JPH08163139A - Packet communication method - Google Patents

Abstract

PURPOSE: To improve a using efficiency of a network by arranging in common the data Q showing the communication quality as a part of a packet header or a trailer of every hierarchy and referring to the data Q to process the packets in every hierarchy. CONSTITUTION: When the data are transmitted, the data string received from a higher hierarchy is divided in a TCP packet 11 and the communication quality request value Q given from the higher hierarchy is buried in a part 13 of an added TCP header 12. Then the part 13 is transformed into an IP packet 14 and the value Q common to the part 13 is buried into a part 16 of an added IP header 15. The value Q is transformed into an ATM cell 17. At the same time, the value Q common to the part 16 is buried into a part 19 of an ATM header 18. Then the packet processing is properly carried out in every hierarchy based on the value Q, and the cell and the packet are disused. Thus it is possible to prevent the useless packet transmission by the preferential processing, to evade the congestion, to effectively use a transmission line, and to transmit the data in real time and with high quality.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a packet communication method in computer communication, in which a data signal is packetized and transmitted using a protocol defined by a hierarchical structure based on an OSI reference model.

[0002]

2. Description of the Related Art Up to now, in a packet communication system (data communication system) in which a data signal is packetized and transmitted according to a format of each layer by a communication protocol layered into a plurality of layers represented by an OSI reference model, Each layer has a format structure consisting of independent headers, and there is no function of exchanging header information indicating a communication quality request or a network state between layers.

As an example, ATM (Asynchronous Trans
The operation will be described for the case where computer data (IP (Internet Protocol) packet data) is divided into ATM cells and transmitted using the fer mode) as a physical layer. FIG. 6 shows that TCP / I is used as a physical layer using ATM.
The data mapping in each layer when data is transmitted between computers by the P protocol is shown. Reference numeral 71 shows the format structure of a TCP packet, 72 is a TCP header, and 21 is TCP data. 73 shows the format structure of the IP packet, and 74 shows I
P header, 22 is IP data, and 23 is a trailer for ATM cells. 75 indicates the structure of the ATM cell, and 76 indicates A
TM header, 24 is a data section. Reference numeral 77 shows the structure of the ATM cell header, and the header length is fixed at 5 bytes and the data length is fixed at 48 bytes. Note that TCP and I
The P packet length is variable and is a maximum of 64 Kbytes.

At the time of data transmission, a data stream from an upper layer application is divided into TCP segments (packets) and a TCP header is added. Next I
It is mapped to a P packet and an IP header is added. Further, the IP packet is mapped to the ATM cell. AT
There are formats called cell types M, M, 3/4, and Type 5 that have different format structures, but here, Type 5 is used for data transmission.
Shows the case. As can be seen from FIG. 6, the header information used for processing in each layer is independent in each layer and cannot be used across layers.

7 to 10 show examples of system configurations for computer communication. This assumes that one computer communicates with another via ATM nodes and ATM routers. 7 is a system block diagram, FIG. 8 is a hierarchical view, and FIG. 9 is A.
FIG. 10 is a diagram showing the functional configuration of the TM node, and FIG. 10 is a diagram showing the functional configuration of the router. 7, 81 and 87 are computers, 82, 84 and 86 are routers having an IP packet routing function, and 83 and 85 are ATM nodes having an ATM cell switching function.

In FIG. 8, 91 is an ATM layer, 92 is a network layer, 93 is a transport layer, and 94 is an upper layer. ATM layer is physical layer, A
Although it is composed of a TM layer and an AAL (ATM Adaptaion Layer) layer, they are collectively referred to as an ATM layer for convenience. Also, the upper layer is composed of a session layer, a presentation layer, and an application layer in the OSI reference model, but they are collectively displayed as the upper layer. Although the network is composed of a plurality of nodes, two ATM nodes are shown in the figure for simplification.

FIG. 9 is a diagram showing the ATM node 83, in which 101 is an ATM header identifying section, 102 is a switching section, and 103 is a header adding section. FIG. 10 is a diagram showing the router 82.
Is an ATM header identification section, 112 is a data reproducing section, and 113 is an I
P header identification unit, 114 is a routing processing unit, and 115 is I
A P header adding unit, 116 is a cell dividing unit, and 117 is a header adding unit.

