JPH01225264A - Packet communicating system - Google Patents

Abstract

PURPOSE:To eliminate the retrieval of a routing table per packet in a node by including a link number, which can directly identify a packet sending destination, in the contents of a header. CONSTITUTION:In a node n0, a link number RNA of a link l0 is 00 and the RNA of a link l2 is 01. In a node n4, the RNA of a link l3 is 000, the RNA of a link l5 is 001, the RNA of a link l6 is 010 and the RNA of a link l8 is 011. When the packet is transmitted from the node n0 to a node n7, the route areas of the header go to be the 01, 011 and 11. In a node n3, the first 01 of the route area is observed and the packet is sent to the link l5. Then, the route area is reloaded so as to be 011 and 11. In the node n4, the first 011 of the route area is observed and the packet is sent to the link l5. Then, the route area is reloaded so as to be 11. In the node n7, the route area 11 is observed and it is known that the packet is its own node addressing packet.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a packet communication system that eliminates the need for a node to search a routing table for each packet and simplifies the node configuration.

(Prior Art) As a conventional packet communication method, a path (hereinafter referred to as VP: Virtual Path) commonly used by virtual lines belonging to multiple calls is defined between each originating node and a terminating node. An identifier that identifies this route (hereinafter V
PI: Virtual Path Identifier
r) The terminating node identifies the call to which the packet belongs and sends it to an appropriate terminal.
There is a method of displaying (:allIdentifier) in the header of a packet and selecting a transfer route within a communication network based on the VPI.

However, in this method, a node needs a routing table that indicates the correspondence between VPI and packet transfer destination.
Furthermore, since it is necessary to search this routing table for each packet, if the routing processing in the node is performed by software, it will increase the load on the processor, and if it is performed by hardware, it will lead to the complexity of the hardware. This will be explained below.

FIG. 5 shows the header configuration of this conventional method, and FIG. 6 shows the notebook configuration when routing processing is executed by hardware in this conventional method. Further, FIG. 4 is an example of VP and VPI in a communication network, where n Or n +
+ ”' * ” B is a node, no.

It l + "" + 111 is a link, the thick line vp is a vP from node n0 to node n, and this v
The VPI of P is 0001000. Now, the routing process when a packet belonging to this vP arrives at node n3 will be explained with reference to FIG. In Fig. 6, the power links 11, 11', 11'' are the links IL2 in Fig. 4.
゜X,,X, input side, output link 12, 12°, 12
” corresponds to the output side of links ftt, i, s, ftt in FIG. 4. Also, output link 1
2゜12°, 12'' are the output terminals (00), (01), (1
0). Here, symbol (xl * X2
)(XI+x2==Q or 1) indicates that the address of the output terminal is Xl*X2. The self-routing packet switch 13 adds the address of a desired destination output terminal to the beginning of the packet as a routing tag 14, thereby sending the packet to the desired output terminal through hardware processing.

As shown in Figure 6, VPI is "0001000"0'
) When the hat 15 arrives, it enters a mechanism 16 that extracts the VPI. The extracted VPI is sent to the routing table 17 in order to identify which output terminal to send out to by the packet switch 13, and the routing table is set as VPI="0001000" (7
) term and recognize that the output terminal address is old. This value is sent to the routing tag adding section 18, where it is added to the packet as the routing tag 14. The packet switch 13 sends the packet to the output terminal (Ol) according to the routing tag. Thereafter, the routing tag is removed from the packet by the routing tag removal unit 19, and the packet is transferred to the output link 12°.

(Problems to be Solved by the Invention) As explained above, in the conventional method, since the VPI does not directly represent the packet destination link or the output terminal address of the packet switch, the routing table is searched for each packet. , it is necessary to find the output terminal address of the packet switch. As a result, the node requires many mechanisms such as a routing table, a VPI extraction mechanism, a routing tag addition section, and a routing tag removal section, and its hardware becomes complicated.

An object of the present invention is to provide a packet communication method that eliminates the need for a node to search a routing table for each packet and simplifies the node hardware.

(Means for Solving the Problems) The present invention has a communication network configured by connecting a plurality of nodes with links, and the header part of the packet transferred therein identifies the transfer route of the packet. In the packet communication method, in which the packet is transferred from the originating node to the destination node by selecting the destination of the packet according to the information in this area, each node is individually own node, and 1
The structure of the area that assigns a link number (RNA) connected to an adjacent node or node via two links and makes it possible to identify the packet transfer route in the packet header is the structure of the area that allows the packet transfer route to be identified in the packet header. It starts with the RNA of the packet destination at , and ends with the RNA representing the destination node's own node.

The packet destination RNA at each node through which the packet passes on the packet transfer route is arranged as a number string in the order in which the packet passes.The node checks the first RNA in the above area of the arriving packet, and if it is its own node number. Identifies that this node is the destination node, identifies the destination of the packet using that number if it is not the own node number, and writes the contents so that the first RNA in the above area of the header becomes the destination of the next node. It is characterized by shifting and sending out packets.

The conventional technology is that in order to eliminate the need for nodes to check the correspondence between the header contents and the packet destination using a routing table, the header contents include RNA that allows the packet destination to be directly identified; The difference is that a processing method at the node is provided that makes it easy to identify the destination.

