JPS63155839A - Routing information management method for packet switching networks - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an adaptive routing method for a packet-switched network, and particularly to a routing information management method for a packet-switched network suitable for reducing the amount of distribution of routing information. [Prior Art] In a conventional adaptive routing method in a packet-switched network, when a network management agent creates a routing table as routing information and distributes it to each switching center, 1 the created routing table is A method is used in which each switching center converts the data into packets and distributes them as they are. This type of scheme is known, for example, by Andraw, S.
“Computer Ne
tworks”, Prautice Hall, (19
1981) pρ200-202.

In addition, if each switching center consists of a management module that has a function of creating a routing table and one or more switching modules, the management module must distribute the routing table to each switching module within the local switching center. In conventional systems, all routing tables are distributed to each switching module, regardless of the routing table created by the management module.

[Problem that the invention seeks to solve]

The above conventional technology is based on the premise that all parts corresponding to the switching center are sent to all switching centers in the network, regardless of the routing table created by the network manager.
It does not consider how to reduce the amount of distribution tables, and it assumes that routing tables are not distributed within the exchange, and does not consider the method of distributing routing tables to multiple exchange modules. There is also no consideration given to the amount of distribution tables within the exchange.

Therefore, especially in large-scale networks with a large number of switching centers, the amount of routing tables distributed by the network management unit to each switching center may be limited, and if the number of switching modules that make up the switching center is large, the amount of routing tables distributed to each switching center may be limited. There is a problem in that the amount of tables distributed to each exchange module increases significantly.

The purpose of the present invention is to solve the above problem and reduce the amount of routing tables distributed by network management to each switching center.

The purpose of this invention is to reduce the amount of routing tables distributed by the management module to each switching module within each exchange, thereby reducing the amount of tables distributed in the mesh body.

[Means for solving problems]

The above purpose is to maintain the routing table that the network manager previously created and distributed to each switching center, and when a new routing table is created, to compare it with the previously distributed routing table. Extract the changed part, distribute only the changed part to each exchange, and if there is no changed part in the routing table corresponding to the distributed exchange, the processing will not distribute the routing table to the exchange. This is achieved by Similarly, within each exchange, the management module maintains the routing table that was previously created and distributed to each exchange module, and when a new routing table is created, it is compared with the previously distributed routing table. By doing this, we extract the changed part,
This is achieved by distributing only the changed part to each exchange module, and if there is no changed part in the routing table corresponding to the exchange module to be distributed, the routing table is not distributed to that exchange module.

[Effect]

In an adaptive routing method with network management,
Periodically or when the status of each switching center (line load or presence or absence of a failure) changes, the network manager creates a routing table taking into account the current status of the network. -Normally, exchanges whose own routing tables change are often part of all exchanges in the network, and there is no need to distribute tables to exchanges whose routing tables do not change. There is also an exchange with changing tables.

It is often the case that only a part of the local station's routing table changes, not all of it. The same thing can be said when the management module within the exchange distributes the routing table to each exchange module.

In the present invention, when the network manager creates a new routing table, by comparing it with the previously created routing table, only the parts of the routing table that need to be distributed to each switching center can be distributed. extract,
Since the portion of the extracted portion for each exchange is distributed to the exchange, the amount of distribution tables in the network is reduced. In addition, when the management module creates a new routing table, by comparing it with the previously created routing table, it can extract only the parts of the routing table that need to be distributed to each exchange module, and By distributing the extracted portion corresponding to each exchange module to the exchange module, the amount of distribution tables within the exchange is reduced.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing the network configuration of a packet switching network in an embodiment of the present invention. A packet-switched network requires network management.
It is composed of packet switching stations 2 to 5 and relay lines that interconnect them. In addition to normal switching functions, the network management unit 1 has a function of creating a routing table based on status reports from each packet switching center 2 to 5 and distributing it to each switching center.

The packet-switched network shown in FIG. 1 is connected to packet-type terminals, host computers, etc. that communicate via the switched network, but these are omitted because they are not directly relevant to the explanation of this embodiment.

FIG. 2 is a diagram showing that a failure has occurred in the switching network in FIG. 1. For example, assume that a failure occurs in the relay line between node A and node 9.

FIG. 3 is a diagram showing the configuration of one of the packet switching centers 2 to 5 in FIG. 1, for example, the packet switching center 3. The packet switching station 3 is a routing information management unit that has a function of creating a routing table for its own station number exchange module from the routing table distributed from the network management unit 1 and the load, status, etc. of each relay line within its own station. A management module 6, and exchange modules 7 to 9 which are exchange processing units that perform packet exchange processing according to a routing table created by the management module 6.

