JP2000253009A - Method and device for fault restoration in atm network - Google Patents

Abstract

PROBLEM TO BE SOLVED: To increase the recovery rate of a virtual path VP. SOLUTION: When a faulty virtual path takes place in a line network nodes 2 and 3 in an asynchronous transfer mode ATM consisting of nodes 1-4, a Sender 3 of a path receiver side node broadcasts a reservation message to a Tandem 1 (4) of an adjacent node. The Tandem receiving the reservation message broadcasts the reservation message to an adjacent Tandem 4 (1) furthermore. When the reservation message reaches a Chooser 2 at a path transmitter side node, the Chooser 2 stores the reservation messages received within a prescribed time all as bypass candidates and adopts the bypass candidates for a bypass until a predetermined minim warrant restoration rate is satisfied with respect to a request band of the faulty virtual path. Thus, the faulty virtual path is switched into virtual paths of a plurality of bypasses and restored.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ATM (Asynchronous
More particularly, the present invention relates to a method and apparatus for restoring a failure in an ATM network, which improves a restoration rate of a failure (a band of a bypass route with respect to a band of a failed virtual path).

[0002]

2. Description of the Related Art As is well known, an ATM network includes a plurality of A networks.
Nodes such as TM exchanges and ATM routers, and physical links such as optical cables connecting the nodes (hereinafter referred to as links)
It is composed of As shown in FIG. 16, a plurality of VPs (Virtual Paths) 51a, 51b,... Are logically accommodated in the link 50, and each VP, for example, the VP 51a further includes a plurality of VPs. VC (Virtua
l Channel (virtual channel)) 52a, 52b, 52
are logically accommodated.

[0003] In this ATM network, if any failure occurs in the physical link and communication between nodes at both ends of the physical link becomes impossible, a message between adjacent nodes can be transmitted without using the configuration information of the entire network. 2. Description of the Related Art A fault recovery method (self-healing method) for autonomously determining a detour by transmission / reception has been proposed.

[0004] An example of a conventional self-healing method will be described with reference to FIG. Now, the node 61 of the ATM network.
If the link between the link 62 and the link 62 fails and the VP of the link fails (hereinafter, referred to as a failed VP), the receiving node 62 sends a message (reservation message) for searching for a detour. Broadcast.
The adjacent nodes 63 and 64 that have received the reservation message reserve a VP constituting a detour, and further broadcast the reservation message. The adjacent nodes 65 and 66 which have received the reservation message reserve a VP constituting a detour, and further broadcast the reservation message. When the reservation message reaches the transmitting node 61 by this broadcast communication, the transmitting node 61 determines a single detour. Then, the transmitting node 61 transmits the assignment message to the receiving node 62 via the nodes 65 and 63 in order to notify the receiving node 62 of the determined detour. When the receiving node 62 receives the assignment message, the failure recovery is completed. The transmitting node 61 is a Chooser, the receiving node 62 is a Sender, and the other nodes 63 to 66 are Tand.
em, commonly called.

[0005] The reservation message includes Sender, Choose
r and fault VP identifier, detour required bandwidth, via Ta
V that does not satisfy the required bandwidth
P and VPs that are equal to or greater than the predetermined maximum number of hops are Ta
Excluded from detour by ndem. If there is a VP that is not subject to the detour, the VP that has been reserved so far
Is transmitted from the Tandem 66 to the Sender 62 to cancel. If a plurality of VPs fail simultaneously due to a link break or the like, the above-described restoration procedure is repeated until all VPs are restored.

In the event of a failure, the order of restoration is set in advance for each VP, and a plurality of V
A method has also been proposed in which, when Ps simultaneously become faulty, a limited band is allocated to a VP with a high priority and restored.
In this case, it has been proposed to use the value of the VP identifier, the size of the band of the failed VP, or the like as a reference for assigning the priority.

[0007]

The conventional fault recovery method recovers a faulty VP by a single detour. Even if a VP has a high priority, the faulty VP can be recovered if the spare band is narrow. In addition, there is a problem that the detour route cannot be determined or the bandwidth to be allocated is insufficient, and the restoration rate of the VP is reduced.

An object of the present invention is to provide a method and an apparatus for restoring a fault in an ATM network which eliminates the above-mentioned problems of the prior art and increases the VP restoration rate.