(Example) FIG. 1 is a diagram for explaining the present invention in detail, in which n O+
n l + “” + nB is a node, flow
j! l * ””+ I-II are links, A, B, C
are all from node n0 to node n 3 * n 4
A represents the format on link 1□, B represents the format on link 1□, and C represents the format on link 18. 2 is a diagram for explaining the header structure of a packet in the present invention, and FIG. 3 is a block diagram of a node for explaining the operation of the node in this embodiment.

To carry out this method, first the link number RNA is determined at each node. The RNA in this embodiment is defined as the encircled number attached to the link near each node in FIG. 1.

That is, at node no, the RNA of link 1° is 0.
0, and the RNA of link It2 is 01. In node n4, the RNA of link 23 is 000, link 1. RNA of is 0011, RNA of link It6 is old 0, link u'8
The RNA of is 011. Also, the RNA of the own node is node no l nu I am, na + n,, l
``''+na is 11 degrees, and node n4 is 111. Since RNA is individually assigned to each node, the same link can take on different RNAs depending on the node. For example, the RNA at node n of link 1° is 01, and the RNA at node n4 is 000. These RNAs
The packet transfer method using this will be explained below.

In the present invention, as shown in FIG. 2, the area in the header of a packet that enables transfer route identification is the destination RNA at the node where the packet sent from the source node first arrives, and the area where the packet arrives next. Destination RNA at the node - Destination RNA of the node that arrives last
(that is, own node number).

Therefore, in this embodiment, from node no, n 3
The configuration of this area (hereinafter referred to as the route area) of the header of the packet transferred to n7 via node n7 is as follows:
At the time of issuing from 0, the number sequence is na: the destination RNA, n4: the destination RNA, and nt: the own node RNA. The destination for na is link 25, and this R
NA is 01. Also, the destination for n4 is link I
L8 and this RNA cjotx. Finally n
The own node RNA of 7 is 11. Therefore, node n0
From na to na, the route area becomes "01,011,11" as shown in A of FIG. Node n3 sees the first “Ol” in the route area and links the packet to link I.
Send to t5. At this time, the contents of the route area are shifted so that the destination RNA in n4 is placed first, so
The contents of the route area are "old 1°11" as shown in B. At n4, the first "011" in the route area is seen and the packet is sent to the link 28. At this time, the contents of the route area are similarly shifted and become "11" as shown in C.

n7 sees the first "11" and knows that this node is the destination node.

The operation of the node in such a transfer method will now be explained with reference to FIG. Since the RNA is assigned independently to each node, it is easy to match the output terminal address of the self-routing packet switch 1 with the RNA. When the RNA is provided in this way, the first RNA 5 in the header 4 of the packet 3 matches the output terminal address and can therefore serve as a routing tag in the conventional method. Therefore, packets arriving from input links 2.2° and 2'' can be directly input to the packet switch 1, and the output is shifted by the route area shift mechanism 6 to shift the contents of the route area to the output link 7.7°. 7”, the processing of the node is completed.

Since the above-described operation is performed, there is no need to search the routing table for each bucket to establish a correspondence between header contents and output terminal addresses, and to add and remove routing tags.

As is clear from this result, the routing table and routing tag addition/removal unit in the node of the conventional method are no longer necessary, and the configuration of the node can be simplified.

(Effects of the Invention) As explained above, since the output terminal of a node can be directly identified from the route area of the header, each component including the routing table can be omitted in the node.
Simplification of nodes is possible.

[Brief explanation of the drawing]

FIG. 1 shows an embodiment of the present invention, FIG. 2 shows a packet configuration according to the present invention, FIG. 3 shows a node configuration in an embodiment that clearly shows the effects of the present invention, and FIG. 4 shows an originating node and a destination node in a conventional method. An example of vP and VPI defined between,
Figure 5 shows the bucket configuration of the conventional method in which VPI is provided in the header as an area that allows the transfer route to be identified, and CI is provided as the area that allows the call to be identified. Figure 6 shows the routing table and addition/removal of routing tags. This is a node configuration of a conventional method with a mechanism. 1... Self-routing packet switch, 2.2
', 2' - Input link, 3... Packet according to the present invention, 4... Header, 5... First RNA of the route area of the header, 6... Route area shift mechanism, 7.- Output link where RNA is 00, 7° - Output link where RNA is Ol, 7'' - Output link where RNA is lO, (00), (old), (10), (11) - Packet switch Output terminal address, 11.11°, 11” - Input link, +2,1
2°, 12''... Output link, 13... Self-routing packet switch, 14-... Routing tag, 15-... Packet, 16-VPI extraction mechanism, 17-... Routing table, 18-... Routing tag addition section, 19--routing tag removal section.

Claims (1)

[Claims] In a packet communication method that transfers a packet from a source node to a destination node according to information in a header part in a communication network in which a plurality of nodes are connected by links, each node is connected to a node. The area indicating the packet transfer route in the header part of the packet includes a number string indicating the link number at each node in the order in which the packet passes, and the area indicating the packet transfer route in the header part of the packet. and a link number indicating the last destination node, and each node determines the link to take in or forward the packet according to the first link number in the area of the packet, and shifts the number string when forwarding the packet. A packet communication method characterized in that data is transferred so that a link number indicating a link to be transferred at a next node appears at the beginning of a number string.