JP2003520525A - Distributed restoration method and system for communication network - Google Patents

Abstract

(57) [Summary] Paths connect nodes, detect disruptions in the traffic span in the network, and immediately restore circuit connectivity within a predetermined maximum time using a new failure detection, isolation and recovery method A network with multiple links connecting multiple nodes, such as a cross-connect in a telecommunications network, where there is spare capacity between as many nodes as possible to accommodate at least some traffic rerouting.

Description

Detailed Description of the Invention

BACKGROUND OF THE INVENTION 1. Technical Field of the Invention The present invention relates generally to communication systems and their methods of operation, and to dynamic recovery of communication traffic through a communication network, and in particular to failure. The origin node and destination node of the path are
It relates to a messaging scheme that can receive information about which spans or links of the failed path are still available. The present invention relates to distributed recovery algorithm (DRA) networks, and more particularly to a method for isolating the location of a fault in the network and an apparatus for implementing the method, and the interruption of traffic by the failure of any of the working links of the network. In this case, it relates to a method of monitoring the topology of the backup link of the network in order to switch the route of traffic.

Further, the present invention relates to a distributed restoration method and system for restoring communication traffic flow in case of detection of a failure in a span of a communication network, and more particularly to a 1633-SX broadband digital cross-connect switch. It relates to a communication network.

Description of Related Art Whether by excavator, freezing storms, or flock of hungry mice, D
Losing a leg or bundle of communication channels, such as S3 or SONET telephone channels, means losing significant revenue. After the first 1.5 seconds of outage, there is a great risk that one of the network's local offices could be outaged due to an excessive carrier group alarm.

Several methods are generally used to restore a communication network. Some of them are well known methods. First, there is a method called route diversity. Route diversity refers to the situation where there are two cables between the source and the destination. One cable takes the north-facing path and the other cable takes the south-facing path. If the northbound path fails, traffic is sent on the southbound path (or vice versa). This is a very high quality recovery mechanism, generally due to its speed. However, route diversity has the problem that it is very expensive to adopt.

Another well-known method for network recovery is the ring. This is a particularly attractive way when a large number of stations are connected to each other. The stations are connected by a ring. If any connection in the ring fails, the circularity of the ring will cause the traffic to be routed in the fault-free direction. Therefore, the ring survives the connection even if there is a single break. However, it has the disadvantage that the nodes of the communication network must be connected in a circle. This type of restoration is not possible without establishing the circular configuration required by the ring.

Another method of network restoration, mesh restoration, requires the rerouting of traffic through the network in some way. Therefore, mesh recovery uses the spare capacity of the network to reroute traffic with spare or busy connections. In general, mesh restoration provides the lowest quality of service in that communication restoration takes longer than route diversity and mesh restoration. On the other hand, mesh recovery has the advantage of not requiring as much reserve capacity as route diversity and ring recovery. There are two possible methods for performing network restoration using mesh restoration.

One is called centralized recovery and the other is distributed recovery. In centralized mesh recovery, a central computer controls the entire process and all associated network elements. All network elements report to and are controlled by the central computer. The central computer verifies the state of the network, calculates alternate paths and sends commands to network elements to perform network recovery. In a sense, centralized mesh recovery is easier than distributed mesh recovery. With distributed mesh restoration, there is no central computer that controls the entire process. Instead, network elements, especially cross-connects, exchange messages to communicate with each other to determine the optimal recovery path.
Therefore, distributed mesh restoration performs one level of parallelism where one restoration program runs on many computers simultaneously.

In this way, parallel processing occurs even if the computers associated with the network elements are geographically dispersed. There are instruction sets running on multiple machines that work together to restore the network.

The communication network is composed of a plurality of nodes connected to each other by spans and links. Most of these spans and links are fiber optic cables. A network path is defined as a connection between two end nodes through which information and traffic flow. One end node is defined as the origin node and the other end node is defined as the destination node. There are many paths that connect a source node to a destination node. Then, if one of the paths carrying the traffic fails, the other path is used as an alternate path and the traffic is rerouted. Therefore, if a path-based approach to recovering interrupted traffic is used in a communication network and the network is equipped with a distributed recovery algorithm (DRA), the origin node and the destination node are considered.

In communication networks, in the event of a failure, one link or one span along the path is often affected. Such disasters occur, for example, due to cable breaks. Therefore, the network paths on either side of the fault are unaffected and may be used to bypass the faulty part. But,
Path-based DRA-equipped networks may ignore intact parts and look for entirely different alternative routes when restoring traffic. Damage to one span of the network can happen to cause many paths to fail through the same span, so choosing an alternate path indiscriminately to recover a failed path can result in
There is the problem that at the same time or subsequently the recovery of other failed paths may be ignored. Therefore, a DRA equipped network must consider all end-to-end paths that need to be restored in order to ensure maximum restoration after a failure.

Therefore, an excellent DRA-equipped path-based recovery method that can be used to create an alternative recovery path by checking the intact part of the failed path and thus improving recovery efficiency and integrity is required. Has been.

In a communication network equipped with a distributed recovery algorithm (DRA), the network can recover traffic that is interrupted by a failure or malfunction of a specific location. In such a DRA equipped network, or part of a network called a domain, a node of the network, or a digital cross-connect switch, routes traffic interrupted by a malfunction or failure at one of the nodes or links of the network, respectively. Each node is equipped with the DRA algorithm and associated hardware so that each node can find an alternate path to fix. Each node is interconnected by a span that includes a working or protection link to at least one other node. Therefore, each node is typically connected to its neighbors using at least one working or protection link. It is through these links that messages as well as traffic signals are transmitted to and received by the nodes.

In a DRA network, when a failure occurs on any working link, traffic is rerouted by the backup link. Therefore, in order to operate effectively, the backup links of the DRA network must always be working, or at the very least the network should know which backup links are working and which are not. There must be.

In addition to routing traffic, links provide each node with signals that inform the node of the health of the network. Each signal has a signal that traffic is being routed effectively between nodes, or that there is a malfunction somewhere in the network that requires one or more alternate routes to reroute interrupted traffic. Provided to the node.

Traditionally, when everything is working well, idle signals and other similar signals are passed between nodes to inform them that traffic is being routed normally. However, if a failure that blocks the flow of traffic occurs somewhere in the network, an alarm is sent from the location of the failure and is communicated to the network nodes. This alarm signal causes the network equipment downstream of the fault location to enter an alarm state. Follow-up signals are sent to suppress alarms in downstream equipment.

This prior art of sending an alarm signal from a faulty location helps isolate faults on a single link. Unfortunately, however, this standard of sending an alarm signal downstream from a fault location requires that the downstream node receiving the alarm signal further propagates the alarm signal downstream. As a result, all nodes in the network will receive an alarm signal shortly after a failure, so it may be difficult for network administrators to identify a failed link or a managed node at a failed site. Make it very difficult, if not impossible. This is because, in addition to the management node, many other nodes in the network also receive alarm signals.

Therefore, there is a need for a way to know that the true management node of a failed link is indeed the management node. In other words, there is a need for a way to distinguish between alarm signals received by nodes other than the management node and alarm signals received by the management node in order to preserve the perceived behavior of sending the alarm signal to downstream equipment.

In most cases, the distributed recovery domain is part of the entire communications network,
Since there are many different networks, the state of the signal received by a node outside the distributed recovery domain, as if there was no difference between the time when the signal was received by the management node and later by the node outside the domain. Must also be retained.

It is also necessary for the DRA network of the present invention to have an up-to-date map of the network's working spare links, ie spare capacity, so that traffic interrupted by a failure can be recovered immediately.

(Summary of the Invention) The present invention connects a plurality of nodes, such as a cross-connect of a communication line network, with a control channel that connects all the nodes to each other, and maintains the continuity of the line within a predetermined maximum time. Includes the concept that spare capacity exists between enough nodes to reroute at least some traffic as soon as possible upon detection of a disconnect in the network's traffic span to recover. .

Further, in order to allow a DRA equipped network to utilize the intact part of a failed path, the messaging method of the present invention provides information about the intact span or link to the point of failure to the origin node of the failed path. And the destination node.