[0009]

In order to achieve the above object, the present invention provides an AT configured by connecting nodes by links.
A first feature of the M network failure recovery method is that a virtual path of a link in which a failure has occurred is switched to a virtual path of a plurality of detours and recovered.

[0010] The present invention also provides an ATM network failure recovery device comprising nodes connected by a link, wherein the node has a failure recovery control module for recovering the failure when the link fails. The failure recovery control module comprises a Sender,
Assigned as a function as a chooser or tandem,
The second feature is that when the function as a chooser is assigned, all the detours of one or more failed virtual paths (hereinafter referred to as failed virtual paths) are determined by a single restoration procedure. There is.

According to the first aspect of the present invention, the restoration rate of the virtual path of the link in which the failure has occurred, that is, the restoration rate of the band of the restoration target VP is restored by a conventional single detour. Compared to this, it can be greatly improved. According to the second feature, high-speed recovery can be performed.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings. FIG. 1 is a conceptual diagram of the simplest model of the ATM network failure recovery method of the present invention. Now, the figure
As shown in (a), when data (packet data) is transmitted from the node 1 to the node 3 via the node 2, if a failure link occurs between the nodes 2-3, the node An operation for recovering from the fault between 2-3 is performed. This failure recovery includes a pre-reservation method and a failure setting method. The failure setting method can be further classified into an inter-bus switching method and an inter-node switching method. In the present invention, the inter-node switching method is used. set to target. Therefore, at the time of recovery from a failure, a detour as shown in FIG. 3B is formed.
P, that is, when a fault VP cannot be restored by a single detour due to a narrow spare band of the detour, it can be restored by using a plurality of detours. The configuration and operation of an embodiment of the ATM network failure recovery apparatus will be described in detail. Hereinafter, the node 3 is referred to as Sender 3, the node 2 is referred to as Chooser 2, and the nodes 1, 4 are referred to as Tandem as necessary.

First, the configuration of the node will be described with reference to the block diagram of FIG. Since the nodes 1 to 4 have the same configuration, the node 1 will be described as a representative.

As shown, the node 1 includes a failure recovery control module 11 and a time monitoring module 12
, Message monitoring module 13, notification monitoring module 14, network management center I / F (interface)
It comprises a module 15 and a communication processing module 16.

The failure recovery control module 11 has a function of initializing and activating each of the modules, receives a notification and a processing request from each module, and performs processing such as Sender, Chooser, or Tandem. The failure recovery control module 11 further includes an initialization block a, an event monitoring block b, a notification / message analysis block c, a timer management block d, a detour management block e, and Se.
It comprises an nder block f, a Tandem block g, a Chooser block h, and a notification / message transmission block i.

The initialization block a is first executed when the node is started, loads the environment settings, establishes a connection with each of the blocks b to i and adjacent nodes, and initializes each module 12.
-16 are performed (shown by dotted lines). When the above processing is completed, the initialization block a
Transfer processing to The event monitoring block b detects and analyzes events such as notification and message reception, timeout from the time monitoring module 12, and detour inquiry and switchback request from a network management center (not shown). According to the type, the process is shifted to the notification / message analysis block c, the timer management block d, or the bypass management block e.

The notification / message analysis block c determines whether the node concerned is a sender or a sender based on the received notification or message.
Decide which node of Chooser or Tandem to function as, and move the process to the block for each node. The timer management block d parses the timer identifier (see the description of the time monitoring module 12), and sends the Sender, Chooser,
Or notify the Tandem block of the timeout. The detour management block e is connected to the network management center (network management center I).
/ F module 15) is processed. Sender block f
Performs the processing when the notification / message analysis block c determines that the sender is Sender. The Tandem block g performs a process when the notification / message analysis block c determines that it is Tandem. Further, the Chooser block h performs a process when the notification / message analysis block c determines that it is a Chooser. Further, the notification / message transmission block i transmits the response to the detour inquiry request from the network management center and the message transmission from Sender, Chooser, and Tandem via the network management center I / F module 15 and the communication processing module 16. Do.

