount of the header compressed packet j can be reduced by the difference value code.

The conditional expressions held in advance on the transmitting side and the receiving side are not limited to those described above. For example, when the uncompressed header N to be compressed is compressed with reference to the n-th uncompressed header and the m-th uncompressed header, the uncompressed header N and the n-th uncompressed header If the difference value Δ (N, N−n) satisfies the conditional expression of Δ (N, N−n) = Δ (N−n, N−n−m) × n / m, the difference value Δ Alternatively, a compressed header in which the difference value code indicating the above is omitted may be generated and transmitted to the receiving side. On the other hand, on the receiving side, when a header compressed packet including the compressed header with the difference value code omitted is received, the difference value Δ to be used for the restoration of the compressed header is determined using the above conditional expression. Can be.

Further, in each of the above embodiments, the data amount is reduced by including the magnification information D1 and D2. However, such a method is used in combination with the method shown in the present modification. Is also good. Specifically, the encoding unit 305 of the transmission-side node 3 a
When the difference values ΔX and ΔY are received from 3 and 304, it is determined whether or not these difference values match the default values shown in the first aspect. As a result, if it is determined that they match, the encoding unit 305 notifies the compressed header generation unit 306 that the difference value code should be omitted.
The information (magnification information D1, D2, etc.) for generating the difference value code by the method described in the embodiment is notified to the compressed header generation unit 306. As a result, the compressed header generation unit 306 generates a compressed header from which the difference value code is omitted when the difference values ΔX and ΔY satisfy a predetermined condition, and generates the first header when the difference value is not satisfied.
A compressed header including the difference value code shown in the embodiment is generated.

[0079]

As described above, according to the present invention,
If at least one uncompressed header among the plurality of uncompressed headers referred to at the time of header compression is held on the receiving side, it is possible to decompress the compressed header by referring to the plurality of uncompressed headers. Therefore, it is possible to avoid a situation where packets are discarded in a chain due to the loss of a certain packet.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of a communication system to which a packet transmission method according to a first embodiment of the present invention can be applied.

FIG. 2 is a diagram illustrating a configuration of a transmitting node 3a and a receiving node 3b in the communication system.

FIG. 3 is a diagram illustrating a configuration of an uncompressed packet.

FIG. 4A is a diagram illustrating a configuration of a header compression packet, and FIG. 4B is a diagram illustrating a configuration of a difference value code.

FIG. 5 is a diagram schematically illustrating the concept of header compression and header decompression in the embodiment.

FIG. 6 is a diagram exemplifying a functional configuration related to a compression operation in the transmitting node according to the embodiment.

FIGS. 7A to 7C are diagrams showing specific examples of a compression operation in a transmitting node.

FIG. 8 is a timing chart showing a specific example of packet exchange performed between a transmitting node and a receiving node in the embodiment.

FIG. 9 is a diagram exemplifying a functional configuration related to a restoration operation in the receiving node according to the embodiment.

FIGS. 10A to 10C are diagrams illustrating a specific example of a restoration operation in a reception-side node.

FIG. 11 is a diagram showing a conventional header compression procedure and its problems.

[Explanation of symbols]

1 ... data terminal on the transmitting side, 2 ... data terminal on the receiving side, 3
..., Network, 3a, transmitting node, 3b, receiving node, 31a, 31b, receiving units, 32a, 32
b transmission unit, 33a, 33b control unit, 34a, 3
4b: storage unit, 301, 310 ... division unit, 302,
316 shift register, 303, 304 subtraction unit, 305 encoding unit, 306 compressed header generation unit, 307 header compressed packet generation unit, 311
Loss detector, 312... Decoder, 313... Header selector, 314... Uncompressed header generator, 315.

Continued on the front page (72) Inventor Takashi Suzuki F-term (reference) in NTT Mobile Communication Network Co., Ltd. 2-1-1, Nagatacho, Chiyoda-ku, Tokyo 5K030 GA12 HA08 JA05 KA13 LA07 MB04 5K034 AA06 BB06 DD03 EE11 FF11 FF13 HH01 HH02 KK21 MM14 SS02

Claims (8)

