JPH0955733A - Method and system for dividing communication network - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a communication network dividing method which reflects a plurality of administering methods in consideration of the morphology of connections between nodes for reliability improvement and efficient fault analysis and management. SOLUTION: This method for dividing the communication network consists of a stage (step 10) wherein a 1st node in a 1st cluster generates adjacency information showing the connectivity of the 1st cluster with a 2nd adjacent cluster, a stage (step 12) wherein a subordination relation setting rule between clusters applied to the generated adjacency information is decided, a stage (step 20) wherein one of the 1st and 2nd clusters is assigned as a subordinate cluster and the other is assigned as an administrator cluster according to the applied subordination relation setting rule, and a stage (step 22) wherein the subordinate cluster generates a new cluster subordinated to the administrator cluster.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of constructing a communication network, and more particularly to a method and system for forming a distributed management form of a communication network in which the network configuration can be changed.

[0002]

2. Description of the Related Art Conventionally, in a communication network consisting of a node group and a link group connecting the node groups, a fixed method has been used to divide the network by forming a cluster composed of at least one adjacent node. There is known a network division method for performing intensive calculations on a communication network, that is, a communication network management mode. In this management mode, one management node is provided in the cluster.

Centralized calculations for this fixed communication network are (1) that the entire network can be grasped, (2) that the management node is fixed, and (3) that the division target is a node. It is assumed that there is. Therefore, the form of the communication network capable of performing the network division based on the above assumption is limited in various forms.

Grasp of the entire network means that the node that divides the network needs to grasp the configuration of the entire network in order to centrally divide the network. Therefore, it is difficult to introduce the above conventional management form in the process of constructing the network. Further, it is necessary to re-divide the entire network every time the network configuration is changed, such as expansion or degeneration of the network, and the cost required for the division is large.

The fixed management node means that the management node that manages or represents the divided network is treated as an independent dedicated node different from the nodes forming the network. Therefore, when the network configuration changes after arranging an appropriate number of management nodes at an appropriate location, the correspondence between the network division mode and the management nodes collapses, and it is necessary to provide a new management node, or The waste of the management node occurs such that the management node already introduced becomes unnecessary. On the other hand, in the case of a communication network in which the configuration can change, if the management node once introduced is used as it is, the network division itself becomes a form specialized in the number of management nodes and the location.
It becomes impossible to manage properly.

The fact that the object of division is a node means that the object of the conventional management form is the operation of an individual node. Therefore, the network division is determined by the relationship between the management target node and the management node such as the number of nodes included in the divided network and the communication distance from the management node. As described above, the conventional network partitioning is different from the network partitioning that inherently represents the characteristics of the network, that is, the network partitioning considering the connection form between the nodes, and therefore, a plurality of nodes exchanging a large amount of information. ,
That is, the node group may be managed by different management nodes. Therefore, when a node in the node group fails, it is necessary to cooperate with a plurality of management nodes to perform failure avoidance measures such as failure analysis or routing change.
Appropriate action may not be taken promptly.

Further, according to the conventional network division described above, it is implicitly assumed that only one specific carrier can manage the communication network. Therefore, it is difficult to incorporate a plurality of management policies (management measures) in consideration of the existence of a plurality of managers. In contrast to the above-mentioned conventional communication network division method, there is a bottom-up clustering method in a large-scale communication network in which each node has a management function and a call processing function, and delegates management authority while cooperating between nodes. Are known.
According to such a bottom-up division method, since the management area is determined bottom-up, it is not necessary to grasp the entire network, and it can be used even when the network configuration may change. Further, since the management node is not fixed but is dynamically determined, it is not necessary to provide a dedicated management node.

[0008]

However, in the case of bottom-up division management that improves the above-mentioned division method by intensive calculation, the connection relation between nodes is not taken into consideration in the logic for determining the management node. . Also,
Since only the management nodes are integrated with respect to the initial configuration of the network, it is difficult to dynamically change the management node in response to changes in the network configuration. It is not envisaged to incorporate more than one management policy.

Therefore, the conventional division management of a large-scale communication network has the following problems. First, in the conventional division management of a large-scale communication network, since the network division does not consider the connection form between nodes, it is difficult to improve reliability, failure analysis, and management efficiency. Second, it cannot be applied to a communication network in which the network configuration can be freely changed, and third, it is difficult to adopt a plurality of management policies.

In view of the above-mentioned problem of division management of a large-scale communication network in the related art, the present invention has a management function and a call processing function in each node, and the management authority is delegated to each other in cooperation with each other. In a bottom-up clustering method for large-scale communication networks, consider the connection form between nodes in order to improve reliability and improve failure analysis and management efficiency, and
An object of the present invention is to provide a communication network division method and division system that reflect a plurality of management policies.