To that end, the failure is first detected by the neighboring management nodes that sandwich the failure. Each of these near-failure management nodes initiates the propagation of the reuse message to either the origin or destination nodes. This reuse message has a variable length route information field and an identifier that identifies it as a reuse message. The reuse message is propagated from node to node, and upon returning to the origin node or destination node, each node through which the reuse message passes adds its own unique node identification (ID) to the route information field. Thus, when the origin or destination node receives the reuse message, the node can read the description of the intact part of the path from the route information field of the reuse message.

By allowing the recovery logic to take into account the intact part of the original path, a better recovery decision can be made that utilizes the intact part of the path. In one embodiment, such restoration logic can only be applied to the restoration of a failed path that originally contained the intact portion. In other words, the use of such intact parts is limited to the restoration of failed paths. These restrictions make the recovery process much easier and avoid unexpected problems.

The invention also provides that upon receipt of a fault signal (or an AIS signal that suppresses alarms in downstream equipment), each node in the domain performs a distinct non-alarm signal (or Modifying the function of each node in the distributed recovery domain to cause the propagation of non-AIS signals). As a result, the management node of the failed link or the nodes sandwiching the malfunctioning site will not be able to receive a true alarm or AI.
Will receive the S signal.

Since neighboring nodes are connected by links or spans, the type of signal passing through each node in the network has a different format depending on the type of connection. In the case of Digital Service 3 (DS3) equipment, each node in the distributed restoration domain is equipped with a converter, and when the AIS signal is received, the converter converts the AIS signal into an idle signal (or modified AIS signal), and the idle signal To the downstream nodes.

At least one of the C bits of the idle signal to perform the conversion of the AIS signal
One is fixed. When direct failure indications such as Loss of Signal (LOS), Loss of Frame (LOF), Loss of Pointer (LOP) occur, the modified idle signal is propagated to the nodes downstream of where the failure occurred.

In the vicinity of the distributed recovery domain, each access / egress node connecting the domain to other parts of the network or to other networks ensures that equipment outside the distributed recovery domain can continue to receive standards-compliant signals. , The incoming modified idle signal is reconverted or replaced by the standard AIS signal.

SONET Synchronous Transport Signal (STS-n) like STS-3 standard
In a network interconnected by optical fibers, where each node in the distributed recovery domain receives an incoming STS-N AIS signal, the AIS
The signal is adapted to be replaced by an STS-N input signal error (ISF) signal. In the preferred embodiment, the Z5 bit of STS-3 is modified. This modification has the same purpose as changing the C bit of the DS3 signal.

In order to provide an up-to-date map of the working spare links of the network, the topology of the networks connected by the working spare links is made available by the management node sandwiching the malfunctioning link as soon as a failure is detected. The management node, designated as the sender or origin node, uses the topology of the backup link to
Reroute traffic through a working backup link.

In order to confirm that the backup link is working, prior to the DRA process, special messages, called live messages in the present invention, are continuously exchanged on the backup link between neighboring nodes. This alive message can identify not only a special field that identifies the alive message that comes from the management node when a failure is detected, but also the port of the transmission source node, the node identification, and the input IP address and output IP address of the node. , Has a large number of fields. This live message is transmitted on the C-bit channel as an idle signal.

As long as the backup link is operating normally, the live message over the backup link is
For example, by an operations support system, a backup link contains data that tells the network the various pairs of backup ports that connect pairs of neighboring nodes. This information is gathered by the network and is constantly updated so that the network has a view of the network's overall topology regarding the availability of backup links. This data can be stored in the database of the network operation support system so that it can be provided to the origin node immediately when a failure occurs.

Furthermore, quality of service (QoS) can be assigned to each link of the network using the information from the survival message that uses C bits as a carrier. QoS can be utilized to assign priorities as to how and which data is rerouted first, or whether data is rerouted upon recovery. In particular, the QoS may include performance quality parameters such as erroneous time (seconds) or very erroneous time (seconds). Once the link is assigned a value, the data is used to run algorithms to prioritize the data and expedite the subsequent recovery process.

One aspect of the present invention is to provide special messages that are continuously exchanged between neighboring nodes prior to the occurrence of a failure in order to continuously collect data on available backup links of the network. Can be characterized as:

Another aspect of the present invention can also be characterized as providing an improved communication failure detection, isolation, recovery scheme or recovery algorithm.

Yet another aspect of the invention can be characterized as providing a method of communicating a description of an intact part of a failed path to a source node and a destination node of the failed path.

Furthermore, another aspect of the invention can still be utilized to carry traffic,
It also provides a way to identify the part of the path that has failed.

Yet another aspect of the invention is the use of reusable links that connect nodes in a failed path, such that the links or spans are used by alternative paths to restore interrupted traffic. It can be characterized as identifying the span.

Yet another aspect of the invention is to provide a method of depicting the topology of the DRA network's reserve capacity such that traffic is routed through a reserve link that functions when the network fails. And

Other aspects and advantages of the invention will be set forth in the description that follows, and will be apparent from the description and examples of the invention. The aspects and advantages of the invention are achieved and realized by means of various elements and combinations of elements pointed out in the appended claims.

In order to realize these advantages, the invention according to the object of the invention, which is described in detail by means of an embodiment, comprises in one aspect a pair of neighboring nodes of a communication network having at least one distributed restoration sub-network. Create the first C-bit alive message at the first node in the pair of neighbors identified and connected by the link,
Embed the first C-bit alive message in the C-bit of the first DS signal, look at the neighbor nodes from the first node to identify the quality of service information for the link, and then the second node from the first node of the neighbor node pair on the link. The first DS3 signal is transmitted to the node, where the first C-bit alive message identifies the first node to the second node when looking at the second node from the first node, and the quality of service of the link. It can be characterized as a method for identifying information.

According to another aspect of the present invention, the present invention includes a plurality of nodes interconnected by a plurality of links and a distributed restoration sub-network, a first node having a first unique identifier, and A second node having a th unique identifier, a link connecting the first node to the second node, and a DS3 signaling channel in the link, where the first node and the second node are It can be characterized as a communication network that can send a mutual message and a DS3 signal with link quality of service information embedded in the C bit.

It is to be understood that the foregoing general description and the following detailed description are by way of example only, by way of illustration only and not as a limitation of the claimed invention.

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate several embodiments of the invention and, together with the description, serve to explain the principles of the invention.

Other objects and advantages will be apparent from the specification and the appended claims, taken in conjunction with the following drawings.

DETAILED DESCRIPTION OF THE INVENTION FIG. 1 illustrates a portion 10 of a communication network that includes a node 12 that can communicate with, for example, nodes 14 and 16. The connection between node 12 and node 14 is
For example, not only is the link 28-30 between node 12 and node 16,
It may also be a collection of links, such as links 18-26. Node 14 and node 16
Can communicate with each other via links 32-36, which, taken together, can be considered span 38, for example.

In the following description, the following terms are used to describe the concept of the present invention. 16
33SX is a cross connect and is called a node here. The link between the nodes is DS3, which is basically the same as DS3, but may also be STS-1 which complies with another standard. The link has three STSs to form one signal
It may also be STS-3 with -l multiplexed. Also, the link is 12 STS-1
In some cases, STS-12 is multiplexed. Alternatively, there are twelve STS-12s, but actually the STSs-combined to form one large channel.
It may be 12C. However, a link is actually a unit of capacity for the purposes of this invention. Therefore, in the following description, a link is a unit of capacity that connects one node to another. A span must be understood as every link between two neighboring nodes. They are connected by a bundle of neighboring or neighboring nodes.

For purposes of the present invention, a link can be classified as active, spare, failed, and recovered.
A working link is a link that is currently carrying traffic. A backup link is a link that is currently unused but can be used. The backup link can be used any time the network wishes to use it. A failed link is a link that was up but failed. A recovery link is a recovered link, as described in detail below.