Next, the time monitoring module 12 manages the time for providing a timer to the failure recovery control module 11. With the introduction of the timer identifier, Sende
Signal multiple timeouts in r, Chooser, or Tandem. The message monitoring module 13 includes:
It constantly monitors and detects the reception of messages for failure recovery. When the reception is detected, a message is transmitted to the failure recovery control module 11. The notification monitoring module 14 constantly monitors and detects a notification issued by a node when a failure occurs or recovers. When a failure occurrence or failure recovery is detected, the failure recovery control module 11 is notified. When there is a network management center that integrally manages a network including a plurality of nodes, the network management center I / F module 15 performs, for example, a notification of the end of failure recovery and a process of requesting a detour return from the network management center. Provide an interface for The communication processing module 16 decodes cells and packets received from the network into messages and notifications. In addition, messages and notifications are encoded into cells and packets and transmitted to the network.

Next, the message format used in this embodiment will be described with reference to FIG. In ATM, cells of a fixed length of 53 octets (1 octet is 8 bits) are used for communication. Of these, the first 5 octets are V
Routing information such as PI (virtual path identifier) and VCI (virtual channel identifier) is stored. The 48 octets following the header include OAM (Operation, Operation) for management.
Administration, and Maintenace) information is divided into several fields and stored.

The fields include OAM type, function type, function-specific information field, reservation, and CRC-
10. The OAM type is a field indicating a management target (VP or VC). The function type includes fields indicating details of the management function, for example, a failure report, a failure recovery, a connectivity test (loopback test), and performance monitoring such as an error rate and transmission delay.
It also has a function to start or end these functions. The function-specific information field is a field in which information and a result required to execute the function type are stored. The reservation is a field reserved for the purpose of expanding functions in the future. CRC-10
Is a field for storing an error detection code for detecting whether or not there is an error in the cell.

In this embodiment, the OAM type is "V
P ", the function type is" failure recovery ", and the" function-specific information field "in this case is as shown in FIG.

In FIG. 3, "length" is the length of the function-specific field, and "s-node" is Se which is uniquely determined in the network.
Choosing the nder identifier, "c-node" is uniquely determined in the network
ser identifier, "s-port" is the S that contains the failed VP
ender port identifier, "c-port" is the VP that failed
Is the port identifier of the Chooser that contains the. The port identifier of the sender and the port identifier of the chooser are uniquely determined for each node. “Fault-VPI” is the VPI of the fault VP for which the detour is determined, and is valid only for the assignment message. “Sig-flag” indicates the type of message (reservation message, assignment message, or cancellation message). “Hop” represents the number of hops (number of nodes passing) of the detour candidate. The value is increased by one each time the signal passes through Tandem, and those exceeding the maximum number of hops (MAXHOP) specified in advance are excluded from the detour route candidates. The “Tandem list” is a set of an identifier i of a transit Tandem (including a Sender) of a detour candidate reserved by the reservation message x and a value of a remaining band r _{x, i} (reservation band and reserved band) after reservation. Is a list of elements. The maximum number of elements in the list is maximum ho
Equal to p number (MAXHOP). "Bx" indicates the band of the bypass route candidate reserved by the message x in the case of the reservation message, and the determined band of the bypass route in the case of the allocation message. “Qos-flag” is a QoS (Quality of Service; Qualit) such as a transmission delay and an error rate to be considered when determining a detour.
y of Service). "Qosi-value"
It represents the QoS value of the detour candidate specified by “qos-flag”. The one that satisfies the required QoS, or the one that does not satisfy the required QoS, for example, the one having the minimum value in the case of transmission delay or error rate is determined as the detour.

Next, the node of FIG. 2 is, for example, the Sennde of FIG.
The operation when r3 is reached will be described with reference to the flowchart of FIG. First, in step S1, the reserved bandwidth, QoS, unused VPI (virtual path identifier) and the like of the node are checked, and a reservation message in the format of FIG. 3 is created based on these. In step S2, after creating the reservation message, the reservation message is transmitted to all adjacent nodes (excluding the Chooser) that satisfy all the QoS conditions. In step S3, a timer is started when transmitting the reservation message. In step S4,
It is determined whether or not the time from transmitting the reservation message to receiving the assignment message has timed out.

