JP2000324127A - Bypass selection method and system, fault recovery method and system, node and network system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten the time to select a bypass. SOLUTION: A path end node receiving a notice of a bypass object discriminates to which processing group a path for retrieving the bypass belongs in the case of selecting the bypass. After the decision, costs of paths calculated as to each bypass object are sequentially ranked on the bases of a selection criterion specific to the processing group decided to which each path belongs so as to decide the selection ranking of the bypass objects.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a detour path selection method for realizing detour path selection and a fault recovery method for fault recovery. In addition, the present invention relates to an apparatus for implementing these methods, a node incorporating the apparatus, and a network system including the nodes.

[0002]

2. Description of the Related Art A network system constructed today has a failure recovery function (self-healing function) for autonomously and decentralized selection of a detour path when a failure occurs. For such a function, it has been proposed to select a detour path by a flooding / ringing process. The outline will be described with reference to the drawings.

In this case, as shown in FIG. 2, a path 01 (shown by a solid line in the figure) between a source node (sender) N0 and a destination node (chooser) N5 passes through a route (nodes N0-N1) shown in the figure. −N2−N5).

In this state, for example, as shown in FIG.
When a failure occurs in the link L3, the nodes N1 and N2 located at both ends of the link L3 send a failure notification. The failure notification is transferred on the network and reaches the sender and the chooser of the path 01 having the link L3 on the transmission path.

[0005] Upon receiving the failure notification, the sender N0 sends out a flooding message to all the links L1 and L2 connected to its own node to search for a detour route. After confirming that the flooding message for the same path has not been received, the node that has received the flooding message adds the ID of the own node to the received flooding message, and adds all the normal nodes except the receiving link. Send to link.

In the case of FIG. 3, the node N 1
1 and the flooding message is transferred to the link L4 except the failed link L3. Similarly, the node N4 transfers the flooding message to the links L5 and L7 except the receiving link L4. Through such a transfer operation, the flooding message reaches the chooser N5.

[0007] The chooser N5 that has received the flooding message obtains connectable bypass route information from the transfer route information recorded in the flooding message.

[0008] In the case of FIG. 3, the chooser N5 determines from the transfer path of the received flooding message the path N0-
The N1-N4-N5 and the route N0-N3-N4-N5 are recognized as candidates for the detour route.

As described above, when the chooser N5 obtains a plurality of alternative route candidates, it selects an appropriate route from the candidates and transmits a ringing message addressed to the sender N0 to the route. Nodes located on this detour route
When transferring a ringing message, necessary resources (bandwidth, etc.) are secured. Then, when the ringing message is transferred to the sender, a bypass route is established.

Documents disclosing this kind of technology include, for example,
Japanese Patent Application Laid-Open No. Hei 8-251161 discloses a “self-healing failure recovery method”. This document discloses a method in which the self-healing process is divided into a search phase, a response phase, a securing phase, and a confirmation phase.

In the search phase, the search signal is transferred from the sender to the chooser, and the amount of spare resources recorded in the search signal is updated as needed. That is, a search is made for a route candidate and a reserve resource amount available for the route. In the response phase, a response signal is transferred from the chooser to the sender. As a result, the amount of spare resources that can be used on the bypass route is notified.

In the securing phase, the securing signal is transferred from the sender to the chooser when the available spare resource amount of the route candidate notified by the response signal is larger than the resource amount required by the path. At this time, the resource amount required by the path is secured from the spare resource amount. In the confirmation phase, a confirmation signal is transferred from the chooser to the sender. As a result, the required resource is notified and the path is switched.

[0013]

However, the prior art has the following problems.

For example, when a bandwidth is secured at the time of transferring a flooding message, a plurality of different routes on the same path overlap to secure a bandwidth, and another faulty path is used until an extra bandwidth is released. A state occurs in which the band cannot be used. For this reason, extra time is required to recover other failed paths.

On the other hand, when the band of the passing link is secured at the time of transferring the ringing message, even if the required bandwidth is surplus at the time of passing the flooding message, the bandwidth is already secured by another path at the time of transferring the ringing message. May not be usable. In this case, it is necessary to select another route candidate to try to establish a detour route, and the time required for recovering the failed path becomes longer.

Furthermore, even when a bypass route is selected in consideration of the route cost, for example, in a method of preferentially selecting a route having a large free band and establishing a bypass route, the bypass route is set to a link having a large free band. It is inevitable that requests for establishing a route are concentrated.

For example, if bandwidths are allocated in the order in which the establishment request arrives earlier, there may be a case where a free path of a link becomes insufficient for a logical path (VP) in which the establishment request arrives later, and a bypass route cannot be established. Can occur. In this case, for a logical path (VP) for which a detour path could not be established, it is necessary to separately search for a new detour path, and extra time is required for path recovery.

In addition, in this method, the larger the number of hops of the route, the higher the possibility that the request for establishing the detour route will arrive late, so that the logical path (VP) having the larger number of hops of the route
With regard to, there is a high possibility that the available bandwidth of the link becomes insufficient. In addition, the logical path (V
If P) decides to find another detour, it takes more time to recover the path.

Further, in the conventional route selection processing based on distributed processing, since it is not known which path is selected by another path, a method of appropriately selecting a path and performing a fault recovery (for example, first, (A method of selecting a route through which an incoming flooding message has passed as a detour route), and the path recovery rate is liable to be greatly deteriorated as compared with an originally possible value.

[0020]

According to a first aspect of the present invention, when a bypass route is selected, the path end node that has received the notification of the bypass route candidate determines which one of the paths to search for the bypass route. Judge whether it belongs to the processing group of
After the determination, the route cost calculated for each detour path candidate is ranked based on a selection criterion specific to the processing group to which the path belongs, and the detour route candidate selection order is determined.

In the second invention, at the time of selection of a detour path, the path end node which has been notified of the detour path candidate,
It is determined to which processing group the path for which the detour path is searched belongs, and after that determination, the path cost of each detour path candidate calculated by a calculation method unique to the processing group determined to belong to the path is The order is determined based on a predetermined selection criterion, and the selection order of the alternative route candidates is determined.

Further, in the third invention, at the time of selecting a detour path, the path end node which has been notified of the detour path candidate,
It is determined to which processing group the path for which the detour path is searched belongs, and after that determination, the path cost notified for each of the detour path candidates is selected specifically for the processing group to which the path belongs. The order is determined based on the criterion, and the selection order of the alternative route candidates is determined.

In the fourth invention, when a detour path is selected, the path end node that has received the notification of the detour path candidate,
It is determined which processing group the path for which the detour route is searched belongs to, and after that determination, a plurality of route costs calculated by a calculation method specific to each processing group, which are notified for each of the detour route candidates. Among them, those of the processing group determined to belong to the path are ranked based on a predetermined selection criterion, and the selection order of the alternative route candidates is determined.

As described above, in the first to fourth aspects of the present invention, the path cost calculation reference or selection reference is switched according to which processing group the path for which the detour path is searched belongs. A detour route can be selected by a method according to each path. Thus, it is possible to reduce the time required for selecting a detour route.

In the fifth invention, (1) notification processing for notifying a failure that has occurred on an existing path, and (2) a detour path to the existing path for which a failure has been notified and the possibility of securing resources thereof. A search process for searching; and (3) as a result of the search process, if there is a bypass route that can secure resources, secure that resource and switch the existing route to a bypass route.
(4) As a result of the search processing, if there is no alternative route that can secure resources, the other path that uses the resource of the location that failed to secure resources is ordered to bypass the route, and resources corresponding to that route are allocated. After that, the detour route passing through the location is provided with a rearrangement process for switching the route of the existing path notified of the failure.

As described above, in the fifth invention, even when there is no detour path capable of securing resources, other paths using the resources of the location where the resources could not be secured cannot be used. When the detour is possible, the route of the other path is detoured, and the free resources generated by the detour are
Since it can be assigned to the detour path of the target path, the possibility of failure recovery can be further improved.

[0027]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (A) Relationship between Embodiments Prior to description of a node and a network system to which a detour route selection device and a failure recovery device (or method) according to the present invention are applied, a relationship between each embodiment. Will be described.

(A-1) First and Second Embodiments Both the first and second embodiments relate to a detour path selection technique that is executed when a plurality of detour path candidates are obtained. And introduces the concept of path cost into the determination method.

Here, the difference between the first embodiment and the second embodiment is the difference in the node that executes the calculation of the path cost.

That is, in the first embodiment, the chooser calculates the path cost, whereas in the second embodiment, each node located on the transfer path calculates the intermediate value of the path cost at the time of transfer. They differ in that they are calculated sequentially.

(A-2) Third and Fourth Embodiments In both the third and fourth embodiments, the use of resources used by other paths is taken into account in the failure recovery fixed processing by distributed processing. This is to introduce the concept of the path rearrangement process.

Here, the difference between the third embodiment and the fourth embodiment lies in the contents of the route rearrangement processing.

That is, in the third embodiment, the route rearrangement process is executed for any one route selected from a plurality of route candidates, whereas in the fourth embodiment, a bottleneck location described later is processed. The point is that the route rearrangement process is executed at once for a plurality of route candidates that share the same route.

(B) Node Configuration Common to Each Embodiment In each embodiment, a node having a functional block configuration shown in FIG. 4 is used. Here, the node has a basic configuration including a switch unit 1 for exchanging lines or packets, a control unit 2 for controlling the switch unit 1, and inter-node interfaces 3 and 4.

The control unit 2 is a functional block unit for implementing a detour path selection function or a failure recovery function described below. It is assumed that the control unit 2 includes a processor and an internal or external memory. That is,
According to a program stored in the memory, the selection of a detour path and the setting of a fault are realized by software. However, the realization method is not limited to software processing, and may be realized by hardware.

It is assumed that the functions mounted on each node are set in advance when designing the system.

(C) First Embodiment (C-1) Network Configuration FIG. 5 shows an example of a network to which a node according to the first embodiment is applied. In the figure, N1 to N5 are nodes, and L1 to L6 are links. FIG. 1 is an example of a network in which five nodes are connected by six links. Of course, the connection configuration is not limited to this.

Further, in the figure, a set of numerical values (i, j) given next to each link represents "free bandwidth" and "cost" of each link. That is, the numerical value i is “free band”
, And the numerical value j represents “cost”. Of course, the set of numerical values to be assigned is not limited to this.

Here, the cost is a value determined for each link. For example, a value considering the installation cost and communication quality is used. For example, a numerical value proportional to the link distance or a numerical value according to the size of the remaining band is provided.
In FIG. 5, the larger the link distance, the greater the value.

In the following description, a node having a smaller node number among nodes terminating a path is called a sender, and a node having a larger node number is called a chooser. Nodes on other route candidates are referred to as relay nodes.