FIG. 2 is an example conceptually showing a restoration sub-network 40 including a source node connected to a destination node 48 through tandem nodes 44 and 46. In the recovery subnetwork 40, paths such as 50, 52, 54, 56 include links between these nodes as well as connections from nodes 42 to 48, for example. As shown by the recovery subnetwork 40, each path enters the recovery network 40 from outside the recovery subnetwork 40 at the origin node 42.

In an embodiment of the invention, each node 42 to 48 includes an associated node identifier. The origin node 42 has the smaller node identifier value and the destination node 48 has the larger node identifier value. In the recovery process of the present invention, the nodes compare the node identification numbers.

The present invention is a restoration sub-network 40 which is part 10 of the entire communication network.
Establish. Within the recovery network 40, there are numerous paths 50. Path 50 includes a number of links 18 that are coupled and cross-connected through node 40. The path 50 does not start within the recovery network 40, but at the customer premises or elsewhere. In fact, the path 50 begins outside the given communication network 10. However, the point where the path 50 enters the recovery network 40 is the origin node 42. The point at origin node 42 where path 50 enters recovery network 40 is access / exit port 58.

In the recovery network, a failure can occur between two tandem nodes. The two tandem nodes on each side of the fault are designated as management nodes. If there is one failure in the network, there can be two management nodes. Therefore, in a network, there may be many origin / destination nodes. There may be two origin nodes and two destination nodes. The combination of the origin node and the associated destination node is
It can be considered as an origin / destination pair. A failure can result in many origin / destination pairs.

FIG. 3 shows the concept of a management node applied to the present invention. Returning to the recovery network 40, the management nodes 62 and 64 are tandem nodes located on either side of the failed span 66. Management nodes 62 and 64 join the failed link and communicate this failure, as described below. FIG. 4 shows processing of a plurality of origin / destination node pairs when a span failure occurs in the present invention. Returning to FIG. 4, the recovery network 40 includes, for example, management nodes 62 and 64.
May include an origin node 42 that connects to a destination node 48 through. There may be multiple origin nodes, such as origin node 72, within the same recovery subnetwork. In fact, the origin node 72 is connected to the destination node 74 through the management node 62 and the management node 64. As shown in FIG. 3, FIG. 4 shows a fault 66 that created management nodes 62 and 64.

The present invention has application in each origin / destination pair of a given restoration subnetwork. However, in the following, one origin / destination pair will describe the operation of the invention. Understanding how the present invention handles a single origin / destination pair reveals how the algorithm can be extended to multiple origin / destination pairs occurring at the same time. However, in the present invention, it is also an important issue to consider that one disconnection produces many origin / destination pairs.

5A and 5B illustrate the concept of loose synchronization in the present invention. Loose synchronization allows the method and system of the present invention to operate as if all steps were synchronized according to a central clock. Known recovery algorithms fall into race conditions during recovery, making the recovery process unpredictable. Due to race conditions, the recovery configuration of a given network will depend on which message arrives first. The present invention eliminates race conditions and provides reliable results for each failure. This makes the restoration process much easier so that one can predict how the restored network will be configured.

Returning to FIG. 5A, the recovery sub-network 40 has tandem nodes 44 and 4
Includes a source node 42 connected to 6. The data flows from the origin node 42 to the tandem node 46 along the data path 76, for example. The origin node 42 may be connected to the tandem node 44 via the path 78. However, the path 80 directly connects the origin node 42 to the destination node 48. Path 82 is
The tandem node 44 and the tandem node 46 are connected to each other. In addition, pass 8
4 connects the tandem node 46 and the destination node 48. As shown in FIG. 5A, data is transferred from the origin node 42 to the tandem node 46 along path 76.
, And from the destination node 48 to the origin node 42. Further, data is exchanged between tandem node 44 and tandem node 46. Destination node 48
Uses the data 84 to pass the data to the tandem node 46 as well as the data path 8
It is directed to the origin node 42 along 0.

All of these data flows occur in one step. At the end of the step, each node in the recovery subnetwork 40 sends a step complete message to its neighbors. Continuing with the example of FIG. 5A, in FIG. 5B there are a number of step completion messages that occur within the recovery subnetwork 40. In particular, data path 7
8 between the origin node 42 and the tandem node 44, and the origin node 4 of the data path 76.
2 and the tandem node 46, and the source node 42 and the destination node 48 of the data path 80 are exchanged for step completion messages. Further, the tandem node 46 exchanges a step completion message with the tandem node 44 on the data path 82, and exchanges a step completion message with the tandem node 46 on the data path 84 and the destination node 48.

In the following description, the hop count is a part of a message moving from a certain node to a neighboring node. A hop occurs each time a message flows from a node to a neighbor node. Therefore, the hop count determines how many hops the message has taken within the recovery subnetwork.

The restoration algorithm of the present invention can be divided into a plurality of steps. The loose synchronization ensures that at each step the node processes the messages it receives from its neighbors. Loose synchronization also causes the node to send a step completion message to each neighbor. If a node has nothing to do in a step, it only sends a step completion message. When a node receives a step complete message from all neighbors, it increments the step counter associated with the node and proceeds to the next step.

When a node receives step completion messages from all neighbors,
The node proceeds to the next step in the recovery process. If you see a message that goes through a link, you can see many messages that move the link. However, the final message is the step completion message. Therefore, during step execution, many data messages are exchanged between the nodes. At the end of the step, all nodes send a step complete message to their neighbors indicating that it is appropriate to proceed to the next step since all the appropriate data messages have been sent. Basic synchronization occurs as a result of the continuous data: step complete, data, step complete, message traffic.

In practice, the operations are not synchronized as shown in the figures, but they are. In operation of the present invention, messages pass through the recovery subnetwork at different times. However, loose synchronization prevents data messages from passing through the recovery sub-network until all step completion messages have been received at the node. That is, one node is in step 3 and another is in step 4.
Is being implemented. In fact, there may be even greater step differences between nodes at some locations within the recovery subnetwork. this is,
Minimize the impact of slow nodes on the steps being performed within the recovery subnetworks.

Each step of the process of the present invention can best be considered by considering it in a numbered manner. Therefore, the process starts at step 1 and proceeds to step 2. Each step has a predefined activity and each node has its own step counter. However, there is no master clock that controls the entire recovery subnetwork. In other words, the network restoration process of the present invention can be regarded as a distributed restoration process. This node is the same for all nodes. All nodes run the same process independently, but with loose synchronization they run the same process.

FIG. 6 shows the general form of a failure notification message through the recovery subnetwork 40. For example, if the origin node 42 wants to initiate a recovery event,
The node first sends a failure notification message to tandem node 44 via data path 78, to tandem node 46 via data path 76, and to destination node 48 via data path 80. Further, as FIG. 6 shows, pass 8
The tandem node 44 sends a failure notification message to the tandem node 46 on path 82 as the destination node 48 sends to the tandem node 46 at 4.

Therefore, the process of the present invention begins with a failure notification message. The failure notification message is broadcast to the entire recovery sub-network, initiating the recovery process from one node to all other nodes. When a node receives a failure message, it sends a failure notification message to the neighbor node, which also sends a message to its neighbor. Thus, the fault notification message reaches all the nodes of the restoration sub-network. If the network has multiple failures, multiple failure notification messages may flood the recovery subnetwork.

[0064]   The first failure notification message triggers the recovery algorithm of the present invention.

Further, the broadcast of the failure notification message is asynchronous in the sense that when a node receives the failure notification message, it immediately broadcasts the message to its neighbors regardless of the timing signal. It is the fault notification message that initiates the loose synchronization process and initiates the recovery process of the present invention at each node in the recovery subnetworks. When a node initiates the recovery process, a series of events occur.

However, it should be noted that a number of events have already occurred in the recovery sub-network before the recovery process of the present invention is performed. Sending and receiving live messages exchanged by neighboring nodes is such an event.

FIG. 7 illustrates the communication of alive messages that the recovery process of the present invention exchanges on the backup link to identify neighboring nodes. Looking at FIG. 7, configuration 90
Indicates a connection between the node 94 and the node 96 via a backup link. For example, it is assumed that the node 94 has a numerical designation I'll and a port designation 11103. It is also assumed that the node 96 has a numerical designation 3 and a port designation 5. On the backup link 92, the node 94 sends a live message 98 to the node 96 to inform it of its node number 11 and port number 103. Also, from node 96, the live message 100 flows to node 94, notifying that the live message came from the port with the number 5 of the node with the number 3.