In step S5, it is determined whether or not the reservation message has been received during the recovery from the failure. If this determination is affirmative, the reservation message is discarded. On the other hand, if the determination is negative, the process proceeds to step S7, where it is determined whether a cancellation message has been received. If this determination is affirmative, the process proceeds to step S8 to cancel the reservation of the VP that is a detour candidate. On the other hand, when the determination is negative, the process proceeds to step S9, and it is determined whether an assignment message has been received. When this determination is affirmative, the process proceeds to step S9,
A detour is determined based on the assignment message. In step S11, it is determined that the failure recovery has been normally performed, and the process ends.

On the other hand, if the determination in step S9 is negative, that is, if the determination in step S4 is affirmative before the assignment message is received and a timeout occurs, the process proceeds to step S12, and a cancel message is transmitted to all nodes. In step S13, the failure recovery ends abnormally as unsuccessful. If there is a network management center, it notifies that the detour cannot be determined.

Next, the node of FIG. 2 is, for example, the Tandem of FIG.
The operation at the time of 1 or 4 will be described with reference to the flowchart of FIG.

At step S21, it is determined whether or not a reservation message has been received, and if this determination is affirmative,
Proceeding to step S22, referring to the Tandem list,
It determines whether it has its own identifier and checks that the detour does not form a loop. If step S22 is affirmative, a loop is formed, so step S2
Proceed to 3 to discard the reservation message. Then, the process proceeds to step S24 to transmit a cancellation message to the transmission source node of the reservation message, thereby canceling the reservation of the detour candidate forming the loop.

If the determination in step S22 is negative, the process proceeds to step S25 to determine whether or not the condition of the number of hops is satisfied. If not satisfied, the process proceeds to step S23, where the reservation message is discarded. On the other hand, if satisfied, the process proceeds to step S26, in which the value of the number of hops is increased by 1, and the reservation message in which the Tandem list is updated is transmitted to all adjacent nodes (Sender and Tandem).
(Excluding the nodes in the list), reserve a detour candidate, and detect the reception of the message again.

When the result of the step S21 is negative, the process proceeds to a step S27 to judge whether or not the assignment message has been received. When the assignment message has been received, the process proceeds to a step S28 to determine the assignment message. Determine a detour based on the Tandem list. continue,
In step S29, the assignment message is transmitted to the adjacent node via the detour. Further, step S30
, A cancel message for canceling the reservation of the detour candidate is transmitted to all other adjacent nodes. Thereafter, the process returns to step S21 again.

If the determination in step S27 is negative, the flow advances to step S31 to determine whether a cancel message has been received. If this judgment is affirmative,
Proceeding to step S32, based on the cancellation message,
Cancel reservation of a detour candidate. In step S33, it is determined whether or not all cancellation messages for the transmitted reservation message have been received. If all of them have been received, the process proceeds to step S34 to transmit a cancellation message to the transmission source node of the reservation message. If step S31 is negative, that is, if all the cancellation messages have not been received yet, the process returns to step S21, and the reception of the message is detected again.

Next, the node in FIG. 2 is, for example, the Choose in FIG.
The operation when r2 is reached will be described with reference to the flowchart of FIG. In step S41, a timeout is set in advance as a time limit for receiving a reservation message. In step S42, it is determined whether or not a reservation message has been received. If it has been received, the process proceeds to step S43, in which the reservation message is held, and a timer for defining a time during which a plurality of reservation messages can be received is started. . In step S44, it is determined whether or not a reservation message has been received. If received, the process proceeds to step S45, where the reservation message is held as needed, and a detour candidate is added. In step S46, it is determined whether or not a timeout has occurred. If the determination is negative, the process returns to step S44,
The above operation is repeated. When a timeout occurs, the process proceeds to step S47, and based on all reservation messages received up to the timeout, a one-shot determination of collectively determining a detour of the recovery target VP in one recovery procedure. Execute the method.

Next, the operation of the one-shot determination method performed by the Chooser will be described with reference to the flowcharts of FIGS. In step S51, all Qo values such as the bandwidth required for the detour, the transmission delay, and the error rate
It is checked whether there is a reservation message that satisfies the S condition. If there is no reservation message that satisfies all QoS conditions, the process proceeds to step S52, and a cancellation message for canceling reservation of a detour candidate not determined as a detour is sent to Senn.
Send to der. Next, the process proceeds to step S53, where the Sender is notified of the end of the recovery procedure of all the faulty VPs.