(C-2) Detour Path Selection Function Mounted on Each Node FIG. 6 shows an outline of how communication processing based on the detour path selection function according to the present embodiment proceeds. FIGS. 1 and 7 show the state of the processing procedure executed by each node when a failure is detected and each message is received. By the way, FIG.
Is a processing procedure unique to the present embodiment executed by the chooser. FIG. 7 shows a processing procedure executed by the sender.

(C-2-1) Detection of Failure All nodes located on the network monitor whether a failure has occurred in a link adjacent to the own node.
For example, in the case of the node N2 in FIG. 5, the links L1 and L2
And three links L6 and L6. If a failure is detected, the node sends a failure notification message to all links except the failed link. For example, if a failure occurs in the link L6 in FIG.
Send out a failure notification message to all links except the failed link L6.

Here, the failure notification message is added with the failure link number and the number of the node sending the message.

(C-2-2) Reception of Failure Notification Message (i) Sender Processing The failure notification message is transferred between nodes on the network, and eventually reaches a sender having a path on the network. Upon receiving the failure notification message, the sender activates a program that performs the processing according to the processing procedure shown in FIG. 7, and determines whether the notified failure location is on the path used by the own node. Is determined (step B1).

Here, if it is determined that the fault is not on the path of the own node, the sender ends the program and returns to the normal operation. On the other hand, if it is determined that the failure is on the path of the own node, the sender sends a flooding message to all normal connection links (step SB2).

Here, the flooding message includes:
For example, sender information, chooser information, path ID information, required resource (band) information, maximum hop number information, and the like are written.

(Ii) Process of the chooser Similarly, the failure notification message is transferred between nodes on the network, and eventually reaches a chooser having a path on the network. Upon receiving the failure notification message, the chooser activates a program that performs the processing according to the processing procedure illustrated in FIG. 1 and checks whether the notified failure location is on the path used by the own node. It is determined, and if it exists, a timer is started (step A1).

Next, the chooser monitors the timer,
The monitoring of the time is continued until the specified time has elapsed (step A2). If no later-described flooding message is received during this period, the processing of this program is terminated (negative result in step A3).

(C-2-3) Reception of Flooding Message (i) Processing of Relay Node Upon receiving the flooding message, the relay node confirms that the route is not in a loop. That is, it confirms that the message forwarded by its own node is not a return message. If the route is not looped, the relay node transfers the message to all normal connected links except the receiving link. On the other hand, if the route is in a loop, the message is discarded.

The relay node writes the passing route information and the passing hop number information when transferring the flooding message.

(Ii) Processing of the chooser When the timer has passed the specified time, the chooser executes 1
Alternatively, when a plurality of flooding messages have been received (that is, when there are alternative route candidates), a process of rearranging the alternative route candidates based on the route cost is performed (steps A4 to A7). More specifically, the routes are rearranged in ascending order.

Then, the chooser sends ringing messages to the sender in order from the one with the lowest path cost, and tries to establish a path (step A9). Then, if the establishment of the route is confirmed, the chooser ends the process, and if the establishment of the route fails, selects the next alternative route candidate and attempts to establish the route (step A10). When there are no more detour route candidates that can be established, the process ends (step A8).

Hereinafter, a method of ranking the detour paths will be described.

First, upon receiving the flooding message, the chooser records the information written in the message. Next, the chooser compares the smallest number of hops (shortest hops) among the alternative route candidates received at the time when the specified time has elapsed with the threshold value Z, and becomes a processing target. It is determined whether the current path route is a short hop route or a long hop route (step A).
4).

The reason for making such a determination is to switch the route cost calculation method between a route having a large number of hops and a route having a small number of hops. Although the details will be described later, the bandwidth is likely to be insufficient on a route with a large number of hops that requires time to transfer the setting information. Because you think it should be.

Here, the chooser determines a path having a minimum hop number smaller than the threshold value Z as a short hop path and a path having a minimum hop number equal to or larger than the threshold value Z as a long hop path.

In this embodiment, the threshold value Z is set as follows. Here, the number of all logical paths (VPs) set in the network is X,
When the total number of hops of these logical paths (VPs) is Y,
A value given by Z = Y / X is defined as a threshold value Z. With such a definition, it is possible to determine the relative positional relationship of the bypass route to be processed. Of course, this definition is an example, and another definition may be used.
Further, the threshold value Z can be determined in advance before a failure occurs.

When the chooser determines that the path to be processed is a short-hop path by the determination in step A4, the chooser obtains an affirmative result and shifts to a path cost calculation process based on algorithm 1 (step S4). A
5).

In algorithm 1, the amount of free bandwidth of link A on the route is CAPA (A), and the cost of link A is CO
When ST (A) is set, for each link passed, CAP
The value of A (A) / COST (A) is calculated, and the sum of all links on the route is defined as a route cost CO. That is, the chooser sets the route cost of each route candidate as CO =
Σ (CAPA (i) / COST (i)).

The chooser ranks the route candidates in ascending order of the value of the route cost CO. This means that, for example, assuming that the total cost is constant, a link having a relatively small free band of the cost of each link is prioritized. However, when there is a route having the same value of the route cost CO, the order of the route having a small hop number may be lowered. If there is a route having the same hop number as the route cost CO, the order of the route where the flooding message arrived earlier may be lowered.

On the other hand, if the chooser determines that the path to be processed is a long hop route by the determination in step A4, the chooser obtains a negative result and shifts to a route cost calculation process based on algorithm 2 (step A4). A
6).

In the algorithm 2, the free bandwidth of the link A on the route is defined as CAPA (A), and the cost of the link A is calculated as CO
When ST (A) is set, for each link passed, the COS
The value of T (A) / CAPA (A) is calculated, and the sum of all links on the route is defined as a route cost CO. That is, the chooser determines that the route cost of each route candidate is CO = Σ
(COST (i) / CAPA (i)).

Then, the chooser ranks the route candidates in the order of the smaller value of the route cost CO. This means that, for example, assuming that the total cost is constant, a link having a relatively large free band of the cost of each link is prioritized. However, when there is a route having the same value of the route cost CO, the order of the route having a small hop number may be lowered. If there is a route having the same hop number as the route cost CO, the order of the route where the flooding message arrived earlier may be lowered.

A specific example will be described. For example, in FIG. 5, if the value of the threshold Z is 1.4, the paths N1-N4
Since the shortest hop number of the route is two hops, the path N1
-N4 is a long hop path. Here, the path N1-N4
Among the route candidates, the cost of the route 1 (N1-N2-N4) is 2.67 (= 0.67 + 2). Route 2
The cost of (N1-N5-N4) is 3 (= 2 + 1). Then, the cost of the route 3 (N1-N2-N3-N4) is 5.17 (= 0.67 + 1.5 + 3). As a result, the order is route 1, route 2, and route 3 in ascending order.

After the ranking, the chooser records the routes in the order of the route costs, and proceeds to the next processing (step A).
7). Here, the chooser determines whether there is a detour path candidate that has not sent a ringing message (step A8). Unless a negative result is obtained in this determination process, the chooser selects a path in order from the path candidate with the lowest rank and returns a ringing message to the path (step A9).

In the ringing message, for example, sender information, chooser information, path ID information, required resource (band) information, maximum hop number information, passing route information (passing node information), passing hop number information, etc. are written.
Of these pieces of information, the passing hop number information is written by the relay node, and the other information is written by the chooser.

By the way, when transmitting a ringing message, the chooser executes a process for securing a required band on the transmission link. However, if the available bandwidth of the transmission link is smaller than the required bandwidth, the next path is selected.

(C-2-4) Reception of Ringing Message (i) Processing of Relay Node Upon receiving the ringing message sent by the chooser, the relay node sends the message to the next node described in the message. Forward. At this time, if the available bandwidth of the transmission link is equal to or larger than the required bandwidth, the relay node secures the required bandwidth. However, if the available bandwidth of the transmission link is insufficient, the relay node does not transfer the ringing message to the next node, and returns a NACK of the ringing message to the reception link.

The node receiving the NACK of the ringing message releases the reserved band. The NACK of the ringing message is forwarded to the chooser. The chooser receiving the NACK of the ringing message,
In step A10, a negative result is obtained, and a path having the next lowest path cost is selected, and a ringing message is transmitted.

(Ii) Processing of Sender When the sender receives the ringing message (FIG. 7)
Step B3), the ACK for the ringing message is returned to the chooser side (Step B4).
Then, the detour route is set and the path is switched, and the process ends (step B5).

If the ringing message is not received within a predetermined time, the sender determines that the setting of the detour route has failed, and ends the process.

(C-2-5) A of ringing message
ACK reception The ACK of the ringing message sent by the sender is
The data is transferred to the chooser via the set detour.
By receiving the ACK of the ringing message, the chooser confirms that the bandwidth of the detour path has been successfully secured (step A10 in FIG. 1), sets the path as the detour path, and executes path switching. Thus, a series of processing is completed (step A11).

If the securing of the band has failed for all the alternative route candidates (if a negative result is obtained in step A8), the chooser ends the process.

(C-3) Effects Obtained by the First Embodiment As described above, in the case of the detour route selection device (method) according to the first embodiment, the route having the smallest route cost is selected. By allocating the detour paths in order, it is possible to alleviate the problem that the path connection requests are concentrated on the link having a large free band.

That is, when calculating the route cost, the route cost calculation criteria are different for the short hop route and the long hop route, so that the route candidates selected for each path are dispersed, and Can be solved.

Furthermore, since the short hop route preferentially selects a route having a relatively small free band, there is a high possibility that another route will be searched due to contention. However, since the number of hops is small, a detour route for each time is required. The time required for selection is short, and the total search time is not long.

On the other hand, since the long hop route preferentially selects a route having a relatively large free band, the effect that the probability of successfully establishing the route is relatively high is obtained.

Thus, the effect of increasing the overall probability of saving more paths by combining the short hop route and the long hop route is obtained.

(D) Second Embodiment Next, a detour route selecting device (method) according to a second embodiment will be described. In this embodiment,
The operation will be described with reference to the network shown in FIG.

(D-1) Detour route selection function mounted on each node The difference between this embodiment and the first embodiment is that the route cost is calculated and updated sequentially on the flood message transfer route instead of the chooser. It is a point. Therefore, the differences between the processing procedures will be described below with reference to FIGS. 8 shows a processing procedure executed by the sender, FIG. 9 shows a processing procedure executed by the relay node, and FIG. 10 shows a processing procedure executed by the chooser.

(D-1-1) Detection of Failure This link failure detection operation is the same as that of the first embodiment. In other words, all nodes located on the network constantly monitor the occurrence of a failure, and when a failure is detected, send a failure notification message to all links except the failed link.