A further object of the present invention is to provide a division method and division system for a communication network that can easily follow changes in the network configuration of the communication network.

[0012]

FIG. 1 is a diagram illustrating the principle of the present invention. A communication network dividing method according to the present invention, in a communication network including a plurality of nodes and a link connecting the plurality of nodes, recognizes a second node adjacent to a first node, Adjacency relation recognition step (step 1) in which information is exchanged between the above node and the above adjacent second node
And; a cluster forming step (step 2) in which the first node and the adjacent second node form a cluster.

In the communication network dividing method of the present invention, the adjacency relation recognizing step (step 1) is performed in order to realize the network division in which the connection form between the nodes is taken into consideration and a plurality of management policies are reflected. An adjacency information generating step (step 10) of generating adjacency information representing a degree of connection of the first node in the first cluster with an adjacent second cluster of the first cluster; Further comprising a subordination determining step (step 12) for determining an inter-cluster dependency setting rule to be applied to the applied adjacent information; the cluster forming step (step 2) is based on the applied subordination setting rule. Accordingly, a step of assigning one of the first cluster and the second cluster to a subordinate cluster and the other to a managing cluster (step 20)
And forming a new cluster in which the subordinate cluster is subordinate to the managing cluster (step 2).
2) consists of

Further, in the communication network dividing method of the present invention, when the new cluster formed in the cluster forming step (step 2) is used as a new first cluster to form another new cluster, the adjacent It further includes a relationship recognition step (step 1) and a cluster hierarchization step (step 3) for repeatedly executing the cluster formation step (step 2).

Further, the adjacent information generating step (step 1)
0) generates the adjacency information consisting of the adjacency representing the degree of connection relative to all the second clusters for each of the first nodes. As described above, the communication network dividing method of the present invention is based on the conventional bottom-up clustering method, and further, when forming clusters, by setting the dependency relationship based on the adjacent information of each cluster. It is characterized by forming a cluster for efficient failure analysis and management, and reflecting a plurality of management policies in the cluster formation.

According to the present invention, the dependency setting rule includes, for example, the cluster information of all adjacent clusters, and the adjacency determined from the capacity ratio and the distance ratio of the links connecting the clusters. A dependency setting rule based on adjacent information and restrictions and subordination rules at the time of cluster configuration based on each management policy are introduced.

In the communication network dividing method according to the present invention, the subordination determining step (step 12) applies the subordination setting rule according to a predetermined priority to obtain an application result. (Step 120)
And; an adoption dependency relationship determining step (step 122) for determining a dependency relationship to be adopted based on the application result and the dependency relationship setting application from the adjacent second cluster. As a result, based on the adjacency information held by the cluster itself, the application status of the dependency setting rule, and the status of the dependency setting application from another cluster,
You can select the dependency setting rule that best suits your current situation. Further, in the communication network dividing method according to the present invention, after the dependency relationship setting step (step 12) determines the dependency relationship setting rule, it is determined whether or not the adjacency information is changed, and the adjacency relationship is determined. If it is determined that it has changed, the method further includes a dependency re-determination step (step 124) of determining again the dependency setting rule between clusters applied to the changed adjacent information. Therefore, it is possible to perform the intended clustering on the connection relation that changes every moment in the clustering process.

Further, according to the division method of the communication network of the present invention,
Saving information about the lower layer cluster applied when the lower layer cluster assigned to the subordinate cluster in the cluster forming step (step 2) forms the new cluster (step 24).
And further, after the cluster is formed in the communication network, an adjacency information change detecting step of sequentially detecting a change in adjacency information of nodes of the cluster of the lower layer from the lowest layer, and the cluster of the lower layer by The first cluster further includes the adjacency relationship determining step and the cluster re-dividing step (step 4) for executing the cluster forming step.

In the cluster subdivision step, after the determination of the inter-cluster dependency setting rule which can apply the change to the detected adjacent information, the cluster sub-division step is performed based on the dependency relationship setting rule stored as the information about the lower-level cluster. Also determines whether a high priority dependency setting rule can be applied,
When it is determined that the dependency setting rule with high priority cannot be applied, the dependency setting rule stored as the information about the cluster in the lower layer is used as the dependency setting rule for the cluster in the lower layer in which the change is detected. It has a dependency confirmation step of determining whether it can be maintained. The cluster re-partitioning step can be applied, if it is determined that a dependency setting rule with a high priority can be applied, the clusters of the upper layer to which the cluster of the lower layer in which the change is detected belong If it is determined that the upper cluster formation stage is executed to form a new cluster and can be maintained by applying the high priority subordination setting rule defined as above, the layer next to the lower layer in which the above change is detected is detected. If it is determined that the cluster cannot be maintained by repeatedly executing the above dependency checking step on the clusters of, the clusters of the upper layer to which the cluster of the lower layer in which the change is detected belong are eliminated, and the adjacent information change detecting step is performed. Have a step back.