If step S51 is affirmative, that is, if there is a reservation message that satisfies all the QoS conditions, the process proceeds to step S54 to reserve the minimum reserved band Bx that is equal to or more than the required band BW of the recovery target VP. Check for message x. Note that the Tandem list of the reservation message x is Lx. If there is no reservation message that satisfies the condition of step S54, the process proceeds to step S55, and one reservation message (x) having the maximum reserved band (Bx) is selected. Next, proceeding to step S56, a detour candidate based on the message x is determined as a detour, and an assignment message for this is transmitted to the Sender. In the next step S57, the reserved band Bx is smaller than the required band BW, and all of Bx is provided to the detour, so that the detour candidate reserved by the message x cannot be used when determining another detour. Therefore, the reserved band Bx is set to 0.

If there is a reservation message that satisfies the condition of step S54, the process proceeds to step S58,
By using a detour candidate reserved by the message x, a detour that satisfies the required bandwidth can be provided. Therefore, the alternative route candidate reserved by the message x is determined as an alternative route, and an alternative route assignment message is transmitted to the Sender. In step S59, since the required bandwidth BW of the reserved bandwidth Bx has been consumed, the reserved bandwidth is updated to Bx-BW.

As described above, since the detour route candidate reserved by the message x is determined as the detour route, it is necessary to update the bandwidth available for the detour route candidates reserved by other messages. In the following, the bandwidth of another message is updated.

In step S60 of FIG. 8, reservation messages other than the reservation message x are named as not updated. In step S61, it is checked whether or not there is an unupdated reservation message, that is, a reservation message other than the reservation message x. If there is an unupdated reservation message, it is determined that all the unupdated reservation messages have been updated. Until the following update.

In step S62, one unupdated reservation message is set (y and its Tandem list is set to Ly).
select. Next, the process proceeds to step S63, where the Tandem list Lx of the message x is compared with the Tandem via the Ly, and
Check if there is a link w shared by x and Ly. If we determine that there is no shared link w,
Proceeding to step S64, the reservation message y is updated. On the other hand, if it is determined that there is a link w to be shared, the process proceeds to step S65, and it is determined whether the remaining band r _{x, w} + Bx on the link w is smaller than the reserved band By.

If it is smaller, that is, as shown on the left side of FIG. 9, if the remaining band r _{x, w} <reserved band By, the process proceeds to step S66, where the detour by the reservation message x is determined. The band of the detour candidate by the message y is reduced to _{rx, w} . Therefore, here, the remaining band r in all the links (excluding the link w) constituting the detour candidate by the reservation message y
_{y, i} (1 ≦ i ≦ | Ly |, i ≠ w) is converted to ry _{, i}
+ By -r _x, the case of _{w -Bx (i = w, r} x, w
= 0), and then By is updated to r _{x, w} + Bx, respectively. Where | Ly | represents the number of elements in the Tandem list Li.

On the other hand, if it is larger, that is, as shown on the right side of FIG. 9, if the remaining band r _{x , w} + Bx ≧ the reserved band By, the process proceeds to step S67, where the detour candidate by the reservation message x is determined. Even if it is determined to be a detour, the band of the detour candidate by the reservation message y can be used as it is. Therefore, updates residual bandwidth _{r y} in the link w when bypass candidate by reservation message y has been determined to _detour, the _{w r x,} the w + Bx-By.

If step S61 is negative,
That is, when all reservation messages other than the reservation message x have been updated and there are no unupdated reservation messages, it is checked whether or not the minimum guaranteed recovery rate of the VP to be recovered is satisfied. If the determination in step S68 is affirmative, the process proceeds to step S69 to check whether there are any more recovery target VPs. If the determination in step S69 is affirmative, the process proceeds to step S70 to update the required bandwidth BW to the value of the required bandwidth of the next recovery target VP. Thereafter, the process returns to step S51 in FIG. 7, and the above operation is repeated.

On the other hand, if the determination in step S69 is negative, the process proceeds to step S52, in which a cancellation message for canceling reservation of a detour candidate not determined as a detour is sent to Se.
nder, and in the next step S53, all the recovery target Vs
Notify the Sender of the end of the P recovery procedure. If the minimum guaranteed recovery rate is not satisfied in step S68, the process proceeds to step S71, where the requested bandwidth BW is set to BW−
In order to increase the restoration rate of the same failure VP by updating to Bx and restoring by a plurality of detours, a detour is determined again using the remaining reservation message (step S51).