The present invention utilizes the transmission of live signals using the C bit of DS3 format messages on the restoration subnetwork 40. The available spare link carries the DS3 message and the C bit carries the special survival message. In particular, each survival message is
The node identifier and port number sending the message, the WAN address of the node,
It contains a "managed node" indicator that is used to evaluate the quality of the backup link.

One important aspect of the present invention relates to the signaling channels that occur when cross-connect nodes communicate with each other. The cross connect can perform two types of communication called in-band and out-of-band. In in-band communication, the signal travels on part of the same physical medium as the actual traffic. The communication goes through the same physical medium as the path or the same physical medium as the link. Out-of-band, you are free to pass signals between cross-connects however you like. Out-of-band signals generally require high data rates.

For example, in FIG. 7, in-band messages may be carried on the link and out-of-band messages may flow on other possible paths between the two nodes. In the present invention, some messages must flow in-band. This includes alive messages, path verification messages, and signal failure messages. There are several signaling channels available in the restoration process of the present invention, depending on the type of link involved. This includes asynchronous links such as SONET links and DS3 links.

The notable difference between SONET and DS3 links is that each uses a different framing standard and special application equipment must comply with it. It is physically impossible for the same port to operate as a SONET port and a DS3 port at the same time. In SONET signaling channels, there is a feature called tandem path overhead, which is the signaling channel that is part of the multiplexed signal. It is possible to separate this signal part from the SONET signaling channel. With the tandem path overhead, sometimes referred to as the Z5 overhead, there is the ability to send messages within the SONET channel.

In the DS3 link, there are two signaling channels. There are C bits and X bits. The C-bit channel cannot be used on the working path, only on the backup or recovery link. This is because the DS3 standard offers the option of using or not using the C bit. If C-bit format signal is used, C-bit can be used for signal transmission. However, in this case, live traffic does not use this format. Therefore, C
The bits are not available for signaling on the working channel. It can only be used on a spare or recovery link.

FIG. 8 shows the restoration subnetwork 40 from the origin node 42 to the tandem node 4.
4 shows the flow of a path validation message going to destination node 48 via 4 and 46. The path verification message 102 is sent from the origin node 42 to the tandem nodes 44 and 4
6 to the destination node 48. In particular, assume that the origin node 42 has the label 18 and the operational path 52 enters port 58. Therefore, the path verification message 102 includes the labels 18 and 53, and this information is tandem nodes 44 and 4.
6 to the destination node 48. Destination node 48 includes label 15 and output port 106 has label 29. The path verification message 104 flows to the origin node 42 via the tandem nodes 46 and 44 to identify the destination node 48 as the destination node of the active path 52.

The path verification message is embedded in the DS3 signal using the X bits that are normally used for low speed 1 bit alert signaling. In the present invention, the state of the X bit is overwritten by a short burst of data that signals the downstream receiving equipment that it is a signal. A burst is a short burst that does not disturb other equipment that relies on the traditional method of using X bits to send an alarm signal.

The present invention also limits the path verification signal in the network. DR
In a network under A control, the path validation message is embedded in the signal carrying the traffic entering the network and stripped from the signal leaving the network.
Within the network, the propagation of such signals is limited based on the DRA enabled state of each port. The path verification message identifies the origin node and the destination node. Path validation messages occur on live links that are actually carrying traffic. The path validation message originates at the origin node 42 and the recovery sub-network and traverses the tandem node until traffic reaches the destination node 48. For example, tandem nodes 44 and 46 between a source node 42 and a destination node 48.
Can read path validation messages, but cannot modify them. At the destination node 48, the path validation message is stripped from the live traffic so that it is not transmitted from the restoration subnetwork.

The present invention uses the X bit to carry the path validation message 104. One signal format in which the present invention can be used is the DS3 signal format. Although it is easy to carry path validation messages on SONET traffic, the DS3 traffic standard does not allow the use of path validation messages 104. The present invention is
Without interrupting the traffic of this signal, and without alerting the entire network,
This limitation is circumvented by adding a path verification message 104 to the DS3 frame X bits in the DS3 signal.

The DS3 standard specifies that signals are provided in frames. Each frame is X
It has special bits called bits. In fact, there are two X bits, X-1 and X-2. However, the original purpose of the X-bit was not to carry the path validation message 104. The present invention provides a path verification message with X bits. This is to avoid alarms and equipment problems that occur when the path verification message 104 is placed elsewhere. An important aspect of using X bits in the path verification message 104 in this embodiment relates to the format of the signal. This embodiment sends the path verification message 104 at a very low data rate, eg 5 bits / second. By sending the path validation message 104 with X bits very slowly,
The likelihood of alerting the network is dramatically reduced. Path verification message 1
04 is sent short burst, long latency, then short burst, long latency, and so on. The method of "hiding" the path verification message 104 in the alert allows the use of the path verification message 104 in a DS3 configured system.

FIG. 9 conceptually shows a time series of the recovery process executed by the present invention. Over time, the time domain 108 shows the state of the network prior to the failure at point 110. At the point of failure, failure notification and failure isolation events occur in timespan 112. At the end of this step, space 114
The first stage of the process begins, as indicated by. This includes, for example, a survey phase 116 with three steps 118, 120, 122.
The return phase 124 then occurs, which is at least two steps 116 and 1
Including 18. These steps are described in detail below.

Upon occurrence of a fault, the process of the present invention causes the fault notification and fault isolation phase 112.
including. Fault notification initiates the process by sending a fault notification message to the entire recovery subnetworks. Fault isolation involves determining which node is the management node. It is important to know which is the management node as the spare is in the same span as the failed span. The present invention avoids the use of this reserve as the probability of failure is very high. Therefore, fault isolation provides a way to identify which node is the management node and to locate the fault on the path.

FIG. 10 shows the flow of the AIS signal 130 through the restoration subnetwork 40. When a failure occurs between the management nodes 62 and 64, the AIS message 130
Is transmitted to the origin node 42 through the management node 62, and the restoration subnetwork 40
To leave. Also, the AIS message goes through the management node 64 and the tandem node 46 to the destination node 48 before leaving the recovery subnetwork 40. this is,
This is a normal way of transmitting the AIS message 130. Therefore, all links in the normally failed path will see the same AIS signal.

On the other hand, FIG. 11 shows “signal impairment” of AIS signal 130, signal 132 and signal 13.
4 shows conversion to 4. The SF message 132 goes to the origin node 42 where it is reconverted to an AIS message 132. The signal 134 then travels halfway through the tandem node 46 to the destination node 48, which reconverts the SF message 134 into an AIF message 130.

Therefore, FIGS. 10 and 11 show how the DS3 standard specifies operation within the restoration sub-network. In a DS3 path with one or more tandem nodes 44, 46, including a source node 42 and a destination node 48, management nodes 62 and 64 are on each side of link failure 66. AIS signal 1
30 is a DS3 standard signal indicating that an alarm is present downstream. Further, the AIS signal 130 may actually be a number of different signals. AIS
The signal 130 is passed downstream so that all nodes see the exact same signal.

There is no way in the AIS signal 130 to determine which are the management nodes 62, 64 and which are the tandem nodes 44, 46. This is because the incoming signal looks the same to each receiving node. The present embodiment takes this into account by converting the AIS signal 130 into a signal impairment or SF signal 132. When the tandem node 46 sees the SF signal 134, the tandem node propagates the signal until it reaches the destination node that returns the SF signal 134 to the AIS signal 130.

Another signal transmitted through the restoration subnetwork 40 is an ISF signal. I
The SF signal is a signal that enters the restoration subnetwork and is a signal that represents an incoming signal failure. An ISF signal is generated when the wrong signal enters the network. If it comes in as an AIS signal, it needs to be identified. The ISF signal already exists in the SONET standard. The present invention is
The SF signal is added as described above. The SF signal already exists in the DS3 standard. The present invention adds ISF signals to the DS3 standard.