At this time, the failure link number and the number of the node sending the message are added to the failure notification message.

(D-1-2) Reception of Failure Notification Message (i) Processing of Sender The contents of the processing procedure executed by the sender having received the failure notification message are the same as those in the first embodiment. That is, as shown in FIG. 8, it is determined whether the notified fault location exists on the path route used by the own node (step B1). If the fault is not on the own node's path route, the program is terminated, and if the fault is on the own node's path route, a flooding message is transmitted (step B2 ').

Here, the sender writes, for example, sender information, chooser information, path ID information, required resource (band) information, maximum hop number information, route cost information, and the like in the flooding message.

The difference from the first embodiment is that at the time of sending this flood message, two types of path costs are calculated and written in the message. The two types of route costs referred to here are a short hop route cost and a long hop route cost.

Here, the sender determines the amount of available bandwidth of the transmission link A as CAPA (A) and the cost of the transmission link as COS (COS).
When T (A), CAPA (A) / COST (A)
Is calculated as the short-hop path cost.

Similarly, the sender determines the amount of free bandwidth of transmission link A as CAPA (A) and the cost of transmission link A as CO
When ST (A), COST (A) / CAPA
The value given in (A) is calculated as the long hop path cost.

[0111] In addition, 1 is added to the passing hop number information.

This processing will be described with a specific example. The following description is made with respect to the paths N1-N4 (routes N1-N2-N4) on the network shown in FIG.

In this example, the vacant band of the transmission link L1 is 3, and the cost is 2. Therefore, the sender N1 writes the value 1.5 of the short hop path cost in the short hop message area of the flood message, writes the value 0.67 of the long hop path cost in the long hop message area, and further writes the passing node information. With link L1
Send a flood message.

(Ii) Processing of the chooser The contents of the processing procedure executed by the chooser that has received the failure notification message are the same as those in the first embodiment. That is, the chooser that has received the failure notification message is as shown in FIG.
Start a program that performs processing according to the processing procedure shown in FIG.
It is determined whether the notified fault location is on the path used by the own node, and if so, the timer is started (step D1).

Next, the chooser monitors the timer,
The monitoring of the time is continued until the specified time has elapsed (step D2). If no later-described flooding message is received during this time, the processing of this program is terminated (a negative result in step D3).

(D-1-3) Reception of Flooding Message (i) Processing of Relay Node Upon receiving the flooding message, the relay node activates the program shown in FIG. 9 and executes the following processing.

First, the relay node confirms that the route is not in a loop (step C1). That is, it confirms that the message previously transferred by the own node is not a feedback message. If the route is in a loop, the relay node discards the message and ends the program.

On the other hand, if the route is not a loop,
The relay node proceeds to step C2, and calculates a path cost for each connection link serving as a transfer destination. Here, too, the relay node calculates two types of route costs for each connection link.

The relay node determines the amount of free bandwidth of the transmission link A as CAPA (A),
If the cost of the transmission link is COST (A), CAP
Calculate the value given by A (A) / COST (A) and rewrite it to the value obtained by adding the calculated value to the value of the short-hop path cost written in the flooding message (step C3).

Further, the relay node sets the vacant bandwidth amount of the transmission link A to CAPA (A) as a long hop route cost.
If the cost of the transmission link A is COST (A), CO
Calculate the value given by ST (A) / CAPA (A),
The calculated value is rewritten to the value of the long hop path cost written in the flooding message (step C3).

Thereafter, the relay node transfers the message to all normal connection links except the reception link. However, when the available bandwidth of the transmission link is smaller than the required bandwidth of the path, the flooding message is not transferred to the transmission link.

The relay node adds the passing route information (passing node information and passing link free band information) and rewrites the passing hop number information when transferring the flooding message. Here, the passing node information includes
Add the identifier information of the own node. In addition, 1 is added to the passing hop number information.

This processing will be described with a specific example. The following description is performed on the path N1-N4 (route N1-N2-N4) on the network shown in FIG. 5, as in the description of the sender.

In the case of this example, the relay node is the node N2. The relay node N2 that has received the flooding message sets the value of the short-hop route cost to 2 (= 1.5 +) because the available bandwidth of the transmission link L6 is 1 and the cost is 2.
0.5). Further, the relay node N2 rewrites the value of the long hop route cost to 2.67 (= 0.67 + 2). Thereafter, the relay node writes the information and the passing node information in the flooding message, and sends out the flooding message to the link L6.

(Ii) Processing of the chooser Each time a flooding message is received, the chooser confirms that the chooser of the flooding message is its own node, and passes through the passing node information, the short hop path cost value, and the long hop path cost value. Is recorded.

When it is confirmed that the specified time has elapsed since the start of the timer (when a positive result is obtained in step D2), the chooser checks whether there is a flooding message addressed to the own node which arrived within the specified time. That is, it is determined whether a detour path candidate exists (step D3 in FIG. 10).

If the existence of the alternative route candidate is not confirmed, the chooser terminates the processing program for selecting the alternative route as described above. On the other hand, when the presence of the alternative route candidate is confirmed, the chooser determines whether the minimum hop number of the path is larger or smaller than the threshold (step D4).

For this determination, the threshold value Z is used as in the first embodiment. Here, the threshold value Z is Z = Y / X, where X is the number of all logical paths (VPs) set in the network and Y is the total number of hops of the logical paths (VPs). Use the given value. This threshold value Z may be determined before a failure occurs.

As a result of the determination, if it is determined that the minimum hop number of the path is less than the threshold value Z, that is, the path is a short hop path, the chooser sorts the stored alternative path candidates in ascending order of the short hop path cost. Ranking is performed (step D5).

On the other hand, when it is determined that the minimum hop number of the path is equal to or larger than the threshold value Z, that is, the path is a long hop path, the chooser sorts the stored detour path candidates in ascending order of the long hop path cost. Ranking is performed (step D6).

In the above description, it has been described that after a specified time, it is determined whether the route is a short hop route or a long hop route. However, the ranking may be performed as follows.

That is, when the flooding message is received, the judgment is made, and when the number of hops of the route is less than the threshold value Z, the route candidates are ranked based on the short-hop route cost. If the number of hops is equal to or greater than the threshold value Z, if both short hop route costs and long hop route costs have not been determined before, record both the short hop route cost and the long hop route cost. If the number of hops of the route candidate is equal to or greater than the threshold value Z, the order may be determined according to the long hop route cost.

In any case, the chooser ranks the route candidates based on the route information written in the received flood message.

The above processing will be described with a specific example. In the following description, the path N1-N on the network shown in FIG.
For route 4, route 1 (N1-N2-N4), route 2 (N
1-N5-N4), Route 3 (N1-N2-N3-N4)
Is recognized by the chooser as a bypass route candidate. Note that the value of the threshold value Z is 1.4.

Here, since the shortest hop number of these three routes is 2 hops, the paths N1-N4 are long hop routes.

For this reason, the chooser N4 determines the order from the value of the long hop route cost among the information written in the received flooding message.

Here, the long hop route cost of route 1 is given by 2.67 (= 0.67 + 2), the long hop route cost of route 2 is given by 3 (= 2 + 1), and the long hop route cost of route 3 is given. The cost is 5.17 (= 0.67 + 1.5 +
Since it is given in 3), the order of route 1, route 2, and route 3 is from the lowest.

As described above, when the rearrangement of the routes is completed, the chooser proceeds in the order from the one with the lowest route cost.
A ringing message is sent to try to establish a route (step D8).

The ringing message includes, for example, sender information, chooser information, path ID information, required resource (band) information, maximum hop number information, passing route information (passing node information), passing hop number information, and the like. Written. Note that, among these pieces of information, for example, the passing hop number information is added by the relay node, and the other information is written by the chooser.

Incidentally, when transmitting a ringing message, the chooser executes processing for securing a required band on the transmission link. However, if the available bandwidth of the transmission link is smaller than the required bandwidth, the next path is selected.

Thereafter, the chooser terminates the processing if the establishment of the route is confirmed, and if the establishment of the route fails, selects the next alternative route candidate and attempts to establish the route (step D).
9). If there are no more detour route candidates that can be established, the process ends (step D7).

(D-1-4) Reception of Ringing Message (i) Processing of Relay Node Upon receiving the ringing message sent by the chooser, the relay node sends the message to the next node described in the message. Forward. At this time, if the available bandwidth of the transmission link is equal to or larger than the required bandwidth, the relay node secures the required bandwidth. However, if the available bandwidth of the transmission link is insufficient, the relay node does not transfer the ringing message to the next node, and returns a NACK of the ringing message to the reception link.

The node receiving the NACK of the ringing message releases the reserved band. The NACK of the ringing message is forwarded to the chooser. The chooser receiving the NACK of the ringing message,
A negative result is obtained in step D9 in FIG. 10, and a ringing message is transmitted to a route having the next lowest route cost.

(Ii) Sender Processing When the sender receives the ringing message (FIG. 8)
Step B3), the ACK for the ringing message is returned to the chooser side (Step B4).
Then, the detour path is set and the path is switched, and the bandwidth on the path is secured (step B5). Thereafter, the sender establishes a detour route and ends the processing.

If the ringing message is not received within a predetermined time, the sender confirms that the setting of the detour route has failed, and ends the process.

(D-1-5) A of ringing message
ACK reception The ACK of the ringing message sent by the sender is
The data is transferred to the chooser via the set detour.
Upon receiving the ACK of the ringing message, the chooser confirms that the bandwidth of the detour path has been successfully secured, sets the detour path as the detour path, switches the path, and ends a series of processing (FIG. Step D of 10
10).

If the reservation of the band has failed for all the alternative route candidates (if a negative result is obtained in step D7), the chooser ends the process.

(D-2) Effect obtained by the second embodiment As described above, also in the case of the detour route selecting device (method) according to the second embodiment, the effect of the first embodiment is obtained. In addition to obtaining the same effect as in the case above, the route cost is calculated at each node, so the message writing area can be reduced, and the route information of the same message length and a larger number of hops can be obtained. It is possible to convey.

(E) Third Embodiment (E-1) Network Configuration FIG. 11 shows an example of a network to which a node according to the third embodiment is applied. FIG. 11 shows the same network configuration as FIG. However, in the case of the third embodiment, the calculation of the route cost is unnecessary, and thus the free bandwidth and the cost of each link are not displayed.

Note that, also in the case of the present embodiment, the sender side with a smaller node number is called upstream, and the chooser side with a larger node number is called downstream. For example, FIG.
, The logical path VP between the node N1 and the node N4
1 and the logical path VP2 between the nodes N2 and N5, the sender of the logical path VP1 is N1 and the chooser is N4. Similarly, the sender of the logical path VP2 is N2 and the chooser is N5.