As is clear from the above description, when the determination in step S68 is negative, steps S71 and S5 are performed.
1, S54,..., And according to the present embodiment, the fault VP
However, when a single detour cannot be restored because the spare band of the detour is narrow, this is restored using a plurality of detours.

Next, the operation and effect when the present invention is applied to the ATM network configuration of FIG. 10 will be specifically described. In the figure, A to P represented by O represent ATM nodes, and → and ← represent links connecting the nodes. In this embodiment, it is assumed that the conditions for failure recovery are as follows, with the priority set to the size of the band (the recovery procedure is executed earlier for VPs with a larger band). (1) The processing time of each message in each node is 2
m seconds. (2) VPs passing through 10 or more nodes are excluded from the bypass route (the maximum number of hops is 9). (3) Timeout in Chooser shall be 20 ms.

The bandwidth of all the links in FIG.
55 Mbps, and the numerical value of each link is a spare band (unit: Mbps). Here, the reserved band is 25 Mb
One numerical value from ps to 55 Mbps is randomly selected and assigned. The failure link is assumed to be a failure of a one-way link from the node J to the node K. The receiving node K of the failed link is Sender, and the transmitting node J is Choose.
r. In the failed link, 48M
It is assumed that six VPs of bps, 15 Mbps, 12 Mbps, 9 Mbps, and 3 Mbps are accommodated, and all of these VPs are to be restored.

Assuming that a failure link has occurred between the nodes J and K as described above, Sender K is determined in step S in FIG.
First, the processing of step 1 is performed to check the spare band, QoS, unused VPI, and the like of each link between the nodes G, L, and O, and create a reservation message based on these. Next, in step S2, a reservation message is transmitted to these nodes. Next, in step S3, a timeout period (for example, 2000 ms) from the transmission of the reservation message to the reception of the assignment message is set. In steps S5, S7, and S9, the reception of various messages is detected until the timeout occurs. I do.

Next, the nodes G, L, and O are connected to the Sender
Receives the reservation message transmitted by the, and performs processing as Tandem. That is, when the reservation message is received in step S21 of FIG. 5, the process proceeds to step S22, where it is detected whether or not the detour route candidate in the received reservation message forms a loop. In step S25, the hop number condition is satisfied. Inspect whether to do. That is, it is checked whether the hop number is 10 or more. When a loop is not formed and the hop number condition is satisfied, the value of the hop number (hop in FIG. 3) is increased by 1, the reserved band (Bx in FIG. 3) is updated, and the reservation message is To nodes (from node G to nodes C, F, H, from node L to nodes H, P, and from node O to nodes N, P)
Send. Thereafter, similarly, the reservation message arrives at Chooser J via the Tandem node.

When Chooser J determines that the reservation message has been received for the first time in step S42 of FIG. 6, it holds the reservation message and starts a timer (timeout 20 ms) in step S43. In this embodiment, a total of 44 reservation messages are received before the timeout of step S46, but FIG.
Shows the first seven reservation messages.

The numerical value on the right arrow in the detour candidate i (1 ≦ i ≦ 7) is the remaining band (the difference between the reserved band and the reserved band) r _{i, j} (M
bps). For example. Reserve band 31 Mbp in FIG.
In the link from s node J to F, detour 1 is 2
To reserve 6 Mbps, the remaining bandwidth r _1,1 = 31
-26 = 5 Mbps.

When the timeout in step S47 is made, a one-shot determination method is executed. First, FIG.
In step S54, the highest priority 48 Mbps
Of the fault VP is determined. Since there is no detour candidate of a band whose reserved band exceeds the requested band of 48 Mbps, the process proceeds to step S55, and a detour candidate 2 having a maximum band of 27 Mbps is selected. Next, the process proceeds to step S56,
The bypass route candidate 2 is determined to be a bypass route, and an assignment message is transmitted to the Sender. In step S57, the reserved bandwidth B2 of the bypass route candidate 2 is set to 0.