As a result, the ISF signal is added to the operation of the present invention in the DS3 standard environment. SON
In operation in the ET standard environment, the present invention adds a SF signal. Therefore, in each standard, the present invention adds new signals.

In order to determine whether the incoming non-traffic signal received by the node has been asserted by an alarm in the DRA control network, a normal alert indicator signal (AIS
), A modified DS3 idle signal is passed downstream. The idle signal created by this alarm is used to signal the presence of a fault within a particular network.
The messaging embedded in the bit maintenance channel distinguishes it from a normal idle signal. The AIS signal is replaced with an idle signal to aid in fault isolation by suppressing downstream alarms. Upon leaving the network, these signals are AI'd to maintain operational compatibility with equipment outside the network.
It can be returned to the S signal. A similar approach has been implemented in SONET networks, where STS-N AIS signals are replaced with ISF signals and ZS bytes carry alarm information.

According to another aspect of the invention, one-way failures can also be managed. In a distributed recovery environment, a failure in one direction of a bidirectional link is handled by first ensuring that the alarm signal persists for a period of time, then propagating the idle signal back in the active direction. . The idle signal produced by this alarm is distinguished from the normal idle signal by the messaging embedded in the C-bit maintenance channel to signal the presence of the far-end reception impairment. In this way, the management node simplifies recovery switching by quickly being identified and handling unidirectional failures as if they were bidirectional failures.

FIG. 12 shows broadcasting of a failure notification message from the management nodes 62 and 64. As shown in FIG. 12, the management node 62 sends the failure notification to the origin node 42 as well as the tandem node 136. Tandem node 136 also broadcasts a failure notification message to tandem nodes 138 and 140. Further, the management node 64 transmits the failure notification message to the tandem node 46, and the tandem node 46 further sends the failure notification message to the destination node 4
8 is transmitted. The management node 64 also broadcasts a failure notification message to the tandem node 140.

FIG. 13 shows a time chart of the first iteration following fault isolation. In particular, FIG. 13 shows a time chart of the survey phase 116 and the return phase 124 of iteration 1. Further, FIG. 14 shows a partial time chart of the completion of iteration 1 and iteration 2. As shown in FIG. 14, iteration 1 includes survey phase 116, return phase 124, maxflow phase 142,
It includes a connection phase 114. Maxflow phase 142 includes one step 146. It should be noted that iteration 2 connection phase 114, represented by region 148, includes six steps from 150 to 160 and is performed concurrently with iteration 2 probe phase 162. Furthermore, it should be noted that the return phase 164 of iteration 2 also includes the six steps 166 to 176.

Each iteration involves a probe, return, maxflow, connect phase. The recovered traffic addressed by the connect message and the remaining non-recovered traffic carried by the probe message are separate sets. Therefore, the synchronous DRA
Propagating or combining these messaging steps in a process at the same time does not cause a collision. This, along with fault queuing, makes the recovery process well coordinated and fast.

The iterations will be longer and contain many steps in subsequent iterations. This is because the alternative path is searched in the subsequent iterations. The path has a certain length when viewed from the hop. The path can be 3 or 4 hops. In the first iteration, the hop count is set to 3, for example. That is, an alternative path of 3 hops or less is searched. The next iteration looks for an alternative path that is less than 6 hops.

Setting a limit on the hop count per iteration improves the efficiency of the process of the invention. In the system of the present invention, the number of iterations and the number of hops per iteration are configurable. However, these values may be preset depending on the degree of flexibility required by a particular implementation. It should be noted, however, that configurability also adds complexity. Increased complexity can result in inappropriate or problematic configurations.

FIG. 15 shows the survey phase 116, which is the beginning of the first iteration 114, for a detailed description of the survey phase. FIG. 16 shows one origin node 4
A portion 170 of the recovery network is shown to represent the idea that a 2 may have multiple destination nodes. In particular, the destination node 180 is the management node 62.
It may also be the destination node of the origin node 42 passing through and 66. Also, as before, the destination node 48 is the destination node for the origin node 42. This occurs because the two working paths 182 and 184 both start at the origin node 42 and pass through the recovery subnetwork portion 170. In the probe phase, the message begins at the originating node and travels out through the recovery subnetwork. Each probe message is stored and transferred in a loosely synchronized manner. Therefore, if the node receives the message in step 1, it forwards the message in step 2. The neighbor node receiving the probe message in step 1 transmits the probe message to the neighbor node in step 2. Since the present invention uses loose synchronization,
It does not matter how fast the message is transmitted from one neighbor to another. The message is just sent in the next step anyway.

If the length of the survey phase is 3 steps, the survey phase will flow 3 hops more, but not more. Although the description that follows deals with one origin / destination pair, there may be origin / destination pairs in the restoration subnetwork 40 that perform similar or similar functions at the same time. If two nodes send a probe message to one neighbor, only the message that the neighbor receives first is transmitted by the neighbor. The second message received by the neighbor node is
Recognized but not transferred. Therefore, the first node to arrive at a neighbor with a probe message is usually the node closest to the neighbor. When the probe message reaches the destination node, the probe message stops. This step reveals the amount of spare capacity in the recovery sub-network between the source and destination nodes.

Due to the gradual synchronization, the first message reaching the source node 42 and the destination node 48 takes the shortest path. There are no race conditions within the operation of the present invention. The survey message contains the distance between the source node and the destination node. This distance, measured in hops, is always less than or equal to the number of steps allowed in a given survey phase. For example, if the destination node is 5 hops away from the origin node on the shortest path, the probe phase with a 3 hop count limit will never produce a return message. On the other hand, the probe phase, which has a limit of 6 hops count, returns 5 hops count information in the return message.

The survey message includes the identification of the origin / destination pair for identifying the node that sent the survey message and the destination node that received the survey message. There are also demands for capacity. For example, the message states that there is a need for 13 DS3s because 13 DS3s have failed. Actually, not only DS3, but also STS-1 and STS-12
There are also C etc. However, the point is that a certain amount of capacity is required. At each node that the probe message passes through, a request for capacity is recorded. The survey phase ends when the predetermined number of steps is completed. Thus, for example, if the survey phase continues for three steps, then the survey phase ends in step 4. This clearly ends the research phase.

FIG. 17 shows the restoration sub-network 40 in one origin / destination pair including the origin node 42 and the destination node 48. In the restoration subnetwork 40, the origin node 42 takes step 1 at the beginning of the survey phase and sends a survey message to the tandem node 44, the tandem node 46, and the tandem node 186. In step 2, tandem node 46 sends a probe message to tandem node 188 and destination node 48. In step 2, tandem node 44 sends an inquiry message to tandem node 46, which sends an inquiry message to tandem node 188 and destination node 48. The tandem node 186 then sends an inquiry message to the tandem node 46 and the destination node 48. It should be noted that the probe message of step 2 from tandem node 44 to tandem node 46 and from tandem node 186 to tandem node 46 is not forwarded by tandem node 46.

FIG. 18 shows a time chart of the next phase, the return phase 24, of the restoration process of the invention, which comprises three steps 126, 128, 129 during the first iteration.

FIG. 19 shows the return phase of the invention during the first iteration. Starting at the destination node 48, in step 4, the return message flows on path 192 to tandem node 46 and path 190 to tandem node 186. In step 5,
The return message follows path 76 to the origin node 42. The return message from the tandem node 186 flows on the path to the origin node 42.

In the return phase, the return message follows the same path as the corresponding probe phase, but in the opposite direction. The message starts at the destination node and flows to the origin node. In addition, the return phase messages are loosely synchronized as previously described. The return phase message contains information about the number of backup links available to connect the source node to the destination node.

In the return phase, information about available capacity is communicated to the originating node. Starting from the destination node 48, and on the way back through each tandem node 44, 46, 186 back to the origin node 42, the return message gets longer and longer. Therefore, the return message contains information about the capacity capacity available in each span on the way to the origin node. As a result of the return message received, it becomes possible to establish at the origin node a map of the recovery network showing where there is spare capacity available for recovery.