Due to the re-division step of the communication network division method of the present invention, the sub-clusters forming the cluster always retain the dependency relation setting rule and the adjacent information applied when forming the cluster, and the change of the adjacent information. Can be detected and the subordination relationship setting rule can be reapplied, so that changes in the network configuration can be easily followed. That is, each cluster can operate autonomously to constantly update the current state and adopt the optimal dependency.

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 2 is a diagram for explaining the configuration of a communication network to which the present invention is applied. As shown in the figure, the communication network 101 includes a node group 102 and a link group 103 that connects the nodes of the node group.

FIG. 3 is a diagram for explaining a cluster configuration of a communication network to which the communication network division system of one embodiment of the present invention is applied. A cluster is composed of at least one adjacent node, and each node corresponds to one cluster in the initial state before network division. FIG. 3 shows a clustering configuration in which clustering is performed halfway. In the figure, cluster 211 is from node 201 and node 202, cluster 212 is from node 203 and node 204, cluster 214 is from cluster 211 and cluster 212, cluster 213 is from node 206 and node 207, and cluster 215 is from node 205. , The cluster 216 is composed of nodes 208, respectively.

For example, the cluster 211 is the cluster 21
2 and the cluster 213 “adjacent” to the cluster 213. Cluster 212 is “adjacent” to clusters 211 and 215. As described above, “adjacent” indicates a state in which clusters are directly connected to each other via a link.

In the cluster 211, the node 201 is subordinate to the node 202. In this case, the node 202 is called a representative node representing the cluster 211. Similarly, the representative node of the cluster 212 is the node 203, and the representative node of the cluster 213 is the cluster 206.

The division processing can be assumed in both the node unit and the cluster unit, and the processing in the cluster unit is
By handling the cluster as a process for the representative node, it is possible to handle the node and the cluster in a unified manner. For example, the cluster 211 is an adjacent cluster, that is, the cluster 212 and the cluster 213.
In the case of exchanging information with the representative node 202 of the cluster 211, the representative node 202 of the cluster 211 exchanges information with the representative node 203 of the cluster 212 and the representative node 206 of the cluster 213 via a link in each cluster or a link connecting the clusters.

A plurality of clusters, for example, cluster 211
And the cluster 212 forms a new cluster 214, the representative node of the original cluster, that is, one of the nodes 202 and 203 becomes the representative node of the newly formed cluster 214. The cluster is composed of at least one node and the cluster, and includes one representative node. A new cluster is formed when the above-described dependency relationship occurs between the representative node of one cluster and the representative node of another cluster.
In the present invention, the representative node serves as a management cluster that manages the nodes belonging to the cluster and the cluster in this way, and a hierarchical management form is realized by setting the subordinate relationships between the representative nodes. When a new cluster is formed, the cluster including the representative node of the new cluster is called the management cluster, and the other clusters are called subordinate clusters. For example, in FIG.
Is the representative node of the new cluster, and the cluster 212
Is called a management cluster, and the cluster 211 is called a subordinate cluster.

FIG. 4 is a block diagram of a communication network division system according to an embodiment of the present invention. In the figure, the cluster 2 is adjacent to the adjacent cluster group 4 via the inter-adjacent cluster link 6, and is composed of a representative node 10, an intra-cluster node group 12, and an intra-cluster link group 14. Representative node 1
0 is one node selected from the in-cluster node group 12.

For example, the cluster 2 may be composed of only the representative node 10. In that case, other nodes in the cluster node group 12 and the intra-cluster link group 14
Does not exist. The representative node 10 which is one of the nodes in the cluster 12 has a group of links 14 in the cluster.
And the adjacent cluster group 4 via the link 6 between adjacent clusters
And an information transmitting / receiving unit 20 that exchanges adjacent information and cluster information with the adjacent cluster group 4; and a dependency setting rule unit 24 that manages the dependency setting rule applicable to the current cluster state. An application rule determination unit 22 that determines the dependency relationship setting rule of the unit 24; calculates the adjacent information of the cluster 2 itself, determines the current adjacent information and the applicable rule, the status of the dependency application obtained from another cluster, and the like. A status holding unit 26 for managing the current processing status; referring to the information managed by the status holding unit 26, reapplying the rules, updating the current status,
It includes a dependency relationship setting determination control unit 28 that controls the shift to the dependency relationship setting process; and a dependency relationship setting processing unit 30 that performs the dependency relationship setting process.