Further, a point where resources required on the network become insufficient is called a bottleneck point, and a node close to the sender among nodes adjacent to the bottleneck point is called a bottleneck node (BNN).

For example, the route candidate 1 (N
1-N2-N4), when the resource of the link L6 is insufficient, the bottleneck point for the route candidate 1 of the VP1 is the link L6, and the bottleneck node (BNN) is the node N.
It becomes 2.

A path for which a detour path is to be established is called a request path, and another existing path passing through a bottleneck point of the request path is called a detour path.

For example, the logical path VP2 is routed (N2-N
4-N5), the logical path V
Logical path VP1 when trying to secure path resources for P1
Is the request path, and the existing path VP2 passing through the bottleneck location L6 is the bypass path.

(E-2) Failure Recovery Function Installed in Each Node The failure recovery function according to the present embodiment proceeds in the order of the processing procedure shown in FIG. Accordingly, in the following description, a failure notification process for notifying a failure (step E1), a search process for searching for a bypass route (step E2), a resource securing process for securing resources for the bypass route (step E3), The reordering process (step E4) for securing resources at the bottleneck location when resources cannot be secured will be described in order.

In the following description, a case will be described in which the rearrangement process is performed on a path candidate having the number of bottleneck portions (the number of bottlenecks) of one.

FIG. 13 shows a time chart of messages transmitted and received between nodes on the network when a bypass path is set.

(E-2-1) Failure Notification Processing (i) Detection of Failure All nodes located on the network periodically execute a failure detection program, and a failure occurs on a link adjacent to the own node. I am watching for

When the node adjacent to the failure point detects the occurrence of the failure, it activates the program shown in FIG. 14 and sends a failure notification message to all links except for the failure point connected to its own node (step F1). , F2).

Thereafter, after transmitting the failure notification message, each node determines whether its own node is a sender (step F).
3) It is determined whether or not the fault location is on the existing path (step F4).

If a positive result is obtained in any of the determination processes, the process proceeds to a search process for searching for a detour route.

On the other hand, if a negative result is obtained in any of the determination processes, the process ends.

(Ii) Reception of Failure Notification Message The failure notification message is transferred between nodes on the network. Upon receiving the failure notification message, each node on the network activates the program shown in FIG. 15 and executes the following processing.

First, the node records the location of the failure described in the failure notification message (step G1).

Next, the node determines whether or not a failure notification message for the same location as the received failure notification message has already been received (step G2).

Here, if a negative result is obtained, the node transfers the message to all connected links except the link that received the failure notification message (step G).
3).

Thereafter, this node determines whether it is the sender (step G5), whether the failure location is on the existing path (step G6), or whether the amount of resources of the transmission link is sufficient (step G7). Is determined.

When a positive result is obtained in any of the determination processes, the process proceeds to the search process as a sender.

On the other hand, if the node has already received the failure notification message for notifying the same failure location, the node does not need to transfer the message redundantly, so the node discards the failure notification message and performs processing. Is completed (step G4).

If a negative result is obtained in any of steps G5 to G7, the processing is similarly terminated.

(E-2-2) Search Process The node that has detected a failure or received a failure notification message and has determined that its own node is the sender has a failure location on the existing path. FIG. 16
Is started and a flooding message is transmitted to a transmission link having a sufficient resource amount (step H1).

Here, the flooding message includes:
Sender information, chooser information, required resource information, maximum hop number information, passing hop number information, and the like are written.

After that, the sender shifts to a band securing process.

The node that has received the flooding message activates the program shown in FIG. 17, and determines whether or not its own node is the destination of the flooding message (step I1).

If the node determines that it is not a chooser recorded in the flooding message, the node shifts to a flooding message transfer process and determines whether a flooding message of the same content has already been transmitted. Is determined (step I2).

When it is determined that the data has already been transmitted,
Since the transfer of the newly received flooding message is unnecessary, the node discards the flooding message and ends the process (step I6).

On the other hand, if the transmission has not been completed, the node determines whether or not the number of hops of the flooding message exceeds the maximum number of hops allowed for the message (step I3). . In this case as well, if it is determined that the number is exceeded, the node ends the process without transferring the message (step I6).

It is to be noted that any one of the steps (I2 and I3)
, The node further determines whether the available resource amount of the transmission link is smaller than the required resource amount (step I).
4).

Also in this case, if the node determines that the resource amount is insufficient, it does not transfer the flooding message and discards it (step I6).

Then, any of the above steps (I2,
If it is determined that the resource amount is also sufficient in I3 and I4), the node adds passing path information (passing node information, passing link free band information, etc.) to the received flooding message, and 1 for the number of passing hops
Is added, and the message is transferred to all connection links except the reception link and the failure link (step I5).

However, when the node that has received the flooding message is the chooser recorded in the flooding message, the node bypasses the passing route information and the like of the detouring route described in the received flooding message. It is recorded as a route candidate, and the process proceeds to the next resource securing process (step I7).

(E-2-3) Resource securing process (i) Chooser process (transmission of ringing message) The chooser that has shifted to the resource securing process starts the program shown in FIG. It is determined whether a ringing message for the previously received flooding message is being transferred (step J1).

At this time, if it is determined that the transfer is in progress, the chooser performs steps J2 and J3 to avoid simultaneous transfer of a plurality of ringing messages.
Jumps to the state of waiting for a reply to the ringing message.

On the other hand, if it is determined that there is no ringing message currently being transferred, the chooser checks whether there is a route candidate that has not transmitted the ringing message. Is selected (step J2).

Here, the route indicated by the flooding message that has arrived first may be selected, or the route with the small route cost may be selected by applying the method of the first or second embodiment described above. You may choose.

In any case, when the selection of the route for sending the ringing message is completed, the chooser secures the required resources of the link based on the passing route information of the alternative route candidates, and sends the ringing message to the route (step J3).

Here, the chooser writes sender information, chooser information, required resource information, passing route information, the number of bottlenecks, and the like in the ringing message. At this time, the initial value of the bottleneck number is set to 0.

[0165] Further, when the ringing message to be transmitted is the last ringing message,
In addition to the above information, a bit indicating the final message is set and transmitted. Here, various methods are conceivable for determining the last ringing message, but a message transmitted after a lapse of a predetermined time after receiving the first flooding message may be used as the last ringing message.

(Ii) Processing of Relay Node Upon receiving the first ringing message, the relay node activates the program shown in FIG. 19 and executes resource securing processing.

When the program is started, the relay node analyzes a message received after the start of the program (step K1).

When the last ringing message or ringing end message is received Here, if the last ringing message or ringing end message is received (step K
(Yes at 2), the relay node transfers the message to the sender (step K3), and shifts to the rearrangement process.

In this case, transmission resources are not secured.

When the ringing message is received On the other hand, if the received ringing message is a ringing message (excluding the last ringing message) (step K4), the relay node determines the destination of the destination written in the ringing message. It is determined whether the link resources are sufficient (step K5).

Here, when the resource amount is sufficient, the relay node secures the band of the transmission link and transfers the ringing message (step K6).

However, even if the resource amount is sufficient, if the number of bottlenecks of the ringing message is one or more, the relay node transfers the ringing message without securing the band of the transmission link.

On the other hand, if a sufficient amount of resources cannot be secured, the relay node transfers a ringing message obtained by adding 1 to the number of bottlenecks to the sender (step K).
7). At the same time, the relay node returns a NACK of the ringing message to the chooser that issued the ringing message (step K8).

At the time of receiving NACK of ringing message On the other hand, what is received is N of the ringing message.
If it is ACK (step K9), the relay node
The resources of the detour path secured during the transfer of the ringing message are released (step K10), and the NACK of the ringing message is transferred to the chooser side (step K11).

Note that N of the ringing message to be transferred is
If the ACK is a NACK for the last ringing message sent by the chooser (step K1
2), the process directly proceeds to the rearrangement process.

At the time of receiving ACK of ringing message On the other hand, what is received is A of ringing message.
If it is CK (step K13), the relay node
The message is forwarded to the chooser.

(Iii) Sender Processing Upon receiving the first ringing message, the sender activates the program shown in FIG. 20 and executes resource securing processing.

When the program is started, the sender analyzes a message received after the start of the program (step L1).

At the time of receiving a ringing end message Here, if the received ringing end message is a ringing end message (step L2), the sender immediately shifts to the rearrangement processing.

When receiving a ringing message On the other hand, if the received ringing message is a ringing message (step L3), the sender first determines whether or not the number of bottlenecks of the route that the arrived ringing message has passed is zero. Check if.

When the number of bottlenecks is not 0 (that is, when the number of bottlenecks is 1 or more and the route cannot be used as a bypass route as it is), the sender:
After recording the route information of the message (step L
5) It is determined whether the ringing message is the last ringing message (step L6).

If the received ringing message is not the last ringing message, the sender shifts to a state of waiting for a ringing message arriving after that, whereas the sender receives the last ringing message. If so, the process shifts to the rearrangement process.

On the other hand, if the number of bottlenecks is 0, the sender determines whether connection of the detour route is possible (step L7).
A NACK of the ringing message is transmitted to the chooser (step L8). On the other hand, if it is determined that the connection is possible, the sender transmits an ACK of the ringing message to the chooser side (step L9), and sets a bypass route (step L10).

(I) Processing of chooser (reception of ringing message ACK / NACK) Returning to FIG. After transmitting the ringing message, the chooser is in a state of waiting for a response to the message to be returned. Eventually, when a response to the ringing message is received (step J).
4), the chooser determines whether the message is ACK or NACK (step J5).

What is received is the ringing message A.
If it is CK, the chooser establishes a detour path and ends the process (step J6).

On the other hand, if the received NACK is a ringing message, the chooser returns to the processing of step J2 to select the next detour. Here, if another route candidate exists, the chooser sends a ringing message to this new route.

If another route candidate does not exist, the chooser determines whether a specified time has elapsed after receiving the first flooding message (step J7).

If it is determined that the specified time has not elapsed, the chooser returns to step J2 because a new flooding message may be received. On the other hand, if it is determined that the prescribed time has elapsed, the chooser notifies the sender of the end of the band securing process to the ringing message transmission path or the like first, thereby notifying the sender of the ringing message. However, in this case, link resources are not secured and only the ringing end message is transferred.

(E-2-4) Rearrangement Processing Next, rearrangement processing which is processing unique to this embodiment will be described. This process is not executed when the bypass route is established at the stage of the resource securing process. In other words, this processing is executed when one or a plurality of bottleneck points exist in any of the detour route candidates.

The rearrangement process will be described below with reference to FIGS.
5 will be described in chronological order. In this process,
Various messages are sent and received between each node.
In FIG. 25 to FIG. 25, the term “rearrangement processing message” is used as a generic term for these.