FIG. 20 shows a tandem node 44 connecting to a tandem node 46 via a span 38. The span 38 has six links 32, 34, 36, 196,
Note that it contains 198, 200. 21 and 22 show the allocation of links between tandem nodes 44 and 46 in accordance with the preferred embodiment of the present invention. Considering initially FIG. 21, the span 38 between tandem nodes 44 and 46 in the previous survey phase, as scenario 202 illustrates,
The first probe message (5, 5 declaring the need for four links to node 46
3) is assumed to be carried. Scenario 204 also shows a message (11, 2) requesting eight links out port 2 of tandem node 44.

FIG. 22 shows how this example assigns the six links of span 38. In particular, in response to survey messages from scenarios 202 and 204 of FIG. 21, we know that tandem nodes 44 and 46 respectively assign three links to each origin / destination pair. Thus, three links between tandem nodes 44 and 46, such as links 32, 34, 36, are assigned to the (5,3) origin / destination pair. For example, links 196, 198, 200 may be assigned to the origin / destination pair (11, 2).

FIG. 23 shows the result of the return phase of the present invention. The restoration subnetwork 40 includes, for example, not only the tandem node 44 but also the origin node 42 and the tandem nodes 208, 210, 212. As FIG. 23 shows, the return message is accompanied by a map of the route followed by itself and the capacity capacity allocated in each span. Origin node 42 collects all return messages. In this example, four links between origin node 42 and tandem node 44 were thus assigned between origin node 42 and node 208. Tandem node 208 has been assigned 10 links to tandem node 210. The tandem node 210 is assigned three links with the tandem node 17. Further, the tandem node 17 is assigned seven links with the tandem node 44.

The next phase in the first iteration of the process of the invention is the Maxflow phase. Maxflow phase is a one-step phase,
As shown in FIG. 24, it is the seventh step of the first iteration. In the embodiment of the present invention, all work in the Max Flow phase is performed by the origin node 42.
At the beginning of the Max Flow phase, each origin node has a model of a part of the network. This is the part assigned to each origin / destination pair by the tandem node.

FIG. 25 shows the restoration sub-network model 214 in the origin node 42. Then, this model shows which part of the restoration subnetwork 40 is allocated to the pair of the source node 42 and the destination node 48. In particular, the model 214 shows that eight links are assigned between the origin node 42 and the tandem node 46, and 11 links are assigned between the tandem node 46 and the destination node 48. Further, model 214 shows that three links may be assigned between tandem node 46 and tandem node 186.

Therefore, as shown in FIG. 26, in the maxflow phase 142 of this embodiment, the origin node 42 calculates an alternative path through the restoration subnetwork 40. Calculated using the Maxflow algorithm. Therefore, the MaxFlow output of FIG. 26 is a flow matrix showing the desired flow of traffic between the source node 42 and the destination node 48. Note that the MaxFlow output does not use tandem node 44 or tandem node 188.

FIG. 27 illustrates the breadth-first search used by the MaxFlowPhase 142 to find a route through the MaxFlowPhase output. In the example of FIG. 27, the first route starting at the origin node 42, to the tandem node 186, to the tandem node 46, and finally to the destination node 48 allocates two units. The second route from the origin node 42 to the tandem node 186 and finally to the destination node 48 allocates three units. The third route allocates eight units from the origin node 42 to the tandem node 46. From tandem node 46, these eight units go to destination node 48.

The last phase of the first iteration in the process of this example is the connection phase 1
44. In the example described here, the connection phases are 150 and 1 respectively.
It includes steps 8 to 13 of the first iteration with reference numerals 52, 154, 156, 220, 222.

The connection phase is loosely synchronized, as described previously, so that each connection message moves one hop in one step. The connect phase 144 overlaps the probe phase 162 in each subsequent iteration, except for the last iteration. The connection phase 144 distributes information about the connection to be established, for example, from the origin node 42 through the tandem nodes 46 and 186 in order to reach the destination node 48.

In the Connect phase 144, the message follows the same route as identified in the MaxFlow phase 142. Therefore, as FIG. 29 shows, the first message M1 flows from the origin node 42 through the tandem node 186 and the tandem node 46 and finally to the destination node 48, which represents a two unit connection. Similarly, the second message M2 is flowing from the origin node 42 through the tandem node 186 directly to the destination node 48, which connects the three unit flow path. Finally, the third connection message M3 is transmitted from the origin node 42.
It spreads to the destination node 48 through the tandem node 46 from k to allocate eight units. The connect phase 144 is synchronized so that each step of the message moves one hop.

A number of extensions are required to implement the process of the present invention in an existing or live network. This extension considers the existence of hybrid networks in which some nodes have both SONET and DS3 connectivity. Further, the present invention provides different priorities for the working paths and different qualities for the backup links. Fault isolation reveals particular challenges in existing operating environments that the present invention addresses. Limited reuse and backup links connected to paths are additional features provided by the present invention. Prohibition functions such as path prohibition and node prohibition are additional functions to the present invention. The present invention also provides the ability to interface with existing recovery processes and systems, such as coordination with existing recovery algorithms, processes and like systems. To ensure proper operation of the present invention, this embodiment provides a practice function for simulating and simulating the restoration process without making an actual connection for subnetwork restoration. In addition, the present invention also includes a termination timer and an emergency shutdown closure function to control or limit malfunctioning of the recovery sub-network. In addition, the present invention handles real-world situations such as anisotropy and disconnection that exist in communication networks. This embodiment also includes holdoff trigger and step timer functions as well as hop count and software modification check mechanisms to ensure proper operation.

30 to 33 show how the present invention addresses hybrid networks. Hybrid networks are asynchronous links and SO
It is a combination of NET links. As a limitation in the method of the present invention for handling a hybrid network, SONE having loads other than all working paths DS3
It may be either a T-path or a DS3 on an asynchronous and SONET working path with a DS3 access / output port. Otherwise,
For example, sending a path validation message within the recovery subnetwork 40 may not help. Referring to FIG. 30 and FIG. 31, the restoration sub-network 40 passes through the SONET tandem port 44, the SONNET tandem port 46, and finally connects to the SONET destination A / E port 48 by the SONET origin A / E port 42. May be included. In FIG. 31, the origin A / E port 42 is a DS3 port, the tandem port 44 is a SONNET port, and the tandem port 46.
Is the DS3 port. Port 106 of destination node 48 is a DS3 port. In a hybrid network, the origin node 42 requests different types of capacity during the probe phase. In the return phase, Tandem Port 44
And 46 allocate different types of capacity.

An important part of the connect phase 144 is to correctly convey the type of traffic that needs to be connected in the connect message. As described above, this includes, for example, DS3, STS-1, OC-3, OC-12C routing. All implementation details need to be tracked for different types of traffic. To that end, the present invention provides different priorities for working paths and different qualities for backup links. In an embodiment of the invention, active traffic is prioritized midway between high priority traffic and active low priority traffic.

SONET traffic includes another rule for addresses. For example,
The SONET path is basically three STS-1 ports, including an OC-3 port with STS-1 representing SONET equivalent to a DS3 port. Therefore, an OC-3 node can carry the same traffic as three STS-1s. Also, an OC-3 node can carry the same traffic as a set of three DS3s or three STS-1 and DS3 nodes. Moreover, OC-3 nodes can carry the same traffic as STS-3. Therefore, OC-3
It can carry the same traffic as three DS3s, three STS-1, or one OC-3. At that time, OC-12 may carry OC-12C. Also, up to 4 OC-3 ports or 12 STS-1 ports, or 12
Can carry the same traffic up to the DS3 port. With all possible combinations, it is important to ensure that the channels with the highest capacity pass the highest capacity first.

Therefore, according to an important aspect of the present invention, a hybrid network service can be provided. Hybrid network means SONET link and DS3
A network that includes both asynchronous links, such as 3 links. The present invention is
Restore the recovery sub-network 40, which includes both types of links. SONET
The standard makes SONET traffic backwards compatible with DS3 traffic. Therefore, SONET links may include DS3 signals within them. A restoration sub-network containing both SONET and DS3 can carry DS3 signals if both the source A / E port 42 and the destination A / E port 48 are DS3 ports. Otherwise, there is no way to send the path validation message 104 within the recovery subnetwork 40.