In the computer 81, the data from the upper layer is divided into ATM cells according to the format shown in FIG. 6 and sent to the router 82. In the router, as shown in FIG. 10, the ATM header identifying section 111 and the data reproducing section 112 reproduce the IP data except the ATM header information, and
The P header identifying unit 113 further identifies the IP header, and the routing processing unit 114 identifies the I in the network layer.
P packet routing processing is performed, the IP header adding unit 115 adds an IP header, the cell dividing unit 116 divides the ATM cells, and the header adding unit 117 adds an ATM header to the ATM node 83 as an ATM cell again. Send out.

At the ATM node 83, as shown in FIG.
The ATM header identifying unit 101 identifies the header, the switching unit 102 switches the cell in the ATM layer, the header adding unit 103 adds the ATM header, and transfers it to the router 84. Further, similar processing is performed by a plurality of nodes and a plurality of routers, and the computer 87 is reached. The computer 87 restores the IP packet from the ATM cell, reproduces the TCP packet from the IP packet, and passes the data to the upper layer application. The processing of each layer is performed based on the header and trailer information of each layer, and in normal cases, internal data is not referred to.

In the system of FIGS. 7-10, ATM
AT at node 83 or 85 due to congestion of transmission line
Discarding of M cells can occur. Further, in the router 82, 84 or 86, data delay may occur due to overloading of the routing process. In the conventional method, the processing in each layer is performed independently based on the header information of each layer, and therefore, when congestion occurs in the ATM layer, the discarding processing is performed for each ATM cell.

Therefore, the ATM that constitutes the IP packet
Even if some cells are discarded and cannot be used as an IP packet, the remaining ATM cells forming the IP packet are sent out to maintain the congestion. In addition, there is a drawback in that the transmission path cannot be effectively used because the IP packet processing is delayed and the useless IP packet that cannot be used by the application requiring the real-time property is also transmitted.

[0012]

SUMMARY OF THE INVENTION It is an object of the present invention to transmit useless packets by commonly using data indicating communication quality in each layer in computer communication using a protocol defined by a layered structure. It is an object of the present invention to provide a method of improving the use efficiency of a network by preventing the above and effectively using the transmission line capacity.

[0013]

According to the present invention, in a packet communication system in which a data signal is packetized and transmitted in accordance with a format of each layer by a communication protocol layered into a plurality of layers, data indicating communication quality is transmitted to each layer. The packet header and the trailer are commonly arranged, and the packet processing can be performed by referring to the data indicating the communication quality in each layer.
As the data indicating the communication quality in the present invention, a discard priority, a congestion state, a delay time, etc. can be used.

[0014]

In the present invention as described above, since the data indicating the communication quality is commonly arranged as a part of the packet header or trailer of each layer, the packet processing is performed using the data indicating the communication quality in each layer. It can be done properly.

[0015]

FIG. 1 shows a packet format according to an embodiment of the present invention. ATM is used as the physical layer and TCP is used.
3 shows data mapping in each layer when data is transmitted between computers by the / IP protocol. The same parts as those in FIG. 6 are given the same numbers. Reference numeral 11 shows a TCP packet format structure, 12 is a TCP header, and 21 is TCP data. 14
Shows the format structure of an IP packet, 15 is an IP header, 22 is IP data, and 23 is an ATM cell trailer.

Shaded parts 13, 16 and 19 in the header part of each packet represent data indicating communication quality commonly arranged in each layer according to the present invention. Although it is arranged in the header in this figure, it may be arranged in the trailer. 17 shows the structure of the ATM cell, and 18 shows the AT.
M header, 24 is a data part. Reference numeral 20 shows an example of the structure of an ATM cell header having a fixed header length of 5 bytes and a data length of 48 bytes. Note that TCP and I
The P packet length is variable.

At the time of data transmission, the data stream from the upper layer application is divided into TCP segments (packets) and a TCP header is added. At this time, the communication quality request value from the upper layer application is set to TCP.
Embed in part 13 in the header. Next, this is mapped to an IP packet and an IP header is added. At this time, similarly, the data corresponding to the part 13 in the TCP header is embedded in the part 16 of the IP header. Further, the IP packet is mapped to the ATM cell. Also at this time, the data corresponding to the part 16 of the IP header is mapped to the part 19 in the ATM header.