FIG. 26 shows the specific contents of the rearrangement processing message. Needless to say, the messages transmitted and received between the nodes are determined based on the functions to be performed by the nodes, and not all the nodes shown in FIG. 26 are transmitted and received between any of the nodes.

(I) Processing up to Sending of the Temporary Reserved Message by the Sender of the Request Path The sender requesting the establishment of the bypass route receives the ringing end message or the last ringing message and receives all the ringing candidates. When it is confirmed that the connection has been unsuccessful, the program shown in FIG. 21 is started, and it is determined whether there is a bypass route candidate (step M1).

Here, if it is confirmed that there is no alternative route candidate, the sender ends the processing.

When it is confirmed that there is a bypass route candidate, the sender selects an appropriate route from the bypass route information of the received ringing message, and determines whether the transmission link has a resource amount necessary for setting the bypass route. Is determined (step M2). Here, the bypass route may be selected, for example, in order from the route with the smallest number of bottlenecks.

If there are enough resources on the outgoing link,
The sender tentatively reserves resources (step M3) and sends out a tentative reservation message (step M4). Here, the temporary reservation message includes path ID information, route information, required resource information, and the like.

On the other hand, if there are not enough resources on the transmission link, the sender registers the ID of its own node as information on the bottleneck node to be included in the provisional reservation message (step M5), and sets the 1 is added (step M6), and a temporary reservation message is transmitted.

In the provisional reservation described below, a certain node provisionally reserves the resources of the transmission link and transmits a provisional reservation message, and the next node receiving the provisional reservation message further transmits the provisional reservation message. Is determined to be successful if it is confirmed that the provisional reservation of the resource is possible.

(Ii) Transfer of Temporary Reservation Message or Return of Temporary Reservation NACK Message by Relay Node The relay node that has shifted to the rearrangement process activates the program shown in FIG. When the relay node confirms receipt of the rearrangement processing message (here, any of the temporary reservation message, the temporary reservation ACK message, the temporary reservation NACK message, and the release request message) (step N).
1) First, it is determined whether or not the received message is a temporary reservation message (step N2).

When the temporary reservation message is received, the relay node confirms whether or not the temporary reservation of the resource has succeeded (step N3). If successful, the relay node transfers the temporary reservation message (step N4).

[0200] On the other hand, if the provisional reservation of the resource is not successful, the relay node executes the following processing depending on whether the unsuccessful location is any of the following.

This is a case in which one link is subjected to resource securing processing for a plurality of paths. When the processing is performed at the same time, although the transmission link bandwidth is successfully secured, the processing time Due to the difference, it is assumed that a shortage of the band may be confirmed when the band of the reception link is secured.

When the resources of the transmission link cannot be secured First, the relay node determines whether or not the number of bottlenecks in the received provisional reservation message is 0 (step N).
5).

Here, when the number of bottlenecks in the received provisional reservation message is 0, that is, when there is only one bottleneck point even when the own node serving as a bottleneck is included (negative result in step N5) , The relay node
The own node ID is registered as a bottleneck node (BNN) in the temporary reservation message, and the temporary reservation message with the number of bottlenecks being 1 is transferred (step N4).

On the other hand, when the number of bottlenecks in the received tentative reservation message is 1, that is, when there are two bottleneck points including the own node serving as a bottleneck (Yes in step N5) , The relay node stops the transfer of the provisional reservation message, and
A CK message is returned (step N6).

When the resources of the receiving link cannot be secured (after confirming the securing of the resources of the transmitting link) When the number of bottlenecks in the received tentative reservation message is 0, the relay node confirms the previous node as a bottleneck node. I do. Then, the relay node registers the node ID of the previous node as a bottleneck node BNN in the tentative reservation message, and transfers the tentative reservation message obtained by adding 1 to the number of bottlenecks to the next node (step N4).
Then, the relay node notifies the previous node that the resource reservation is unsuccessful, and also notifies that the previous node is a bottleneck node.

On the other hand, if the number of bottlenecks in the received provisional reservation message is 1, the relay node further determines whether or not the registered bottleneck node is the previous node.

[0207] If it is determined that the bottleneck node is the previous node (No in step N5), the relay node transfers the temporary reservation message as it is.

However, if it is determined that the registered bottleneck node is a node other than the previous node (Yes in step N5), the relay node
Return the temporary reservation NACK message (step N
6).

In the case where resources cannot be secured on both the transmission link and the reception link In this case, the relay node determines that there is one or more bottleneck points, and returns a temporary reservation NACK message (step N6).

(Iii) Processing when Received Temporary Reserved Message by Chooser The chooser that has shifted to the rearrangement processing activates the program shown in FIG.

When the receipt of the reordering processing message is confirmed (step O1), the chooser first determines whether or not the received message is a provisional reservation message (step O2).

[0212] When the temporary reservation message is received, the chooser checks whether or not the temporary reservation of the resource has succeeded (step O3). Here, the chooser immediately adds a value of 1 to the number of bottlenecks (step O4) when it confirms that the number of bottlenecks is successful, and determines whether the number of bottlenecks is 0 if it does not succeed. It is determined whether or not it is (step O5).

If the bottleneck number of the received provisional reservation message is 0, the chooser sends a release request ACK message to the sender after securing the required resources (step O6). Thereafter, a detour path is established, and the process of the chooser ends (step O7).

On the other hand, if the number of bottlenecks in the received tentative reservation message is 1 (if a negative result is obtained in step O5), the chooser determines that the bottleneck node BNN registered in the tentative reservation message is It is determined whether the node is the previous node.

Here, when a positive result is obtained (ie, when the bottleneck node is the previous node), the chooser issues a release request message requesting the bottleneck node BNN to release the temporarily reserved resources. Is transmitted (step O9). The release request message includes a request path ID, a required resource amount of the request path, and the like.

This release request is issued only to the bottleneck node BNN.
In the portion excluding the bottleneck portion of the route including the NN, the resources remain temporarily reserved. In this way, the temporarily reserved path is the request path.

On the other hand, when a negative result is obtained (that is, when the bottleneck node is not the previous node), the chooser returns a temporary reservation NACK message (step O10).

(Iv) Processing of Relay Node Having Received Temporary Reservation NACK Message Return to FIG. 22 again. The relay node that has received the temporary reservation NACK message obtains a positive result in step N7 of FIG. 22, and proceeds to step N8. Here, the relay node releases the resource if the resource for the path corresponding to the message is provisionally reserved, and transfers the provisional reservation NACK message (steps N8 and N9).

(V) Processing of Sender Upon Receiving Temporary Reservation NACK Message Return to FIG. 21 again. Upon receiving the temporary reservation NACK message, the sender obtains a positive result in step M8 in FIG. 21 and proceeds to step M9. Here, the sender releases the resource if the resource for the route corresponding to the message has been provisionally reserved, selects the next route candidate, and transmits the provisional reservation message (steps M1 to M6).

If there is no next alternative route candidate, the process is terminated (step M1).

(Vi) Processing for Sending Detour Request Message by Bottleneck Node The bottleneck node that has received the release request message from the chooser (the sender who obtained a positive result in step M10 in FIG. 21, or the steps N10 and N11 in FIG. 22) 24) activates the program shown in FIG. 24, and uses a resource that is equal to or greater than the amount of resources required by the request path that passes through the bottleneck location and that passes through the bottleneck location. A detour request message for requesting detour of the route is sent to the senders of all other paths (step P1).

The detour request message records the bottleneck location of the request path, the required resource amount of the request path, and the like.

Here, the path to which the detour request message has been sent is the detour path.

(Vii) Response Sending Processing by the Sender of the Detour Path Having Received the Detour Request Message The sender of the detour path having received the detour request message is:
The route shown in FIG. 25 is a route that does not pass through the bottleneck point of the notified request path and that can be established without using the resources temporarily reserved by the request path. Search for existence (step Q
1).

The search for the detour path of the detour path may be performed by using, for example, a method similar to the method using flooding.

If there is no detour path of the detour path, the sender of the detour path notifies the detour request NACK message to the bottleneck node (step Q2). In contrast,
If the detour path sender has a detour path,
The detour request ACK message is notified to the bottleneck node (step Q3).

(Viii) Processing of Bottleneck Node Having Received Response Message to Detour Request When the bottleneck node receives the detour request ACK message (when a positive result is obtained in step P3 in FIG. 24), It is confirmed whether the detour has already been completed in the detour path (step P4).

When the detour has already been completed, no processing is required for the detour request ACK message, and the bottleneck node waits for the next reordering message to be received.

On the other hand, if there is no path for which detour has been completed, it is confirmed whether or not a detour notification is being sent (step P5).

When the bottleneck node is not transmitting, the bottleneck node selects one of the detour paths that can be detoured, and sends a detour notification message to the sender of the selected detour path. For example, the detour path of the detour request ACK message received first may be selected. On the other hand, if transmission is currently being performed, the information of the received detour request ACK message is recorded, and preparation is made for transmission of the next detour notification message.

On the other hand, when the detour request NACK message is received (when a positive result is obtained in step P8 in FIG. 24), the bottleneck node determines whether or not detour is possible in all detour paths. Unless it is confirmed that it is not possible in the detour path, the process returns to a state of waiting for a reordering message (step P9).

On the other hand, if it is determined that the detour is not possible in all the detour paths, the bottleneck node sends a release request NACK message to the sender and the chooser (step P10), and the processing as the bottleneck node is performed. finish.

(Ix) Processing of the Sender of the Detour Path Having Received the Detour Notification Message Return to FIG. After returning the detour request ACK message, the sender of the detour path that has received the detour notification message within a predetermined time attempts to establish a detour path of the detour path.

Here, when the detour path succeeds in establishing the detour path (step Q5 in FIG. 25), the sender of the detour path sends a resource release message to the existing path to secure the current path of the detour path. Released resources (step Q7). However, when the bottleneck node receives the resource release message, it releases the resources of the bypass path and secures the resources of the requested path.

If the establishment of the detour path of the detour path has failed, the sender of this detour path transmits a detour notification NACK message to the bottleneck node BNN (step Q6).

(X) Processing of Bottleneck Node Having Received Response Message to Detour Notification Message Return to FIG. When the received is a detour notification NACK message (that is, step P11 in FIG. 24).
(If a positive result is obtained in step (1)), the bottleneck node determines whether there is another bypass path that can be bypassed, and if so, sends a bypass notification message to the sender (steps P12 and P13).

On the other hand, if there is no alternate path that can be bypassed, the bottleneck node sends a release request NACK message to the sender and chooser of the request path,
The temporary reserved resources on the route are released (step P10).

On the other hand, if the received resource release message is a bottleneck node,
A negative result is obtained in step P11 of FIG.
The bottleneck node that has confirmed that the message is a resource release message in step 4) sends a release request ACK message to the sender and chooser of the request path to validate the provisionally reserved resource on the path (step P15).