In hybrid networks as well as in pure networks, the survey message requires capacity for network recovery. This message is
Specify what capacity you need. It is important to know whether DS3 or SONET capacity is needed. In addition, another type of SONET
Since there are also links, it is also necessary to identify the type of SONET format required. In the return phase, the tandem node allocates capacity to the origin / destination pair. This requires the tandem node to know the types of spares available in the span. There are DS3 and SONET as spares. Capacity can be assigned knowing what types of spares are available.
Therefore, it is necessary to add expansions that allow different kinds of capacity in the execution of the investigation phase and the return phase. Therefore, the probe message of the present invention includes a request for capacity to determine how many DS3 links and how many SONET links are needed. For example, STS-1, STS-3C, or STS-12C may be required. Further, in the return phase, the return message must include information that the network has multiple types of capacity. You must be aware of this rule as traffic passes through your network. For example, a failed working link in DS3 can be replaced by a SONET link, but not the other way around. In other words, the DS3 cannot replace the SONET failure working path.

32 and 33 illustrate this feature. For example, referring to FIG. 32, the origin node 42 has five DS3s, three STS-1, two STS-3, 1
An inquiry message to the tandem node 44 can be created requesting one STS-12 (c). As shown in FIG. 33, from the return phase, the origin node 42 receives the return message from the tandem node 44, and the origin node 42 receives 5 DS3s, 1 STS-1, 1 STS-3 (c). , 0 STS
Notify that you have received -12.

In the hybrid restoration sub-network 40, the present invention routes the OC-12C fault working capacity first on the OC-12 protection link in the MaxFlow phase. Thereafter, the MaxFlow phase routes the OC-3C fault working capacity on the OC-12 and OC-3 protection links. Next, the present embodiment describes OC-12.
, OC-3, and the STS-1 spare link to route the STS-1 failed working link. Finally, Max Flow Phase is OC-12, OC-3, ST
S-1 and then the DS3 protection link routes the DS3 failed working link.
In the connection phase, the restoration sub-network of the present invention responds to the hybrid network so that the tandem node receives instructions to cross connect multiple types of traffic.

FIG. 34 concerns the characteristic of the invention of assigning different priorities to working paths and different qualities to spare links. The embodiment of the present invention has 32
Includes level priorities. The priority configuration is performed at the origin node 42, for example. In addition, the preferred embodiment provides four levels of quality for the spare link as follows: A SONET 1-to-N protection spare link in a span with no failed links has the highest quality. The next highest quality is SON in spans with no failed links
It is a 1-to-N protected port of ET. The next highest is the SONET 1-to-N protected port of the span that has the failed link. The lowest quality is SONET 1-to-N protected ports on spans that have failed links.

In this configuration, different paths are associated with different priorities, and spare links are associated with different qualities. At some stages of using prioritized working paths and different quality spare links to use this process, it is possible to simplify different levels of priority and different levels of quality into two, high and low. it can. For example, a high priority working link is a link having a priority of 1 to 16, and a low priority working link is a link having a priority of 17 to 32. For example, a high quality reserve is a quality 1 reserve and a low quality reserve is a quality 2-4.

In assigning various priorities and qualities, the present invention can provide a method for restoring traffic through the restoration sub-network. For example, the present invention
It is possible to try to restore a high-priority failed working link of a high-quality backup link as soon as possible. It then attempts to recover a high quality failed working link with a low quality spare. The low quality backup link is then tried to recover the low quality working path, and finally the high quality backup link recovers the low quality failed working path.

To carry out this function, the invention adds a special iteration at the end of the normal iteration. The special iteration has the same number of steps as the previous iteration. However, this feature gives priority to working paths and quality to protection links. Figure 3
4, during normal iterations, the present invention restores a high quality working path with a high quality backup link. In a particular iteration, the invention operates on a high priority working path of a low quality backup link, then a low priority path working on a low quality backup link, and finally on a high quality backup link. Restore medium low priority paths. This involves several additional runs of the MaxFlow algorithm.

The network recovery process of the present invention, including the inquiry, return, and connection messaging phases, is repeated multiple times for a single failure, sequentially increasing the hop count limit. The first iteration is performed only on high priority traffic. Subsequent or additional iterations are used to recover the remaining lower priority traffic. This approach prioritizes high quality traffic from the path length perspective.

35 to 37 are diagrams illustrating in detail how the present invention handles fault isolation. In FIG. 35, a backup link 92 is provided between the tandem nodes 44 and 46.
Exists. Between the management nodes 62 and 64 is a working link 18 with a fault 66 and a backup link 196. When a backup link, such as backup link 196, is on a span, such as span 38, that has a failed working link, the backup link is more than a backup link, such as backup link 92 on a span that has no failed links. With low quality. In FIG. 35, the backup link 92 between the tandem nodes 46 and 48 is part of a span that has no failed links. Therefore, in this example, the backup link 92 has a higher quality than the backup link 196.

Within each node, there are specific criteria that sort the list of backup ports and paths to be recovered. This provides consistent mapping and highest priority allocation to the highest quality recovery paths. In particular, spare ports are sorted by type (eg, STS-12 or STS-3 bandwidth), then quality, and third by port label number. The restored paths are first sorted by type and then by the assigned priority value. The quality of one recovery path is limited by the lowest quality on the path.

In addition to these classification criteria, a process of allocating traffic to a spare port while optimally utilizing high capacity and high quality resources is performed on these lists multiple times. This includes, for example, all other STS-12 and STS
-3 High priority ST to STS-12 remaining after traffic is assigned
Includes S-1 stuffing.

In performing the recovery process, rules determine the appropriate way to handle working paths of varying priorities and spare links of varying quality. In an embodiment of the invention, there are 32 priority levels, for example. The priority of live traffic can also depend on business-related issues such as who the customer is, how much the customer pays for communications services, and the nature of the traffic. High priority working channels are more expensive than low priority channels. For example, "in operation" is a priority assigned based on these considerations. Such default configuration information can be stored at the origin node of the recovery subnetwork. Therefore, the priority information is saved for all the paths of the origin node. There is no functional difference between the high priority working path and the low priority working path, but the traffic on the higher priority working path is restored first and the lower priority working path is restored. The traffic on the inside path will be restored later.

This embodiment includes four qualities of the spare link. The quality of the backup link is related to two factors. The link may or may not be protected by other protection schemes. Given the priority of the faulty working path and the quality of the protection link, the present invention uses several rules. The first rule seeks to recover a failed priority working link that has a high priority with high and low quality spare links. The following rule attempts to recover a low priority working failed path with high and low quality backup links. The third rule attempts to recover the low priority working failure path of the low quality backup link. Finally, the low priority working paths of the high quality and low quality backup links are restored.

The present invention allows to know when a node became a management node. However, since there are no live messages on working links, the management node does not know in which span the faulty link is. Referring to FIG. 36, the management node 64 knows that the management node 62 is at the other end of the backup link 196. However, the question is whether the management nodes 62 and 64 can know if the working link 18 with the fault 66 and the backup link 196 are in the same span, because the management node 62 also Management node 64
This is because the link 18 in operation does not know in which span it is.

FIG. 37 illustrates how the present invention overcomes this limitation. For example, the management node 64 sends a non-management tandem node 46 with a "management node" flag in a live message sent by a backup link. Also, both management node 64 and management node 62 send the "management node" flag to each other over the spare link 196. If a non-managed node, such as the receiving tandem node 46, itself is not a managed node, that node ignores the "managed node" flag. Otherwise, the receiving node determines that the failure lies between itself and the receiving management node receiving the "management node" flag.

However, since it may be fooled by a span that does not have anisotropy or reserve,
There are some limitations to this approach. However, the worst that can occur is that the alternate path may be placed on a faulty link, that is, a span with a poor quality reserve.