The subordination relation setting processing unit 30 performs subordinate processing when the cluster is a subordinate cluster subordinate to another cluster, and a management process which performs management processing when the cluster is a management cluster. Processing unit 34
Consists of FIG. 5 is a flow chart of the network division operation of the communication network division system of the embodiment of the present invention.

The information transmitting / receiving unit 20 recognizes the adjacent cluster group 4 and notifies the cluster information for identifying the cluster of the cluster 2 itself (step 200). Here, the cluster information includes, for example, a cluster name reflecting a hierarchical cluster structure so that the cluster 2 can be specified.

The state holding unit 26 collects information from all adjacent clusters via the information transmitting / receiving unit 20, calculates adjacent information, and records the calculated adjacent information as the current state (step 202). As the adjacent information, for example,
It is possible to use an adjacency list including a list of nodes forming the adjoining cluster group 4, and a capacity ratio (adjacency) of links to individual adjoining clusters with respect to the total capacity of the inter-adjacent cluster link group 6 included in the cluster 2.

In accordance with an instruction from the subordinate relationship setting judgment control unit 28, the application rule determining unit 22 causes the subordinate relationship setting rule unit 2 to operate.
The application of the rule No. 4 is tried, and the result is recorded in the state holding unit 26 (step 204). According to one embodiment of the present invention, the following dependency setting rules are applied for each adjacent cluster. The priority of rules is the order of rule 1, rule 2, and rule 3.

Rule 1: If the degree of adjacency ≧ α (0.5 ≦ α ≦ 1), the cluster is subordinate to the adjacent cluster. Rule 2: In the adjacent list of the cluster, the ratio of the elements included in the adjacent list of the adjacent cluster is β (0.
If 5 ≦ β ≦ 1) or more, a subordinate relationship is set between the cluster and the adjacent cluster.

Rule 3: In the adjacency list of the adjoining cluster, the adjacency of the matching portion is γ (0.5≤γ≤1)
If the above clusters exist, the cluster depends on any of the clusters. When applying the rules 2 and 3, the adjacent information such as the adjacent list of the adjacent cluster is acquired via the information transmitting / receiving unit 20.

The subordination relation setting judgment control unit 28 judges whether or not there is a change in the adjacency information newly obtained through the information transmitting / receiving unit 20 (step 206). If the adjacency state has changed, the step 202 Then, the adjacency information is calculated again, and the rule is applied again.

When the newly obtained adjacency information has no change, the subordination relation setting judgment control unit 28 determines whether the applicable rule judgment unit 22 is applicable.
In response to the result of the application of the dependency setting rule in the above, the state holding unit 26 is referred to, and the rule selected by the application rule determining unit 22 and the dependency setting application proposed from the adjacent cluster group 4 at that time Those having a high application priority are selected (step 208).

The subordination setting determination control unit 28 determines whether there is a rule that can be adopted (step 210), and if there is no applicable rule, holds the current state (step 212) and sets the adjacent cluster group. Wait for a change in 4 to occur (step 214). Adjacent cluster group 4
If there is an information notification from, the process returns to step 202.

If there is a rule that can be adopted, the dependency setting processing section 30 sets the dependency based on the adopted rule. It is determined whether the adopted rule is Rule 1 or 2, and in the case of Rule 1 or 2, the subordination processing unit 32 makes a subordination application, otherwise, makes a management application, and makes a management application or subordination from another cluster. In consideration of the application, it is determined whether the cluster is a management cluster or a subordinate cluster (step 216), and when it is determined that the subordinate cluster is the subordinate cluster, the subordinate processing unit 32 sets the subordinate relationship (step 218). ), The adjacent information held by the cluster 2 is transferred to the management cluster so as to form a new cluster (step 220), and the clustering process ends.

When it is determined that the cluster is the management cluster, the management processing unit 34 sets a subordinate relationship (step 222). Then, the management processing unit 34 collects the adjacency information notified from the subordinate clusters, creates the cluster configuration list and the adjacency information for the newly formed cluster in addition to the adjacency information of the cluster 2, and the state holding unit 26, and notifies the adjacent cluster of new cluster information for performing clustering with the adjacent cluster (step 224). It is determined whether the condition is an end condition (step 226), and if the condition is an end condition, the clustering is ended. If the termination condition is not met, step 20
Returning to 2, the clustering process is continued.

When the subordination application is made in step 216 and the subordination application is also received from the adjacent cluster group 4, for example, the cluster with a small adjacency is determined to be the management cluster. It can be determined which of the subordinate clusters it is.