It is composed of a module bus 10 that connects each of the above modules. The exchange modules 7 to 9 include trunk lines 11.
13, subscriber lines connecting to host computers and terminals are accommodated. Other packet switching stations also have the same configuration as the packet switching station 3 described above.

FIG. 4 is a diagram showing an inter-switching center distance table created by the network manager 1 based on reports from each switching center 2 to 5 in the network. Figure 4(a) shows the inter-exchange distance table in the normal state in Figure 1, and Figure 4(b) shows the table in Figure 2 when a failure occurs in the trunk line between node A and node 9. This is the inter-exchange distance table. Here, the distance between exchange offices is
This is the minimum number of relay exchanges between the two concerned exchanges (including the originating station, but excluding the destination station). For example, the bottom row of the inter-switching center distance matrix in FIG. 4(a) is the packet switching center 5 (node D).
The distance from network management node 1 (node N) is 2, and similarly the distances to nodes A, B, C, D are 1, 1, 1, . . . , respectively. Indicates that O. In FIG. 2, when a trunk line failure occurs between nodes A and D, the distance between nodes A and 9 changes from 1 to 2. In the inter-exchange distance matrix in FIG. 4(b), elements marked with 0 indicate that the elements have changed due to the occurrence of a failure.

FIG. 5 shows a routing table for each exchange created by the network manager 1. The routing table indicates the distance from an adjacent switching center to a destination switching center when each relay route is selected when a packet is relayed from each switching center to a destination switching center.

Figure 5(a) is the routing table in the normal state in Figure 1, and Figure 5(b) is the routing table in Figure 2 when a failure occurs in the trunk line between node A and node 9. be. In the routing table of FIG. 5(a), for example, at the top of the part where the source station is node B, when node B relays a packet addressed to the destination station (node N),
When g, h, and e are selected as relay routes, the distances from the adjacent station to the destination station (node N) are 0 and 1.2, respectively. The above routing table has the fourth
It is created by editing the corresponding rows of the inter-exchange distance matrix shown in the figure.

FIG. 6 is a diagram showing a routing table held by the packet switching center 3 (node B). Figure 6(a) shows
The ei management station 1 distributes the portion of the routing table 15 whose originating station corresponds to the node B, and the management module 6 of the node B creates the routing table based on the distributed portion. FIG. 6(b) shows the management module 6
is a routing table created from the routing table 16 and distributed to each exchange module. The routing table 17 indicates the priority of the sending route when sending a packet to each destination station.

Each switching module 7 to 9 of the node B maintains a routing table 17 shown in FIG. 6(b), and actually performs packet routing processing by referring to the routing table. For example, routing table 1
The top row of 7 shows that the first, second, and third priority routes are each g* h+ 1 as the sending route when node B relays a packet addressed to node N. .

Next, with reference to one or more figures, a description will be given of the process of changing and distributing the routing table at the network management station 1 when a failure occurs.

When the network is operating in the normal state shown in Figure 1,
For example, suppose that a failure occurs in the relay line between node A and node 0 as shown in FIG. Nodes A and D report to the network management station 1 that a failure has occurred in their own relay lines using status notification packets. By receiving reports from nodes A and D, network management station 1 learns that a failure has occurred in the relay line between node A and node 0, changes the connection status between each node, and changes the current status. The distance between each switching center is calculated based on the connection state between the nodes, and an inter-switching center distance matrix shown in FIG. 4(b) is created. Here, the network service station 1 compares the inter-exchange distance matrix shown in FIG. 4(a) that was held during normal operation and the newly created inter-exchange distance matrix shown in FIG. 4(b) row by row. Then, extract the changed part. In this example, the second row (
(originated from node A) and the fifth row (originated from node 0) are extracted as changed parts.

Next, the network management station 1 creates a routing table 15 for each exchange. Check the originating nodes (A and D) in the row of the inter-exchange distance matrix extracted as the changed part above and the node in the adjacent station column of the routing table 15, and set the node A or D in the adjacent station column of the routing table 15. D
If exists, the element in the corresponding vertical column of the routing table 15 is replaced with the corresponding change row of the inter-exchange distance matrix. For example, the column where node N originates and adjacent station is A is updated by substituting the fifth row (originating node A) of the inter-exchange distance matrix. In the manner described above, the portions (1) to (2) of the routing table 15 indicated by arrows at the bottom of the table are updated.