(Xi) Processing of Node Having Received Response Message to Release Request Return to FIG. The relay node that has received the release request ACK message validates the temporarily reserved resources and transfers the release request ACK message (step N1 in FIG. 22).
4).

Further, the sender and the chooser that have received the release request ACK message validate the temporarily reserved resources and establish the path of the request path (step M in FIG. 21).
12 and M14, and steps O13 and O1 in FIG.
4).

On the other hand, if the response is a release request NACK message, the relay node receiving the release request NACK message releases the resource if provisional reservation has been made (step N14 in FIG. 22).

If the sender of the request path receives the release request NACK message, the sender selects the next path candidate of the request path after steps M12 and M13 in FIG. Is sent.
If there is no alternative route candidate for the next request path, a positive result is obtained in step M1, and the process ends.

(E-3) Effects obtained by the third embodiment As described above, in the case of the failure recovery apparatus (method) according to the third embodiment, distributed processing is performed when a failure occurs. In addition to recovering from a failed path, recovery of the failed path is performed by re-establishing the established path for the path that could not be recovered. Can be improved.

Further, since the route reassignment process can also be realized by the distributed process, it is possible to perform a recovery process that is robust against failure.

(F) Fourth Embodiment Next, a failure recovery device (method) according to a fourth embodiment will be described.

The fourth embodiment is a modification of the third embodiment. Therefore, here, the operation contents will be described using the network shown in FIG. 11 used in the description of the third embodiment.

(F-1) Failure recovery function mounted on each node Also in the case of the failure recovery function according to the present embodiment, the process proceeds in the order of the processing procedure shown in FIG. 12 described in the third embodiment. The only difference between the fourth embodiment and the third embodiment is the content of the rearrangement process, and the other processes are the same as those of the third embodiment.

That is, the fourth embodiment differs from the fourth embodiment in that provisional reservation of resources is performed at once for a plurality of route candidates having a common bottleneck point, and the route is rearranged.

In the following, in order to clarify the difference, only the rearrangement process will be described. FIG. 27 shows the transmission and reception relationship of messages exchanged between nodes on the network in the case of the present embodiment. This FIG.
FIG. 14 is a diagram corresponding to FIG. 13 described above.

(F-1-1) Rearrangement Processing This rearrangement processing is not executed when a bypass route is established at the stage of the resource securing processing, as in the case of the third embodiment. In other words, this processing is executed when one or a plurality of bottleneck points exist in any of the detour route candidates.

Hereinafter, the rearrangement process will be described in chronological order with reference to FIGS. 28 to 34 when the number of bottlenecks is one.

(I) Processing up to Sending of Confirmation Message by Sender of Request Path A sender requesting establishment of a detour route receives a ringing end message or receives the last ringing message, and connects with all route candidates. If it is confirmed that the ringing message has been unsuccessful, the program shown in FIG. 28 is started, and a confirmation message is transmitted to all the detours that have received the ringing message (step R1).

For example, the sender sends confirmation messages to all routes in order from the route with the smallest number of bottlenecks. The confirmation message includes path ID information, route information, required resource information, and the like.

(Ii) Transfer of Confirmation Message by Relay Node The relay node that has shifted to the rearrangement process activates the program shown in FIG.

When the reception of the reordering message is confirmed (step S1), the relay node first determines whether or not the received message is a confirmation message (step S2).

If the received confirmation message is a relay message, the relay node determines whether or not the available resource amount of the link is equal to or more than the required resource amount of the confirmation message (step S3). If the resources are more than necessary, the relay node transfers a confirmation message (step S4).

On the other hand, if the resources are insufficient, the relay node registers the bottleneck node in the confirmation message, adds 1 to the number of bottlenecks, and transfers the confirmation message (step S5).

Here, as in the third embodiment, the relay node determines whether the bandwidth can be secured for both the transmission link and the reception link.

(Iii) Processing at Receiving Confirmation Message by Chooser The chooser that has shifted to the rearrangement processing activates the program shown in FIG.

Upon receiving the confirmation message, the chooser determines whether or not there is a route candidate with a bottleneck number of 1 (step T1).

If there is a corresponding path candidate, the chooser proceeds to step T2, and selects one bottleneck point from one or a plurality of bottleneck points confirmed from the path candidate (step T2). For example, a bottleneck point is selected from the path of the confirmation message received first.

When the selection of the bottleneck portion is completed in this way, the chooser selects the route candidate having the bottleneck number of 1 and selects all the routes having the bottleneck portion at the location selected in step T2. On the other hand, a temporary reservation message is transmitted (step T3).

At this time, the chooser tentatively reserves a transmission link resource for each of the selected route candidates.
However, when making a temporary reservation of a plurality of routes of the same request path on the same link, the chooser temporarily reserves resources for only one route (required resource amount of the path), and a plurality of route candidates assigns one temporarily reserved resource. Share. At this time, the chooser records the ID of the route candidate sharing the temporary reservation resource.

(Iv) Processing of Relay Node Having Received Temporary Reservation Message The relay node having received the temporary reservation message issued by the chooser obtains a positive result in step S6 of FIG. 29, and proceeds to step S7. Here, the relay node determines whether a temporary reservation of the resource has already been made for the same path.

Here, if the provisional reservation has already been made for another route of the same path (Yes in step S7), the relay node does not newly make a provisional resource reservation, and Only the ID is registered (step S8), and the temporary reservation message is transferred (step S9).

On the other hand, if it is determined that the tentative reservation has not been made yet (negative result in step S7), the relay node
It is determined whether a tentative reservation of resources is possible (step S10). If possible, a link resource is tentatively reserved and a message is transferred (steps S11, S9).

On the other hand, if it is determined in step S10 that the provisional reservation of the resource cannot be made, the relay node determines that the bottleneck location includes two nodes including its own node (or the previous node if the resource of the reception link cannot be secured). It is determined whether or not this is the case (step S12).

Here, when the bottleneck location written in the temporary reservation message is 0 (step S12)
, The relay node does not make a temporary reservation of the link resource, transfers the temporary reservation message, and adds 1 to the bottleneck number of the temporary reservation message.

On the other hand, if the bottleneck location written in the temporary reservation message is 1 (the bottleneck location including the current bottleneck location is 2), the relay node sends the temporary reservation NACK message Is returned to the chooser (step S13).

(V) Processing of Relay Node and Chooser Having Received Temporary Reservation NACK Message The temporary reservation NACK message is transferred to the chooser via the relay node on the corresponding route.

At this time, the relay node that has received the provisional reservation NACK message obtains a positive result in step S14 of FIG. 29, and only shares the route for which the provisional reservation failed (sharing the provisional reservation resource with another route). It is determined whether there is a temporarily reserved resource (step S15).

If a negative result is obtained, the relay node deletes only the path candidate ID (step S1).
6). On the other hand, if a positive result is obtained, the relay node releases the temporary reservation resource (step S17), and then deletes the route candidate ID (step S16).

Thereafter, the relay node transfers the provisional reservation NACK message to the chooser side (step S1).
8).

Eventually, the provisional reservation NACK message reaches the chooser. The chooser that has received the temporary reservation NACK message determines whether replies to the temporary reservation message have been received from all the route candidates including the temporary reservation NACK message (step T4 in FIG. 30).
Then, while there is a path without a response, the process waits until all responses are obtained.

(Vi) Processing of Sender Having Received Temporary Reservation Message On the other hand, if the temporary reservation message is transferred to the sender without sending a temporary reservation NACK message back to the chooser, the sender who has received the temporary reservation message Obtains an affirmative result in step R3 in FIG. 28 and proceeds to step R4. Here, the sender determines whether or not a tentative reservation for the same path has already been made on the reception link where the message arrived, and if a tentative reservation exists, obtains a positive result and reserves any resources. The state shifts to a state of waiting for the next rearrangement processing message without any change.

[0276] On the other hand, if the temporary reservation does not exist for the same path (when the temporary reservation of the resource is successful in the portion excluding the bottleneck point on the bypass route), the sender determines that the reception link A temporary reservation is made and a temporary reservation ACK message is returned to the chooser (step R6).

In this way, a path in which a required resource has been provisionally reserved on a link excluding a bottleneck point on the path is called a request path. In the case of the fourth embodiment, a plurality of request paths can exist at the same time.

(Vii) Processing of Relay Node and Chooser Having Received Temporary Reservation ACK Message
In steps S19 and S21 of No. 9, a negative result is obtained, and the process proceeds to step S22, where the received tentative reservation ACK message is transferred to the chooser side.

Eventually, the provisional reservation ACK message reaches the chooser. Here, in Step T4 of FIG. 30, the chooser determines whether or not there is a response for all the route candidates that have transmitted the provisional reservation message including the received provisional reservation ACK message. Waits for another response.

On the other hand, if it is confirmed that responses have been received for all the route candidates, the chooser determines whether the provisional reservation has been successful (step T5). Here, that the provisional reservation has been successful means that at least one provisional reservation ACK message has arrived.

Accordingly, in this case, the provisional reservation ACK message has arrived, the chooser proceeds to step T7, and transmits a release request message to the bottleneck node.

Here, the release request message records required path information, bottleneck location information, and the required resource amount of the required path.

[0283] If all the responses are provisional reservation NACK messages, the chooser still has a route candidate that passes through the next bottleneck location in order to restart the process for another bottleneck location. (Step T6).

Then, as long as the affirmative result is obtained, the process returns to step T2 and the above process is repeated. On the other hand, if no such route candidate exists, the chooser gives up setting the bypass route,
The process ends.

(Viii) Processing of Relay Node and Bottleneck Node Having Received Release Request Message The relay node that has received the release request message is the destination of the release request message, although not shown in FIG. 29 due to space limitations. It is determined whether or not the bottleneck node is its own node. If a negative result is obtained, the received release request message is transferred to the next node as it is (step S22).

On the other hand, if the relay node determines that the bottleneck node which is the destination of the release request message is its own node, it activates the program shown in FIG.
A bypass request message is issued to the sender and the chooser of a path that uses resources equal to or more than the required resource amount of the request path among the paths passing through the bottleneck point (step U).
1).

Here, the bypass request message records the required path information, bottleneck location information, and the required resource amount of the required path.

The path for which the detour request has been issued is called a detour path.

(Ix) Processing of Sender of Detour Path Receiving Detour Request Message The sender of the detour path receiving the detour request message is:
The program shown in FIG. 32 is started, and whether or not there is a detour path that can be established without using the temporarily reserved resources on the path that does not pass through the bottleneck ring of the notified request path Is sent to the chooser of the bypass path (step V1).