FIG. 6 is obtained by giving an appropriate link capacity to the link group for the communication network 101 shown in FIG. 3 and executing the network division operation of the communication network division system of the embodiment of the present invention. It is a figure which shows an example of a communication network. By applying the embodiment of the present invention to the communication network 101 of FIG. 3, a hierarchical cluster based on the connection relationship as shown in FIG. 6 is formed, and a cluster of an arbitrary hierarchy is set as a unit based on the management policy. Can be divided into nets.

A modified example of the network division operation of the communication network division system of the embodiment of the present invention shown in FIG. 5 will be described below. The cluster information notified by the information transmitting / receiving unit 20 in the above step 200 includes the cluster name reflecting the hierarchical cluster structure for identifying the cluster, and the clustering source name designating whether or not the management form is the same. , And a cluster key that unconditionally specifies that the cluster is the same, and a representative node attribute that explicitly specifies that the node becomes the representative node.

Furthermore, as the rule 0 of the above-mentioned dependency setting rule, rule 0: When the same cluster key is used, unconditional clustering is performed regardless of the connection form. May be added.

When such rule 0 is adopted, in step 218, the subordinate processing unit 32 is activated and a subordinate application is made. Furthermore, when a representative node is designated for the representative node attribute of the cluster, it can be determined that the representative node is always the representative node. Therefore, according to the modification of the embodiment of the present invention, the dependency setting in the dependency setting rule applying step and the determination of the cluster of the cluster to be managed in the cluster forming step are performed at the time of cluster configuration based on the management policy. It is carried out based on the above-mentioned dependency setting rules including restrictions or subordination rules.

Next, a method of re-dividing a communication network according to another embodiment of the present invention will be described. In order to re-divide the communication network, it is necessary to hold the current state once it has been divided. Therefore, in the re-division method, the information when the new cluster is created in step 224 of the communication network division method according to the embodiment of the present invention described in FIG. 5 is recorded as the network division phase. Separately, reclustering phases 1 and 2 are provided.

FIG. 7 is a flow chart for explaining the re-division judging method in the re-division method of the communication network of another embodiment of the present invention. Step 300: When the information transmitter / receiver 20 is notified of the cluster information, the information transmitter / receiver 20 checks the change in the adjacent state of the cluster 2 in the current reclustering phase. Here, when the re-clustering phase is not set, it is processed as clustering for the hierarchy of each node. When there is a change in the adjacent state, the state holding unit 26 updates and records the adjacent information. The information updated and recorded here is the re-clustering phase 1
Is recorded separately from the information recorded in the normal network division stage. If there is no change in the adjacent state, the cluster to be re-divided is changed to a higher cluster and the processing is continued.

Step 302: The applicable rule judgment unit 22
It is checked whether a rule having a higher application priority than the dependency rule applied in the raster hierarchy of the current reclustering phase can be applied, and if it cannot be applied, the process proceeds to step 304. If higher rules apply,
Go to step 308.

Step 304: For new adjacent information,
Find out if existing dependencies can be maintained in the cluster hierarchy for the current reclustering phase. If it can be maintained, the process proceeds to step 306. If it cannot be maintained, the process proceeds to step 310. Step 306: Notify the upper cluster of the adjacent information, shift the upper cluster processing, and notify the cluster information of the upper layer to the adjacent cluster that has caused the change of the adjacent information.
The upper cluster performs the process of step 300 based on the adjacency information received from the sub cluster.

Step 308: Since the rule having the higher application priority is applied, the existing dependency relation is canceled and the process proceeds to step 314 so that the dependency relation is newly established. Step 310: Since the existing dependency cannot be maintained,
Proceed to step 312 to cancel the existing dependency and apply another rule.

Step 312: The process moves to the re-clustering phase 2 on the basis of the adjacency information updated in the re-clustering phase 1. Reclustering phase 2
Performs clustering based only on the adjacency information in the re-clustering phase 1 without referring to existing dependency relationships. The cluster information is updated, the cluster information is notified to the adjacent cluster, and the normal network division method shown in FIG. 5 is executed in the re-clustering phase 2. In the re-clustering phase 2, it is determined whether another rule can be applied, and if another rule is applicable, clustering is performed based on the rule. On the other hand, if another rule cannot be applied, after detecting a new change in adjacent information, the process returns to step 300 to execute the re-clustering phase 1.

Step 314: Based on the neighbor information updated in the re-clustering phase 1, the re-clustering phase 2 is entered. Reclustering phase 2
Performs clustering based only on the adjacency information in the re-clustering phase 1 without referring to existing dependency relationships. The cluster information is updated, the cluster information is notified to the adjacent cluster, and the normal network division method shown in FIG. 5 is executed in the re-clustering phase 2. Since it is predetermined that the subordinate relationship is set, the subordinate relationship is newly established. When the cluster in the reclustering phase 1 receives the dependency application from the reclustering phase 2, the process proceeds to step 350 described in FIG. 9.