Next, the alternative route 2 is determined as an alternative route.
Therefore, in steps S60 and S61, a detour candidate
Updating the remaining bandwidth of another detour candidate sharing the link with 2
I do. In step S63, the detour candidate 5 is a detour candidate.
Complement 2 and the first link (link between nodes JN)
It is determined that it has, and the process proceeds to step S65. Step S65
Then, the remaining bandwidth r in the link between the nodes J and N
_2,1= 20 (Mbps) is the reserved bandwidth of the detour candidate 5
Since B5 is smaller than 26 (Mbps), B5 = 20
(Mbps). Therefore, step S
At 66, the remaining links of all the links constituting the detour candidate 5
Band r_{5, j}(1 ≦ j ≦ 5) by r_{5, j} + B5 -r
_2,1= R_{5, j}+ 26-20 = r_{5, j}+6
(Mbps) and B5 = 20Mbps
Update. Similarly, for each link of the detour candidate 2,
Update of the remaining bandwidth of other detour candidates with links
Do.

In step S68, 27/48 = 0.5
Since 625 <0.9, which does not satisfy the predetermined minimum guaranteed recovery rate of 90%, the process proceeds to step S71, where the required bandwidth is updated to 48-27 = 21 (Mbps), and 48M
The second detour for the recovery target VP of bps is determined. Thereafter, the same processing is continued until there are no detour candidates whose bandwidth is larger than 0 (step S51), or until all restoration target VPs are restored (step S69). Finally, in step S52, a message of canceling the detour candidate not determined as the detour is transmitted to the Sender.
Notify nder.

The detour determined by the above operation is as shown in FIG. The relationship between the pre-failure band of the recovery target VP and the post-recovery band in the present embodiment is shown in FIG.
As shown in FIG. 48Mbps of recovery target VP and 9
Mbps recovery, two detours (27 Mbps each)
s and 21 Mbps; 7 Mbps and 2 Mbps), and the total recovery rate is 85/93 (= 92.47%).
On the other hand, in the conventional method of recovering a faulty VP by a single detour, as shown in FIG.
(= 77.42%), and the total recovery rate according to the present embodiment is improved by about 15% as compared with the conventional method.

As can be seen from FIG. 13, in the proposed method, a high-priority fault VP may be recovered at the minimum guaranteed recovery rate, whereas a low-priority fault VP may not be recovered at all. , As compared with the conventional method. However, the recovery rate of each failed VP is based on the priority, and the recovery rate can be differentiated by the priority.

FIG. 14 illustrates a failure recovery operation according to the second embodiment of the present invention. In this embodiment, multiple QoS
Is taken into account, and the following transmission delay cost conditions (4) to (7) are added to the conditions (1) to (3) in FIG. (4) A value of 1 to 10 is randomly selected and assigned to the transmission delay value of each link. (5) The value of the VP transmission delay is the value of the transmission delay of the accommodated link. (6) The value of the transmission delay of the detour is determined by each VP constituting the detour.
And the sum of the transmission delays. (7) When the value of the transmission delay is 33
Only the following are detours. In FIG. 13, A and B in (A, B) represent the reserved band and the transmission delay, respectively.

In the ATM network having the configuration shown in FIG. 14, the bandwidth before restoration of the VP to be restored is different between the case where the proposed method in which the processing of FIGS. FIG. 15 shows a comparison between the bandwidth and the post-restoration band. In the proposed method, two detours (27 Mbps and 21 Mbps, respectively) are performed when the recovery target VP recovers 48 Mbps and 15 Mbps.
bps; 10 Mbps and 5 Mbps), and the total recovery rate is 65/93 (= 69.89%). On the other hand, in the conventional method of recovering a faulty VP by a single detour, it is 59/93 (= 63.44%), and the total recovery rate according to the present embodiment is improved by about 6% compared to the conventional method. I do.

In this embodiment, although there is a single detour candidate having a bandwidth of 15 Mbps (see FIG. 13), the condition of the transmission delay is not satisfied. Of the maximum bandwidth (here, 10 Mbps) among the detours satisfying the above.

According to the present invention, as described above,
In some cases, a single failed VP is divided and restored by a plurality of detours, so that packet data transmitted on the VP before the occurrence of the failure is divided and transmitted by the plurality of detours. In other words, the existing technology P
PP multilink protocol (for example, IETF, RF
C1990. The PPP Multilink Protocol (MP), Aug. 199
6) can be restored to one original packet.