(X) Processing of Relay Node Receiving Detour Flood Message The relay node that has received the detour flood message transfers the message to a transferable link if necessary resources are available. Here, it is determined whether there is a resource necessary for the bypass of the bypass path based on the following criteria.

In the case where the available resources of the link are equal to or more than the required amount of resources In the case where the total of the available resources of the link and the resources provisionally reserved in the request path is equal to or more than the necessary resource amount, The detour-flooding message is recorded by recording the detour-path ID of the request path that uses the existing resource in the detour-flooding message.

(Xi) Processing of the bypass path chooser that has received the bypass request message The bypass path chooser that has received the bypass request message activates the program shown in FIG. 33, and checks whether the bypass flood message has been received within a predetermined time. Is determined (step W1).

When the reception of the bypass flooding message is confirmed, the chooser continuously determines whether the bypass of the bypass path is possible (step W2).

Here, the chooser of the detour path, if the route candidate ID of the provisionally reserved request path is not recorded in the received detour flooding message (this is the case, and if a negative result is obtained in step W4), It is determined that the bypass of the bypass route is possible without using the temporarily reserved resources.

On the other hand, if the route candidate ID of the temporarily reserved request path is recorded in the received detour flooding message (this is the case and a positive result is obtained in step W4), the temporarily reserved resource is If it is used, it is determined that the detour on the detour route is possible, and the temporary reservation resource use information is recorded in the detour request ACK (step W5).

When a negative result is obtained in step W1 or W2, the chooser of the detour path determines that detour is impossible.

Thus, the chooser of the detour path obtains any of the following determination results (a), (b) or (c),
One of them is notified to the bottleneck node (step W3 or W6).

(A) It is possible to bypass the bypass route without using the temporarily reserved resources.

(B) By using the temporarily reserved resources, it is possible to bypass the bypass route.

(C) It is impossible to bypass the detour route.

However, in the case of (b), as described above, the chooser of the detour path notifies the combination of the path IDs of the temporary reservation resources of the request path required for detour of the detour path.

(Xii) Processing of each node when there is a response of (a) Here, as a response to the detour request message,
Subsequent processing when the bottleneck node receives the response of the content of (a) from the chooser will be described.

The bottleneck node that has received the response obtains a positive result in step U2 of FIG. 31, and sends a detour notification message to the chooser of the detour path that has been notified (step U4).

This detour notification message reaches the detour path chooser via the relay node. The chooser of the detour path that has received the detour notification message is shown in FIG.
After an affirmative result is obtained in step W7, the process proceeds to a process of securing resources of the bypass route and establishing the route (step W8).

However, when there are a plurality of detour paths, the chooser of the detour path that has received the notification selects one path and secures the resources of the detour path up to the sender of the detour path.

[0306] If the detour path that has received the detour notification fails to establish the detour path, the detour path chooser returns a detour notification NACK message to the bottleneck node to notify the failure of the process ( Step W
9).

In this case, the bottleneck node is as shown in FIG.
Then, the process returns to step U2 via step U5, selects the next detour path that can be detoured, and repeats the process of transmitting the detour notification message to the chooser of the detour path.

On the other hand, if the detour path succeeds in establishing the detour path, the chooser of the detour path sends a resource release message to the existing path to release the resources secured by the current path (FIG. 33). Step W10).

[0309] This resource release message is notified to the bottleneck node via the existing path.

In this case, the bottleneck node is as shown in FIG.
A positive result is obtained in step U5, and the process proceeds to step U6.
Release the resources of the detour path and secure the resources of the requested path.

Thereafter, the bottleneck node selects one path from the path candidates of the request path, and transmits a release request ACK message to its sender and chooser.
The provisional reservation resource on the route is validated (step U7).

[0312] At this time, the bottleneck node sends a temporary reservation release message to its sender and chooser, and instructs nodes secured by other routes to release the temporary reservation resources (step U8). .

As a result, the relay node obtains an affirmative result in step S19 in FIG. 29 and proceeds to step S20, releases any temporary reservation resources if any, and transfers the message.

Further, the sender of the request path releases the temporary reservation resource, if any, and establishes a bypass route (steps R7, R8 and R11, R14 in FIG. 28).

Similarly, the chooser of the request path establishes a detour (steps T9 and T10 in FIG. 30).

(Xiii) Processing of each node when there is a response of (b) Here, as a response to the detour request message,
The response after the bottleneck node receives the response of the content of (b) (that is, when the detour path of the detour path exists, but the request path requires the resource temporarily reserved) from the chooser. Will be described.

[0317] The bottleneck node that has received the response selects the chooser of the detour path for transmitting the detour notification message in step U2 in Fig. 31 (step U4).

For example, as shown in FIG. 34, there are three routes # 1, # 2, and # 4 in the alternative route candidates of the request path, and
This is a case where two alternative routes exist in the alternative route of the alternative route 1, and the alternative route candidate 1 of the alternative route 1 is # 1 of the request path.
When the tentative reservation resources # 2 and # 4 are used, detour is possible, and the detour path candidate 2 of the detour path 1 can be detoured using the tentative reservation resources # 1 and # 4 of the request path. If the bypass path candidate 2 is adopted, it is determined that the request path can be bypassed by the path # 2 even if the temporary reserved resources of the request paths # 1 and # 4 are used in the bypass path of the bypass path. .

[0319] However, it is determined that it is impossible to detour the request path in the alternative path candidate 1.

[0320] The bottleneck node that has selected the detour path that can be detoured by the above-described method sends a detour notification message to the chooser of the detour path and notifies the route (step U4 in Fig. 31). This detour notification message includes information on which detour path should be selected.

[0321] This detour notification reaches the detour path chooser via the relay node. After receiving the detour notification message, the chooser of the detour path obtains an affirmative result in step W7 of FIG. 33, and then proceeds to a process of securing resources of the detour path and establishing the path (step W8).

However, when there are a plurality of detour paths, the chooser of the detour path that has received the notification selects one path and secures the resources of the detour path to the sender of the detour path.

When the detour path receiving the detour notification message fails to establish the detour path, the detour path chooser sends a detour notification NACK message to the bottleneck node to notify the failure of the process. (Step W9).

In this case, the bottleneck node is as shown in FIG.
Then, the process returns to step U2 via step U5, selects the next detour path that can be detoured, and repeats the process of transmitting the detour notification message to the chooser of the detour path.

On the other hand, when the detour path succeeds in establishing the detour path, the chooser of the detour path sends a resource release message to the existing path to release the resources secured by the current path (FIG. 33). Step W10).

In this case, the bottleneck node is as shown in FIG.
A positive result is obtained in step U5, and the process proceeds to step U6.
Release the resources of the detour path and secure the resources of the requested path.

Thereafter, the bottleneck node selects one path from the path candidates of the request path, and transmits a release request ACK message to its sender and chooser.
The provisional reservation resource on the route is validated (step U7).

At this time, the bottleneck node sends a temporary reservation release message to its sender and chooser, and orders the nodes secured by other routes to release the temporary reservation resources (step U8). .

As a result, the relay node (step S19 in FIG. 29) releases the temporary reservation resource, if any, and transfers the message (step S20).

Similarly, the chooser of the request path establishes a detour (steps T9 and T10 in FIG. 30).

(Xiv) Processing of each node when there is a response of (c) Here, as a response to the bypass request message,
Subsequent processing when the bottleneck node receives the response of the content of (c) (that is, when there is no detour path of the detour path) from the chooser will be described.

The bottleneck node receiving the response obtains a negative result in step U2 in FIG. 31, and sends a release request NACK to the sender and the chooser of the request path.
The message is sent (step U3 in FIG. 31).

The relay node, having received the release request NACK message, obtains a positive result in step S21 of FIG. 29 and proceeds to step S20, releases the resource if the resource is provisionally reserved, and transfers and processes the message. To end.

Eventually, the release request NACK message
Reach the sender and chooser of the request path.

The sender of the request path that has received the release request NACK message obtains a positive result in step R9 in FIG. 28 and proceeds to step R10, and releases any temporarily reserved resources, if any.

After that, the sender of the request path determines whether there is a next bottleneck point and a detour (step R12). If there is, the sender of the request path sends a confirmation message to the detour path candidate. Is repeated (step R13).

[0337] If another alternative route candidate no longer exists, the sender of the request path ends a series of processing.

Similarly, when the release request NACK message is received in the case of the chooser of the request path, the release request NACK message shown in FIG.
A positive result is obtained in step T8, and the process proceeds to step T6.
It is determined whether there is another detour path candidate that passes through another bottleneck location where the number of bottlenecks is one.

[0339] If a positive result is obtained, the chooser selects one of them, and sends a temporary reservation message again to all the paths that pass the bottleneck node and have a bottleneck number of one. And tries to establish a detour route.

If it is not possible to establish a detour path of a detour path at all bottleneck locations where the number of bottlenecks is one, the search for a detour path of the requested path is terminated.

(F-2) Effect obtained by the fourth embodiment As described above, in the case of the failure recovery apparatus (method) according to the fourth embodiment, a common bottleneck node is provided. By making a temporary reservation for a plurality of detour routes at the same time, it is possible to simultaneously check whether a detour is possible on a plurality of routes and to search for a detour route in a shorter time than in the third embodiment. Can be expected.

Further, by sharing the tentatively reserved resources on different paths of the same path passing through the same link, the resources to be tentatively reserved on the entire network can be reduced, and the factors hindering the establishment of the paths of other paths are reduced. be able to.

(G) Other Embodiments (1) In the above-described first and second embodiments, the detour route is divided into two groups, a short hop route and a long hop route, and the detour routes are divided by different methods. Although the case of calculating the path cost of the path has been described, the cost of the path may be calculated by dividing the path into more groups.

(2) In the above-described first and second embodiments, the short hop route and the long hop route use different route cost calculation criteria, and then, in order from the one with the smallest route cost, In the above description, the case of trying to set a detour route is described, but other calculation criteria and selection criteria may be used.

For example, the same calculation criterion may be applied to the short hop route and the long hop route, and the selection order may be reversed between the short hop route and the long hop route.

For example, in the first embodiment, when the calculation criterion used for calculating the route cost of the short hop route is also applied to the long hop route, the detour routes are selected in the short hop route in the descending order of the route cost. Alternatively, on the long hop route, the detour route may be selected in the descending order of the route cost.

(3) In the above-described first and second embodiments, the route cost is calculated using the free band of the link A as CAPA (A). However, the route cost is calculated using another index. You may go. For example, assuming that the free bandwidth of the link A is B (A) and the link capacity of the link A is C (A), the value given by CAPA (A) = B (A) / C (A) is calculated as the above-mentioned path cost. May be applied to the calculation.

(4) In the above-described first and second embodiments, the value of the threshold value Z is determined in advance. However, the threshold value Z may be determined after a failure has occurred.