FIG. 8 is a flow chart of the management cluster process when the subordination is eliminated in the subdivision method according to another embodiment of the present invention. The upper cluster is notified that the cluster has been resolved (step 330). Upon receiving the notification of cluster cancellation, the upper cluster notifies the sub-cluster managed in the hierarchy that the cluster has been cancelled, and the dependency with the sub-cluster is canceled, thereby canceling the cluster in the hierarchy (step 332). ).

The cluster which has received the message of cluster elimination executes the clustering process in the re-clustering phase 2 of step 312 (step 334) and returns to step 330. FIG. 9 is a flowchart of the processing in the reclustering phase 1 on the partner side when a subordination relation is newly established in the repartitioning method of another embodiment of the present invention.

The cluster that has received the application determines whether or not to receive the application for setting the dependency in the current re-cluster hierarchy in the dependency setting determination controller 28 (step 35).
0). If the priority is the same or lower than the existing dependency,
Reject the application and proceed to step 306 (step 35
2). If the application priority is higher than the rule in which the cluster itself sets the dependency relationship, the process proceeds to step 308 (step 354).

The change of the communication network obtained by applying the communication network re-partitioning method of another embodiment of the present invention will be described below with reference to FIGS. FIG. 10 is a diagram showing an example of a communication network obtained by the network division method according to the embodiment of the present invention, which is different from the communication network shown in FIG. 902 is added so as to be adjacent to the nodes 903 and 904. FIG. 11 is a diagram showing the progress of the cluster reconfiguration situation obtained when the repartitioning method according to another embodiment of the present invention shown in FIGS. 7 to 9 is applied to the communication network shown in FIG. . On the other hand, FIG. 12 is a diagram showing a final communication network obtained by applying the above-mentioned subdivision method.

First, in step 300 node 9
03, the node 904 checks the change of the adjacency information of the node itself by receiving the cluster information from the node 901.
The node 904 changes the neighboring information to the cluster 506.
, The cluster 506 is resolved in step 310. On the other hand, the node 903 cannot maintain the existing dependency due to the change of the adjacent information, and thus the cluster 504 is resolved in step 308.

In step 314, the node 901 establishes a subordinate relationship with the node 904 among the nodes 903 and 904 which are adjacent clusters. At this point, as shown in FIG. 11, the communication network 101 includes a cluster 1001 including nodes 904 and 901, a cluster 1002 including only nodes 903, and a cluster 501.
And a cluster 502.

Since the cluster 506 has been eliminated, the cluster 1001, the cluster 1002, and the cluster 501 are in the reclustering phase 2, and are included in the cluster 1001 and the cluster 505 that receives new cluster information from the cluster 501. The nodes are
In the re-clustering phase 1, steps 300 to 306 are repeatedly executed and the cluster 505 is maintained.

Next, the cluster 1002 becomes the cluster 1
The cluster information of 001 is received, the setting of a new dependency relationship with the cluster 1001 is determined, and the new cluster 1101 is set.
(Step 314). Further, the cluster 110
1 and the cluster 501 set a subordinate relationship, and the cluster 11
02 is configured (step 314).

Furthermore, according to the method of re-dividing a communication network according to another embodiment of the present invention, even when the network configuration is changed by changing the link capacity or deleting the link or the node, the adjacent information can be updated. It is possible to respond similarly as a trigger. The communication network 1 described in the above embodiment of the present invention
The 01 link group 103 may be replaced with, for example, a virtual path or a virtual channel of the ATM network.

[0061]

As described above, according to the present invention, the dependency relation setting rule in consideration of the connection form between the nodes is handled, the application priority is further set, and the dependency relation in which the connection form is better reflected. Is set, the network division that fully considers the connection form between the nodes is realized, the failure analysis and management of the communication network are made efficient, and the reliability is improved.

Further, by using the dependency relation setting rule reflecting the management policy of the manager together, a management mode in which the intention of the manager is reflected at the time of network division can be obtained. Furthermore, by making optimum use of the dependency setting rule according to the update of the adjacent information based on the existing dependency, there is an advantage that it is possible to follow the change of the network configuration only by the local correspondence.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an explanatory diagram of a configuration of a communication network to which the present invention is applied.

FIG. 3 is an explanatory diagram of a cluster configuration of a communication network to which the communication network dividing system according to the embodiment of the present invention is applied.

FIG. 4 is a configuration diagram of a communication network division system according to an embodiment of the present invention.

FIG. 5 is a flowchart of a network division operation of the communication network division system according to the exemplary embodiment of the present invention.

FIG. 6 is a diagram showing an example of a communication network obtained by performing a network division operation of an embodiment of the present invention by giving an appropriate link capacity to the communication network shown in FIG.