Type 1 is used for the cell conversion method to ATM.
There are formats called 2, 3/4 and Type 5, and the format structures are different respectively, but here, the case of Type 5 intended for data transmission is shown. In type 5, only the trailer is added to the end of the user data (IP packet) and the user data (IP packet) is placed in the cell data area.

The word lengths of the data 13, 16 and 19 in the headers or trailers that are commonly arranged do not necessarily have to be the same, and the data lengths corresponding to the data lengths of the respective layers are based on the conversion table created in advance. Values can be used. Further, the position in the header or the trailer can be arbitrarily set using the spare data or the like.

2 to 5 show examples of system configurations for computer communication. This assumes that one computer communicates with another via ATM nodes and ATM routers. 2 is a system block diagram, FIG. 3 is a hierarchical view, and FIG. 4 is an AT.
FIG. 5 is a diagram showing the functional configuration of the M node, and FIG. 5 is a diagram showing the functional configuration of the router. In FIG. 2, 31 and 37 are computers, 32, 34 and 36 are routers having an IP packet routing function, and 33 and 35 are ATM nodes having an ATM cell switching function.

In FIG. 3, 41 is an ATM layer, 42 is a network layer, 43 is a transport layer, and 44 is an upper layer. ATM layer is physical layer, A
Although it is composed of a TM layer and an AAL (ATM Adaptaion Layer) layer, they are collectively referred to as an ATM layer for convenience. Also, the upper layer is composed of a session layer, a presentation layer, and an application layer in the OSI reference model, but they are collectively displayed as the upper layer. Although the network is composed of a plurality of nodes, two ATM nodes are shown in the figure for simplification.

FIG. 4 is a diagram showing the ATM node 33.
Is an ATM header identification unit, 52 is a switching unit, and 53 is a header addition unit. FIG. 5 is a diagram showing the router 32, in which 61 is an ATM header identifying unit, 62 is a data reproducing unit, 63 is an IP header identifying unit, 64 is a routing processing unit, 65 is an IP header adding unit, 66 is a cell dividing unit, 67 is a header addition part. Figure 7
The system shown in FIG. 10 has the same basic functions as the system shown in FIG. 10, but is different in that it includes a processing function using header information according to the present invention.

In the computer 31, the data from the upper layer is divided into ATM cells according to the format shown in FIG. 1 and sent to the router 32. In the router 32, as shown in FIG. 5, the ATM header identifying section 61 and the data reproducing section 62 reproduce the IP data except the ATM header information.
The IP header identification section 63 further identifies the IP header, and the routing processing section 64 identifies the IP in the network layer.
Performs packet routing processing and adds IP header
An IP header is added at 65, and ATM is applied at the cell division unit 66.
The cells are divided into cells, an ATM header is added by the header adding section 67, and the cells are sent again to the ATM node 33 as ATM cells.

In the ATM node 33, as shown in FIG. 4, the ATM header identifying section 51 identifies the header, and the switching section 52 performs cell switching in the ATM layer.
An ATM header is added by the header adding unit 53 to the router 34.
To transmit. The same processing is performed by a plurality of nodes and a plurality of routers, and the computer 37 is reached. Computer 37
Then, the IP packet is restored from the ATM cell, the TCP packet is reproduced from the IP packet, and the data is passed to the upper layer application.

An embodiment in which the discard priority is used as common header or trailer information (data indicating communication quality) will be described based on the system configuration of FIG. For example, congestion of a transmission line or overload of cell switching processing may occur in the ATM node 33, and data delay may occur in the router 32 due to overloading of routing processing.

In such a case, according to the present invention, when cell congestion occurs in the ATM layer, the ATM layer of the node performs cell discard according to the discard priority identified by the ATM header identification section 51 of FIG. When an overload of the routing process occurs in the IP layer, the IP layer can similarly drop the IP packet according to the priority identified by the IP header identification unit 63 in FIG. Congestion processing can be performed independently according to the capability and the priority of data to be discarded is common.

As described above, in the past, congestion processing was performed only in a limited layer, so that when congestion occurred in another layer, a large quality deterioration occurred. However, in the present invention, the discard priority is set in each layer. The quality deterioration can be suppressed to a minimum because the processing can be performed accordingly.