(5) In the first and second embodiments described above, the ringing message is transmitted after the specified time has elapsed, but the transmission of the ringing message may be determined by another method. For example, a ringing message may be transmitted when a specified number of flooding messages are received.

(6) In the third embodiment,
Although a method of notifying a failure by transferring a failure notification to an adjacent node has been described, other failure notification methods may be used. For example, if the node that has detected the failure is a transmission link in which the available resource amount of the link is smaller than the required resource amount in the search processing in the third embodiment, the flooding message is not transmitted, but the number of bottlenecks is one. When the rearrangement process is performed for the route of the above, the number of bottlenecks is counted even when the flooding message is transferred, and when the number of bottlenecks is 1, the flooding message is transferred, and the route candidate is also rearranged. May be.

(7) In the third embodiment described above,
Although the case where the sender of the request path sends a temporary reservation message after receiving the ringing end message has been described, also in the present embodiment, as in the fourth embodiment, the sender of the request path sends a confirmation message. Alternatively, the sender of the request path that has received the confirmation message may transmit the temporary reservation message.

(8) In the third and fourth embodiments described above, a bypass route having only one bottleneck node has been described, but a route having a plurality of bottleneck nodes may be bypassed. .

[0353]

As described above, according to the first to fourth aspects of the present invention, the path cost calculation criterion or the selection criterion is switched according to which processing group the path for which the detour path is searched belongs. By doing so, a detour path can be selected by a method corresponding to each path, and the time required for the detour path selection can be reduced.

According to the fifth aspect of the present invention, even when there is no alternative route that can secure resources, other paths that use the resources of the location where the resources could not be secured can be used.
When the route can be detoured, the route of another path can be detoured, and the free resources generated by the detour can be assigned to the detour route of the target path. It can be further improved.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a procedure of an execution process of a chooser according to a first embodiment.

FIG. 2 is a diagram provided for explanation of a conventional technique (No. 1).

FIG. 3 is a diagram provided for explanation of a conventional technique (No. 2).

FIG. 4 is a diagram showing a schematic configuration of a node.

FIG. 5 is a diagram for explaining a network and terms to which the first embodiment and the second embodiment are applied;

FIG. 6 is a diagram schematically illustrating a communication procedure in the first and second embodiments.

FIG. 7 is a diagram showing a procedure of an execution process of a sender according to the first embodiment.

FIG. 8 is a diagram showing a procedure of an execution process of a sender according to the second embodiment.

FIG. 9 is a diagram illustrating an execution processing procedure of a relay node according to the second embodiment.

FIG. 10 is a diagram illustrating an execution processing procedure of the chooser according to the second embodiment.

FIG. 11 is a diagram provided for describing a network to which the third embodiment is applied;

FIG. 12 is a diagram showing a schematic procedure of a communication process;

FIG. 13 is a diagram showing an outline of a communication procedure in the third embodiment.

FIG. 14 is a diagram illustrating a processing procedure executed by a node adjacent to a failure point;

FIG. 15 is a diagram illustrating a processing procedure executed by a node that has received a failure notification message.

FIG. 16 is a diagram illustrating a processing procedure executed by the sender that has shifted to the search processing.

FIG. 17 is a diagram illustrating a processing procedure executed by a node that has received a flooding message.

FIG. 18 is a diagram illustrating a processing procedure executed by the chooser that has shifted to the resource securing processing.

FIG. 19 is a diagram showing a resource securing processing procedure executed by the relay node receiving the first ringing message.

FIG. 20 is a diagram illustrating a resource securing processing procedure executed by the sender that has received the first ringing message.

FIG. 21 is a diagram illustrating a processing procedure executed by the sender that has shifted to the rearrangement processing;

FIG. 22 is a diagram illustrating a processing procedure executed by the relay node that has shifted to the rearrangement processing;

FIG. 23 is a diagram illustrating a processing procedure executed by the chooser that has shifted to the rearrangement processing;

FIG. 24 is a diagram illustrating a processing procedure executed by the bottleneck node that has started the bottleneck processing.

FIG. 25 is a diagram illustrating a processing procedure executed by a sender of a detour path.

FIG. 26 is a diagram illustrating a description of terms appearing in the description of the rearrangement process.

FIG. 27 is a diagram schematically illustrating a communication procedure according to the fourth embodiment.

FIG. 28 is a diagram illustrating a processing procedure executed by the sender that has shifted to the rearrangement processing;

FIG. 29 is a diagram illustrating a processing procedure executed by the relay node that has shifted to the rearrangement processing;

FIG. 30 is a diagram illustrating a processing procedure executed by the chooser that has shifted to the rearrangement processing;

FIG. 31 is a diagram illustrating a processing procedure executed by the bottleneck node that has started the bottleneck processing.

FIG. 32 is a diagram illustrating a processing procedure executed by a sender of a detour path that has received a detour request message.

FIG. 33 is a diagram illustrating a processing procedure executed by the chooser of the bypass path that has received the bypass request message.

FIG. 34 is a diagram provided for explanation of a case where a bypass of a bypass path is possible if a request path uses a temporarily reserved resource.

[Explanation of symbols]

 N0 to N6 ... nodes, L1 to L7 ... links.

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Fumito Kubota Inventor 4-2-1 Nukikitamachi, Koganei-shi, Tokyo F-term in the Communications Research Laboratory, Ministry of Posts and Telecommunications 5B045 BB11 JJ46 5K030 JA11 LB08 LC09 5K051 FF17 FF18 LL02

Claims (16)

[Claims] When a bypass route is selected, a path end node that has received a notification of a bypass route candidate determines which processing group a path to search for the bypass route belongs to, and after the determination,
A detour route selection characterized by ranking the route costs calculated for each detour route candidate based on a selection criterion specific to the processing group to which the path belongs, and determining a detour route candidate selection order. Method. 2. When a detour path is selected, a path end node that has been notified of a detour path candidate determines which processing group a path for searching for a detour path belongs to, and after the determination,
The route cost of each detour path candidate calculated by a calculation method specific to the processing group determined to belong to the path is ranked based on a predetermined selection criterion, and the selection order of the detour route candidates is determined. Characteristic detour route selection method. 3. When a detour path is selected, the path end node that has received the notification of the detour path candidate determines which processing group the path for which the detour path is to be searched belongs.
The route cost notified for each of the detour route candidates is ranked based on a selection criterion specific to the processing group to which the path belongs, and the selection order of the detour route candidates is determined. Detour route selection method. 4. When a detour path is selected, a path end node that has received a notification of a detour path candidate determines which processing group a path for searching for a detour path belongs to, and after the determination,
Among the plurality of route costs calculated by the calculation method unique to each processing group, which is notified for each of the detour route candidates, the processing cost for the processing group determined to belong to the path is determined by a predetermined selection criterion. And selecting a detour route candidate selection order based on the detour route selection method. 5. The alternative route selection method according to claim 1, wherein the processing groups are classified based on the number of hops required for setting each path. . 6. A notifying process for notifying a fault that has occurred on an existing path, a searching process for searching for a detour path for the existing path for which a fault has been notified and the possibility of securing resources thereof, If there is a detour route that can be secured, the resource is secured, and the resource securing process for switching the existing route to the detour route is performed. Instructs other paths that use the resources of the location that failed to secure the route to detour the route, secures the resources corresponding to the route, and then replaces the route of the existing path for which the failure was notified with the detour route passing through the location. And a rearrangement process for switching between them. 7. The failure recovery method according to claim 6, wherein a part where required resources cannot be secured is detected during the search processing for searching for a detour path by transferring the search message, the search message is included in the search message. A failure recovery method characterized by transferring a search message to an adjacent node after updating the value of the number of bottlenecks to be performed. 8. The fault recovery method according to claim 6, wherein, in the rearrangement process, the resource at the location is used for one of the detour paths having a location on the route that is a failure in securing resources. A failure recovery method comprising: determining whether a detour of a path of another path to be performed is possible; and, if the detour fails, determining the possibility of detour for the next one. 9. The failure recovery method according to claim 6 or 7, wherein at the time of the rearrangement processing, a plurality of detour paths having a failure point in resource securing on a common path are simultaneously determined. It is determined whether it is possible to bypass the path of another path that uses the resource at the relevant location, and if the bypass fails,
The possibility of detour is determined for one or more alternative routes that have another location on the path in common with the resource securing obstacle, and if the detour is successful, a plurality of locations having the location on the route are determined. A failure recovery method characterized in that one of the alternative routes is selected as a final alternative route. 10. When a detour path is selected, if the own device is a path end node that has been notified of a detour path candidate, it is determined which processing group a path for searching for a detour path belongs to. Thereafter, a function is provided in which a route cost calculated for each detour path candidate is ranked based on a selection criterion specific to a processing group to which the path belongs, and a selection order of detour route candidates is determined. Detour path selection device. 11. When a detour path is selected, if the own apparatus is a path end node that has been notified of a detour path candidate, it is determined which processing group a path for searching for a detour path belongs to. Thereafter, the route costs of the detour route candidates calculated by the calculation method specific to the processing group determined to belong to the path are ranked based on a predetermined selection criterion, and the detour route candidate selection order is determined. A detour route selection device having a function. 12. When a detour path is selected, if the own device is a path end node that has been notified of a detour path candidate, it is determined which processing group a path for searching for a detour path belongs to. Thereafter, a function is provided in which the route costs notified for each of the detour path candidates are ranked based on a selection criterion specific to the processing group to which the path belongs, and the selection order of the detour route candidates is determined. A detour route selection device characterized by the above-mentioned. 13. When a detour route is selected, if the own device is a path end node that has been notified of a detour route candidate, it is determined which processing group a path for searching for a detour route belongs to. After that, of the plurality of route costs calculated by the calculation method specific to each processing group, which is notified for each of the detour route candidates, the processing cost of the processing group determined to belong to the path is determined by a predetermined method. An alternative route selection device, which has a function of determining the order of selection of alternative route candidates based on the selection criteria. 14. A notifying function for notifying a failure that has occurred in an existing path, a search function for searching for a detour path to the existing path for which a failure has been notified, and the possibility of securing resources thereof, If there is an alternate route that can be secured, the resource is secured and the resource securing function that switches the existing route to the alternate route.If the alternative route that can secure the resource does not exist as a result of the search function, Instructs other paths that use the resources of the location that failed to secure the route to detour the route, secures the resources corresponding to the route, and then replaces the route of the existing path for which the failure was notified with the detour route passing through the location. A fault recovery device comprising a re-arrangement function for switching between the two. 15. A node comprising: the detour path selecting device according to claim 10 or the failure recovery device according to claim 14; and a switch for switching input / output of transmission data. . 16. A network system comprising one or more nodes according to claim 15.