FIG. 7 is a flowchart illustrating a re-division determining method in a re-division method for a communication network according to another embodiment of the present invention.

FIG. 8 is a flowchart of a management cluster process when a subordination relationship is resolved according to another embodiment of the present invention.

FIG. 9 is a flowchart of processing in re-clustering phase 1 on the partner side when a subordination relationship is newly established in the re-division method according to another embodiment of the present invention.

FIG. 10 is a diagram showing an example of a communication network obtained by a network division method according to an embodiment of the present invention.

FIG. 11 illustrates a subdivision method according to another embodiment of the present invention.
FIG. 9 is a diagram showing the progress of the cluster reconfiguration status obtained when applied to the communication network shown in FIG.

FIG. 12 illustrates a subdivision method according to another embodiment of the present invention.
It is a figure which shows the final communication network obtained by applying to the communication network shown in FIG.

[Explanation of symbols]

2 Cluster 4 Adjacent Cluster Group 6 Adjacent Cluster Link 10 Representative Node 12 Cluster Node Group 14 Cluster Link Group 20 Information Transmission / Reception Section 22 Applicable Rule Judgment Section 24 Dependency Setting Rule Section 26 State Holding Section 28 Dependency Setting Judgment Control Section 30 subordination setting processing unit 32 subordination processing unit 34 management processing unit 101 communication network 102 node group 103 link group 201, 202, 203, 204, 205, 206, 2
07,208 nodes 211,212,213,214,215,216
cluster

Claims (14)