After the network management station 1 finishes creating the routing table 15, it distributes the changed portion of the routing table to each node in the network. Here, there is a changed part (3) in the part where the NRC of the local station is issued, but this is the routing table of the local station and is not distributed. Further, since there is no change in the routing table of the exchange for nodes A and D, distribution is not performed. That is, the network management station 1 sends only the changed portions ■ and ■ to the node B and the changed portions ■ and ■ to the node C as packets. As shown above, the network management local station 1 uses one of the created routing tables.
There is no need to distribute the whole of 5 to each node, it is sufficient to distribute only the changed parts. In this embodiment, since the creation process of the routing table 15 is performed only for the changing part of the inter-exchange distance matrix shown in FIG. 4(b),
This method also has the advantage that the table creation processing load can be reduced compared to the case where all tables are created.

Next, the creation and distribution of the routing table at each switching center when the routing table is received from the network management station 1 will be explained.

For example, the management module 6 of the node B stores the changed part of the routing table 15 distributed above by the network management station 1.
Suppose that and ■ are received. The management module 6 is the route of the routing table 16.

Update the part corresponding to i with the received changed part and ■. Thereafter, the management module 6 creates a routing table 17 addressed to each exchange module 7 to 9 from the created routing table 16.

Here, the management module 6 creates the routing table 17 by considering the load on the relay line of its own station, the line status of its own station, etc. However, for simplicity, there is no extreme difference in the load of each relay line. Let us consider that they are approximately equal.

The management module 6 uses the created routing table 1
6 is compared line by line with what was created last time, and the changed parts are extracted.The changed parts extracted here are ■ and ■0.
Next, priority is given to each changing part in order from the direction with the smallest distance, that is, 0 For example, for the changing part ■, the first priority route is given in order from the first priority route, and this is also The routing table 17 is updated accordingly. Similar processing is performed for the changed portion (■).

After completing the creation of the routing table 17, the management module 6 distributes the routing table 17 to each exchange module 7-9. At this time, the management module 6 distributes only the updated portions (2) and (2) of the created routing table 17 to each of the exchange modules 7-9. Each exchange module 7-9 has a routing table 17
It receives the changed parts (2) and (2), updates its own module's routing table, refers to the routing table, and performs packet routing processing.

As shown above, it is not necessary for the management module 6 to distribute the entire routing table 17 that has been created, and it is sufficient to distribute only the changed parts.
Since the creation process of the routing table 17 is performed only for the changed part number 2 of the routing table 16, it also has the advantage that the table creation process load can be reduced compared to the case where all tables are created.

〔Effect of the invention〕

As described above, according to one aspect of the present invention, a routing table is distributed from a network management station to each switching station.

When distributing the routing table from the routing table and management module to each exchange module, by extracting the parts that have changed from the previously distributed routing table and distributing the changed parts only to the exchanges and exchange modules that need to be distributed, The amount of routing tables distributed from the current station to each switching station and from the management module to each switching module can be reduced compared to when all of them are distributed, and the amount of table distribution processing for the mesh body can be reduced. It has this effect.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the network configuration in an embodiment of the present invention, and FIG.
Figure 1 shows the failure that occurred, Figure 3 shows the configuration of the packet switching center, Figure 4 shows the inter-switching center distance matrix,
FIG. 5 is a diagram showing the routing table of the network management station;
FIG. 6 is a diagram showing the routing table of the management module. 1...Network management station, 2-5...Packet switching center, 6.
・・Management module, 7 to 9 ・・Replacement module, 1
o...Module bus, 11-13...Relay line,
14゜15...Network control local station routing table, 16
゜Early lI2] 2nd station 2-5... 8th Kesoto intersection 17th station 3rd [] 412th] /1-13... Middle group line

Claims (1)

[Claims] 1. An exchange having a function of creating information for routing distributes the routing information to each other exchange, and each exchange performs routing processing by referring to the routing information. In the routing method of a packet switching network that performs the A routing information management method for a packet switching network characterized by reducing the amount of routing information to be distributed by distributing the information. 2. In the routing information management system as set forth in claim 1 above, each switching center includes one or more switching processing units that perform packet switching processing according to the routing information, and creates the routing information, and and a routing information management unit that has a function of distributing to each exchange processing unit, the routing information created by the routing information management unit is distributed to other exchange processing units as required by the exchange processing unit, A routing information management method for a packet switching network, characterized in that the amount of routing information distributed throughout the network is reduced by extracting the portion that has changed since the previous distribution and distributing it to each switching processing unit.