[Claims] In a network including a transmitting side and a receiving side, the transmitting side converts an uncompressed packet to be transmitted into a full header packet having a full header or a header compressed packet having a compressed header, and sends the converted packet to the receiving side. A method for transmitting a packet, wherein the receiving side restores a full header packet or a header compressed packet received from the transmitting side to an uncompressed packet, wherein the transmitting side includes a packet included in each of the plurality of uncompressed packets. An uncompressed header to be transmitted is held as a transmission-side reference header, and an uncompressed packet to be transmitted is converted into the header-compressed packet by using two or more transmission-side reference headers. The uncompressed header in the uncompressed header obtained by restoring the received full header packet or header compressed packet Held as a reference header, using the reception-side reference header corresponding to at least one transmission-side reference header among the two or more transmission-side reference headers used for conversion to a header compressed packet on the transmission side, A packet transmission method characterized by decompressing the header compressed packet into an uncompressed packet. 2. The uncompressed header includes information that can change by a predetermined difference value for each uncompressed packet, and difference variable information that can change the difference value between each uncompressed packet. The transmitting side obtains a difference value of the difference variable information between each of the two or more transmitting side reference headers and an uncompressed header of the uncompressed packet to be transmitted to the receiving side, and While converting the compressed packet into a header compressed packet including difference value information corresponding to the respective difference values, the receiving side refers to the difference value information included in the received header compressed packet and the at least one transmission side reference. Using the difference variable information included in the reception-side reference header corresponding to the header to restore the difference variable information in the uncompressed packet to be restored from the header compressed packet. Packet transmission method according to 1. 3. The transmission side and the reception side hold a common factor which is a minimum unit of change of the difference variable information, wherein the difference value information includes a magnification of each of the difference values with respect to the common factor, The side, by using the respective difference values obtained by multiplying the magnification included in the difference value information by the held common factor,
3. The method according to claim 2, wherein variable difference information in an uncompressed packet to be restored from the header compressed packet is restored.
2. The packet transmission method according to item 1. 4. The transmitting side, when the difference value satisfies a predetermined condition, converts the uncompressed packet into a header compressed packet from which the difference value information has been omitted. When a header compressed packet from which the difference value information has been omitted is received, the difference value information is generated based on the predetermined condition, and a header to be restored from the header compressed packet using the difference value information is generated. 4. The packet transmission method according to 2 or 3, wherein the variable difference information in the compressed packet is restored. 5. A packet relay device interposed between a plurality of data terminals and relaying uncompressed packets exchanged between the data terminals, comprising: receiving means for receiving an uncompressed packet; Holding means for holding an uncompressed header included in a packet as a transmission-side reference header, and using two or more transmission-side reference headers to compress a part of the received uncompressed packet into a header having a compressed header. A packet relay device comprising: a conversion unit that converts a packet into a packet; and a transmission unit that transmits the header compressed packet. 6. A packet relay device for relaying a plurality of uncompressed packets exchanged between a plurality of data terminals, comprising:
Communicates with the sender that converts some uncompressed packets of multiple uncompressed packets into header-compressed packets using the uncompressed headers contained in the other two or more uncompressed packets and sends them. A packet relay device, receiving means for receiving a packet transmitted from the transmitting side, holding means for storing an uncompressed header obtained from the received packet as a receiving side reference header, and the header from the transmitting side. When a compressed packet is received, the transmitting side uses the receiving side reference header corresponding to at least one of the uncompressed headers among the two or more uncompressed headers used for the conversion to the header compressed packet. Decoding means for restoring the header compressed packet into an uncompressed packet, and transmitting means for transmitting the restored uncompressed packet. And a packet relay device. 7. A data terminal capable of exchanging packets with another data terminal, wherein an uncompressed header included in a plurality of uncompressed packets to be transmitted to the other data terminal is a transmission side reference header. A conversion unit configured to convert a part of a plurality of uncompressed packets to be transmitted into a header compressed packet having a compressed header by using two or more transmission-side reference headers; A data terminal, comprising: transmitting means for transmitting a compressed packet. 8. A data terminal capable of performing packet exchange with another data terminal via a packet relay device, wherein a part of the plurality of uncompressed packets is transmitted to another data terminal. In a data terminal capable of communicating with a packet relay device that converts a packet into a header compressed packet using the uncompressed header included in the above uncompressed packet and transmits the packet, receiving a packet transmitted from the packet relay device Means, holding means for holding an uncompressed header obtained from the received packet as a reference header on the receiving side, and when the header compressed packet is received from the packet relay device, the packet relay device converts the header compressed packet to the header compressed packet. Out of the two or more uncompressed headers used for the conversion, the reception side reference corresponding to at least one of the uncompressed headers. A decompression means for decompressing the header compressed packet into an uncompressed packet using the reference header.