Next, an embodiment in which the congestion state is used as the common header or trailer information will be described. In the ATM node 33, assume that congestion occurs in the ATM layer and cell discard occurs. In that case, the ATM of FIG.
The header identification unit 61 and the data reproduction unit 62 put the congestion state on the IP header so that the IP header identification unit 63 can refer to it in the processing in the IP layer.
Therefore, by discarding the entire IP packet,
This has the advantage that unnecessary packet transmission can be stopped and congestion can be avoided in subsequent nodes.

Further, when the delay time is used as the common header or trailer information, for example, routing processing takes a long time, and an IP packet which cannot be used by an application requiring real-time processing is provided on the way to a router or node on the way. It is possible to make effective use of the transmission line by detecting and discarding this. Since the technology for synchronizing the time of each network node is already widespread, the delay time can be easily calculated.

As described above, according to the present invention, the data indicating the communication quality is commonly arranged as a part of the packet header or the trailer of each layer (layer), and the data indicating the communication quality is referred to in each layer. Since packet processing can be performed in each layer, cell discard processing or packet discard processing with a delay time exceeding the allowable value can be independently performed for each layer, so that useless packet transmission can be prevented, congestion avoidance and transmission path It can be effectively used.

Further, by referring to the above common header information and preferentially processing a packet having a high priority or a packet requiring a low delay, real-time transmission or high-quality transmission can be realized. It is possible.

[0032]

As described above, according to the present invention, in computer communication, data indicating communication quality is commonly used in each layer to prevent useless packet transmission and improve transmission line capacity. It is possible to provide a network with high usage efficiency by using the network.
Further, real-time transmission or high-quality transmission can be realized.

[Brief description of drawings]

FIG. 1 is a diagram showing a packet format according to an embodiment of the present invention.

FIG. 2 is a block diagram of a configuration of an embodiment according to the present invention.

FIG. 3 is a diagram showing a hierarchical structure of an embodiment of the present invention.

FIG. 4 is a diagram showing a functional configuration of an ATM node according to an embodiment of the present invention.

FIG. 5 is a diagram showing a functional configuration of a router according to an embodiment of the present invention.

FIG. 6 is a diagram showing a packet format of a conventional example.

FIG. 7 is a block diagram of a configuration of a conventional example.

FIG. 8 is a hierarchical view of the configuration of a conventional example.

FIG. 9 is a diagram showing a functional configuration of a conventional ATM node.

FIG. 10 is a diagram showing a functional configuration of a conventional router.

[Explanation of symbols]

11 TCP packet format structure 12 TCP header 13, 16, 19 Data indicating communication quality commonly arranged in each layer 14 IP packet format structure 15 IP header 17 ATM cell structure 18 ATM header 20 ATM cell header 21 TCP data 22 IP data 23 ATM cell trailer 24 Data section 31, 37 Computers 32, 34, 36 Routers 33, 35 having an IP packet routing function A having a switching function for ATM cells
TM node 41 ATM layer 42 Network layer 43 Transport layer 44 Upper layer 51 ATM header identification part 52 Switching part 53 Header addition part 61 ATM header identification part 62 Data reproduction part 63 IP header identification part 64 Routing processing part 65 IP header addition part 66 Cell Division Unit 67 Header Addition Unit 71 TCP Packet Format Structure 72 TCP Header 73 IP Packet Format Structure 74 IP Header 75 ATM Cell Structure 76 ATM Header 77 ATM Cell Header Structure Example 81, 87 Computers 82, 84, 86 router having routing function of IP packet 83, 85 A having switching function of ATM cell
TM node 91 ATM layer 92 Network layer 93 Transport layer 94 Upper layer 101 ATM header identification unit 102 Switching unit 103 Header addition unit 111 ATM header identification unit 112 Data reproduction unit 113 IP header identification unit 114 Routing processing unit 115 IP header addition unit 116 cell division unit 117 header addition unit

Claims (4)

[Claims] 1. In a packet communication system for packetizing and transmitting a data signal according to a format of each layer by a communication protocol layered into a plurality of layers, data indicating communication quality is part of a packet header or trailer of each layer. The packet communication method is characterized in that the packet processing can be performed by referring to the data indicating the communication quality in each layer. 2. The packet communication method according to claim 1, wherein the discard priority is used as the data indicating the communication quality. 3. The packet communication method according to claim 1, wherein a congestion state is used as data indicating communication quality. 4. The packet communication method according to claim 1, wherein a delay time is used as the data indicating the communication quality.