[Claims] 1. In a communication network including a plurality of nodes and a link connecting the plurality of nodes, a first node recognizes an adjacent second node, and the first node and the adjacent node are connected to each other. A method of dividing a communication network, comprising: an adjacency relation recognizing step in which a second node exchanges information, and a cluster forming step in which the first node and the adjacent second node form a cluster, The adjacency information recognizing step includes an adjacency information generating step of generating adjacency information representing a degree of connection of the first node in the first cluster with an adjacent second cluster of the first cluster; A dependency relationship determining step of determining a dependency relationship setting rule between clusters to be applied to the generated adjacency information, wherein the cluster forming step is performed in accordance with the applied dependency relationship setting rule. Assigning one of the one cluster and the second cluster to a subordinate cluster and assigning the other to a managing cluster; and the subordinate cluster forming a new cluster subordinate to the managing cluster. The division method of the communication network. 2. A cluster hierarchy in which the new cluster formed in the cluster forming step is used as a new first cluster, and the adjacency recognition step and the cluster forming step are repeatedly executed to form a new new cluster. 2. The method of dividing a communication network according to claim 1, further comprising a composing step. 3. The adjacency information generated in the adjacency information generating step has adjacency representing the degree of connection relative to all the second clusters for each of the first nodes. 2. The method of dividing a communication network according to 2. 4. The subordination determining step comprises: applying the subordination setting rule according to a predetermined priority to obtain an application result; the subordination setting rule applying step; the application result, and the adjacent second 4. The method of dividing a communication network according to claim 3, further comprising: an employment dependency determination step of determining a dependency to be adopted based on a dependency setting application from the cluster. 5. The cluster forming step further comprises the step of storing information about the lower layer clusters applied when the lower layer clusters assigned to the subordinate clusters form the new clusters. After the cluster is formed in the communication network, an adjacency information change detecting step of detecting a change in adjacency information of nodes of the cluster of the lower layer is sequentially executed from the lowest layer, and the cluster of the lower layer is set to the first layer. 4. The communication network partitioning method according to claim 3, further comprising a cluster re-partitioning step for executing the adjacency relationship determining step and the cluster forming step as the cluster of. 6. The subdivision setting step, wherein the subdivision setting step stores the subordination setting stored as information about the subordinate clusters after determining a subordination setting rule between clusters that can apply the change to the detected adjacent information. A dependency confirmation step of determining whether or not the rule can be maintained as a dependency setting rule applied to the cluster of the lower layer in which the change is detected; When it is determined that the dependency check step cannot be maintained by repeatedly executing the dependency check step for the cluster of the next layer, all the clusters of the upper layer to which the cluster of the lower layer in which the change is detected belong are eliminated, and the adjacent information change The method according to claim 5, further comprising the step of returning to the detecting step. 7. The subordination determining step comprises: applying the subordination setting rule according to a predetermined priority to obtain an application result; the subordination setting rule applying step; the application result, and the adjacent second 6. The division method of the communication network according to claim 5, further comprising: an adoption dependency relationship determining step of determining the dependency relationship to be adopted based on a dependency application setting application from the cluster. 8. The cluster subdivision step comprises, after the determination of an inter-cluster dependency setting rule that can apply the change to the detected adjacent information, the dependency setting stored as information about the lower-level cluster. If it is determined whether a dependency setting rule with a higher priority than the rule can be applied and it is determined that a dependency setting rule with a higher priority cannot be applied, the above-mentioned dependency stored as information about the cluster in the lower hierarchy A dependency confirmation step of determining whether or not the relationship setting rule can be maintained as a dependency setting rule for a cluster of a lower hierarchy in which the above change is detected; and when it is determined that a higher priority dependency setting rule can be applied, All the clusters in the upper layer to which the clusters in the lower layer in which the above change is detected belong, are resolved, and the priority level If it is determined that the upper cluster formation step is performed to form a new cluster and can be maintained, the above dependency is applied to the cluster of the next lower layer of the lower layer where the above change is detected. When it is determined that the relationship cannot be maintained by repeatedly executing the relationship confirmation step, the step of further canceling all clusters of the upper layer to which the cluster of the lower layer in which the change is detected belongs and returning to the adjacent information change detection step is further included. The method of dividing a communication network according to claim 7. 9. The subordination determining step, after determining the subordination setting rule, determines whether or not the adjacency information has changed. If it is determined that the adjacency has changed, the subordination determining step includes: 8. The method of dividing a communication network according to claim 4, further comprising a dependency re-determination step of determining again a dependency relation setting rule between clusters applied to changed adjacent information. 10. Dependency setting in the step of applying the dependency setting rule and determination of the cluster to be managed in the cluster forming step include the dependency setting including a restriction or a dependency rule at the time of cluster configuration according to a management policy. The method for dividing a communication network according to claim 4 or 7, which is performed based on a rule. 11. If the dependent relationship that can be adopted is not obtained in the adoption dependent relationship determination step, the adjacent second
8. The method of dividing a communication network according to claim 4, further comprising the step of calculating the adjacency information based on the information notified according to the change of the cluster, and executing the subordination determining step again. 12. In a communication network comprising a plurality of nodes and a logically set virtual path connecting the plurality of nodes, a first node recognizes an adjacent second node, A communication network comprising an adjacency relation recognizing step in which a first node and the adjacent second node exchange information, and a cluster forming step in which the first node and the adjacent second node form a cluster. And a step of recognizing the adjacency relationship, wherein the adjacency recognition step generates adjacency information indicating a degree of connection of the first node in the first cluster with an adjacent second cluster of the first cluster. And a dependency determination step of determining a dependency setting rule between clusters to be applied to the generated neighboring information, wherein the cluster formation step includes the applied dependency relationship setting step. According to a fixed rule, assigning one of the first cluster and the second cluster to a subordinate cluster and assigning the other to a managing cluster, and the subordinate cluster depends on the managing cluster. A method of dividing a communication network, which comprises the steps of forming a new cluster. 13. In a communication network comprising a plurality of nodes and a logically set virtual channel connecting the plurality of nodes, a first node recognizes an adjacent second node, A communication network comprising an adjacency relation recognizing step in which a first node and the adjacent second node exchange information, and a cluster forming step in which the first node and the adjacent second node form a cluster. And a step of recognizing the adjacency relationship, wherein the adjacency recognition step generates adjacency information indicating a degree of connection of the first node in the first cluster with an adjacent second cluster of the first cluster. And a dependency determining step for determining a dependency setting rule between clusters to be applied to the generated neighboring information, wherein the cluster forming step includes the applied dependency relationship. Assigning one of the first cluster and the second cluster to a subordinate cluster and assigning the other to a managing cluster according to the assignment setting rule; and the subordinate cluster depends on the managing cluster. And a step of forming a new cluster, which divides the communication network. 14. A cluster, which is adjacent to an adjacent cluster group via an inter-adjacent cluster link and is composed of an in-cluster node group, a representative node selected from the in-cluster node group, and an in-cluster link group. A division system of a communication network, wherein the representative node recognizes the adjacent cluster group through the intra-cluster link group and the adjacent inter-cluster link, and outputs adjacent information and cluster information to and from the adjacent cluster group. An information transmitting / receiving unit that sends and receives information, an application rule determining unit that manages the dependency setting rule and determines an applicable dependency setting rule regarding the current cluster state, and calculates the adjacency information of the cluster itself and calculates the current adjacency. A status holding unit that manages the current processing status such as information and applicable rules, the status of dependency applications obtained from other clusters, and dependency setting The dependency relationship setting processing unit that performs processing, and the dependency relationship setting determination control unit that controls reapplication of rules, update of the current state, and transition to the dependency relationship setting processing by referring to the information managed by the state holding unit. A communication network division system consisting of and.