[0058]

As is apparent from the above description, according to the present invention, the virtual path of the link in which a failure has occurred is switched to a virtual path of a plurality of detours and restored.
The restoration rate can be greatly improved as compared with the case where the restoration is performed using the virtual path of the conventional single detour.
In addition, the restoration rate of a virtual path having a high priority can be greatly improved.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram of a failure recovery method for an ATM network according to the present invention.

FIG. 2 is a block diagram showing a configuration of a node of the ATM network of the present invention.

FIG. 3 is an explanatory diagram of a message format for failure recovery.

FIG. 4 is a flowchart illustrating processing of a Sender.

FIG. 5 is a flowchart showing a Tandem process.

FIG. 6 is a flowchart showing a Chooser process.

FIG. 7 is a flowchart illustrating a process of a one-shot determination method.

FIG. 8 is a flowchart continued from FIG. 7;

FIG. 9 is an explanatory diagram of processing of the one-shot determination method.

FIG. 10 is a circuit diagram of a first embodiment of the ATM network of the present invention.

FIG. 11 is a diagram showing an example of a reservation message received by a chooser.

FIG. 12 is a diagram showing a detour to be restored.

FIG. 13 is a diagram illustrating a pre-failure band and a post-recovery band of a recovery target VP.

FIG. 14 is a circuit diagram of a second embodiment of the ATM network of the present invention.

FIG. 15 is a diagram showing a pre-failure band and a post-recovery band of a recovery target VP.

FIG. 16 is a diagram showing a configuration of an ATM transmission path.

FIG. 17 is a schematic explanatory view of a conventional self-healing method.

[Explanation of symbols]

1, 4, Tandem, 2 Chooser, 3 Sender, 11 Fault recovery control module, 12 Time monitoring module, 1
3: Message monitoring module, 14: Notification monitoring module, 15: Network management center I / F module, 16:
Communication processing module.

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Keizo Sugiyama, Inventor 2-1-15 Ohara, Kamifukuoka, Saitama Prefecture K.D. Institute (72) Inventor Sadao Obana 2-1-15, Ohara, Kamifukuoka, Saitama F-term in the Caddy Laboratory (reference) 5K030 GA12 HA10 LB07 LB08 LC08 LC09 MA04 MB01 MD02

Claims (7)

[Claims] 1. An AT configured by connecting nodes with links
A fault recovery method for an ATM network, wherein a faulty link virtual path (hereinafter referred to as a faulty virtual path) is switched to a plurality of detour virtual paths for recovery in an M network fault recovery method. Method. 2. The failure recovery method for an ATM network according to claim 1, wherein the plurality of detours are set so as to satisfy a predetermined minimum guaranteed recovery rate with respect to a required bandwidth of the failed virtual path. A fault recovery method for an ATM network. 3. The ATM network failure recovery method according to claim 1, wherein said plurality of detours are set so as to satisfy a predetermined quality of service (QoS). Network recovery method. 4. The ATM network failure recovery method according to claim 3, wherein the quality of service (QoS) includes at least one of a transmission delay and an error rate. 5. The ATM according to claim 1, wherein:
In the network failure recovery method, all the detours of one or more failed virtual paths are determined by the Chooser which is the transmitting node between the failed links in one recovery procedure. ATM to be
Network recovery method. 6. An AT configured by connecting nodes with links
In the M network failure recovery device, the node comprises a failure recovery control module for recovering the failure when the link has failed, the failure recovery control module according to the link failure location, When assigned as a function as Sender, Chooser, or Tandem, and when assigned as the function as the Chooser, all the detours of one or more failed virtual paths (hereinafter referred to as “failed virtual paths”) are performed once. A fault recovery device for an ATM network, which is determined by a recovery procedure. 7. The failure recovery device for an ATM network according to claim 6, wherein the Chooser holds all reservation messages received within a predetermined time as detour candidates, and with respect to a required bandwidth of the failed virtual path, A fault recovery apparatus for an ATM network, wherein the detour candidate is adopted as a detour until a predetermined minimum guaranteed recovery rate is satisfied.