The present embodiment provides this function by including the "management node" flag in the live message. Recalling that the management node is a node at either end of the failed link, the "management node" flag is set when the management node is identified. If the flag is on the backup link, this means that the neighboring link is the managing node. That is, the node is near the fault. If the node receiving the flag is also the managing node, the spare is in the span containing the failed link. Thus, a managed node that sends a flag to an unmanaged node will never get that flag back from the unmanaged node,
The backup link is not in the faulty span.

38 to 42 illustrate the limited reuse feature of the present invention. The present invention has a limited reuse function. The restored link is associated with the limited reuse feature. Given the faulty path, the link that is restored is between the two nodes. The restored link is a good link, but on the failed path. FIG. 38 shows a recovery subnetwork 40 that includes a source node 42 of link 18 and connects to destination node 48 through management nodes 62 and 64. The fault 66 exists between the management nodes 62 and 64. The limited reuse feature of the present invention also includes what is done on a restored link, such as link 224.

There are at least three reuse modes in the present invention. One reuse mode is simple no reuse. This prevents the recovery link from being used to carry alternate path traffic. Another reuse mode is unlimited reuse, which allows the recovery link to carry alternate path traffic in any way. Another reuse mode provided by the present invention is limited reuse. Limited reuse allows the recovery link to carry alternative path traffic, but only traffic that was being carried before the failure.

FIG. 39 illustrates the concept of restricted reuse used by the present invention. Link 18
Enters the origin node 42, passes through node 226 of links 228 and 230, and passes through management node 64 and recovery link 48.

Limited reuse involves modification of the probe and return phases of the present invention, where the process determines where in the network the recovery link is. The process finds the restored link and sends this information to the origin node. The origin node collects information about where in the network the restored link is to develop a map of the restored links in the restored sub-network. The tandem node sends information about the location of the reuse link directly to the originating node through the wide area network.

40 to 42 show how the present embodiment implements limited reuse. Referring to the portion 40 of the restoration sub-network of FIG. 40, the origin node 42 is the tandem node 44 via link 78, the tandem node 46 via link 82, the tandem node 186 via link 84, and the link 19
Connect to destination node 48 via 0. Tandem node 46 and tandem node 18
It should be noted that between 6 there is an obstacle 66.

In order to realize the limited reuse in the present invention, in the investigation phase and the return phase, the origin node 42 acquires the map of the link to be recovered. Thus, the recovery links 232, 234, 236 are saved at the origin node 42, as FIG. 40 shows within the origin node 42. This map is created by sending in-band and reuse messages during the probe phase on the recovery link from the management node to the source and destination nodes, such as source node 42 and destination node 48. Therefore, as shown in FIG. 41, in the probe phase, the reuse message is issued from the tandem node 46 to the tandem node 44 and from the tandem node 44 to the origin node 42. From the tandem node 186, the reuse message goes to the destination node 48.

In the return phase, as shown in FIG. 42, the destination node sends the information acquired through the reuse message to the origin node in the return message. Therefore, as FIG. 42 shows, the destination node 48 sends the return message and the reuse message to the tandem node 46 on link 192. In response, tandem node 46 sends a return message and a reuse message over link 76 to source node 42.

In the restricted reuse function, in the max flow phase, the origin node 42
Knows about recovered links and "pure" backup links. When the origin node executes the MaxFlow algorithm, the recovered link is populated with a pure backup link. When a breadth-first search is performed, the present invention does not confuse another working failure link and recovery link on the same alternate path.

Another function of the invention relates to the protection link connected to the path. This path often has an idle or test signal when the protection link is connected to the path. If the backup link has a test signal, it is not possible to distinguish it from the working path. In that case, the present invention avoids this by using a "busy" signal on the backup link.

In the Max Flow phase, we have found what can be considered a pure backup link. The origin node also receives information about the recovery link, which according to the invention is restricted to limited reuse, when the maxflow algorithm is executed in the maxflow phase of the invention, whether the link is initially pure or A recovery map of the recovery subnetworks is created using pure backup and recovery links, whether spare or restored.

Another aspect of the present invention is a function called path prohibition. 43 and 44 show the path prohibition function of the present invention. For various reasons, it may be desirable to disable network recovery protection on one port of a node. It may also be desirable later to turn recovery protection back on without turning off the entire node. In short, I want to turn off a port and then turn it back on again. This is what is desired when maintenance of a particular port is needed.
When performing such maintenance, it is desirable not to automatically activate the recovery process of the present invention. The present invention provides a means to turn off subnetwork recovery on a particular port. Therefore, as shown in FIG. 43, the origin node 42 includes the path 2 to the tandem node 44. Note that there is no link between 44 of nodes 42. This indicates that the recovery process of the present invention is prohibited on the link 240 between the origin node 42 and the tandem node 44. On the other hand, an active node exists between the origin node 42 and the tandem node 46. The link 76 indicates that the recovery process of the present invention will be restored later.
This pass indicates that it is not prohibited.

In the path prohibition function, the process of the present invention prohibits recovery on the path by blocking the recovery process at the beginning of the investigation phase. The originating node either sends no probe message or sends a probe message that does not require the ability to restore the forbidden path. This is an instruction to go to the origin node. Thus, in the case of a path ban, the process of the present invention tells the originating node 42 to ban recovery on the path, for example by sending a message over the associated wide area network.

Therefore, referring to FIG. 44, the tandem node 46 sends a path prohibition message to the origin node 42. For example, tandem node 46 receives a TL1 instruction that temporarily instructs the port to inhibit the recovery process. The tandem node 46 also sends a message for that path to the origin node 42 via the wide area network, as indicated by arrow 246.

Since the tandem node 46 is a part of the path verification message, the tandem node 46 knows the internet protocol address of the source node and sends the path prohibition message 246. Executing this function involves a protocol. Its purpose is to cover situations where a node fails when the path is forbidden.

Another feature of the present invention is the permission to prohibit a node. In the node prohibition function,
The recovery process of the present invention can be temporarily prohibited at a node. This is for example T
This is done by the L1 command. The node keeps sending step completion messages under this condition. Further, the simulation function operates on the node of this condition.

In order to support the use of conventional field technologies such as node port inspection access and path loopback functionality, the recovery process can be implemented locally so that inspection signals and alarm conditions can be achieved without triggering the recovery process. It must be able to be disabled. With this method applied to a particular path, a port instructed to enter check access, loopback, or DRA invalid mode tells the origin node of the path to suppress DRA protection on the path. Further, the present invention includes the functionality of automatic timeout in invalid mode and suppression of automatic loopback detection / recovery algorithms when a port receives an in-band signal with its own local node ID.

Direct communication between nodes is done via a dedicated wide area network.
This approach avoids the use of existing in-band and out-of-band call processing signals and network control links for speed and simplicity. In addition, the WAN approach provides certainty with diversity.

The activation mechanism of the distributed recovery process applies a verification timer to the collection of alarm inputs, always holds the number of verified alarms, and creates a trigger output when the count exceeds a predetermined threshold. This approach reduces false or premature DRA activation and provides automatic protection switching with the opportunity to recover individual link failures.
This approach also allows the trigger sensitivity to be locally adjusted based on the number and co-occurrence of multiple alarms.

The preferred embodiment provides a step completion timer in synchronous DRA. For each DRA process started in the network node, logic is provided to automatically terminate the local DRA process if it cannot receive a step completion message within a certain time period monitored by the failsafe timer. . The process is also terminated by loss of alive signal across the inter-node WAN link, normal termination of the last DRA iteration, exhaustion of all available spare ports, or operation support system overwrite command.

Another aspect of the present invention is a method of handling a relocation failure event in a DRA. In a protected sub-network, an initial link failure, or a collection of failures that occur at about the same time, also initiates a series of DRA processing phases with the message flow of the network. Other disconnects that occur during messaging similarly initiate the recovery process, creating confusion and uncontrollable contention for spare resources. This method is an improvement over conventional methods. In particular, during the probe and return messaging phase, any subsequent disconnection will be "upgraded" until the next probe phase.
I'm kept waiting. " Furthermore, in the multiple-iteration approach, the probe messaging for new disconnects will be sent during the final probe / return / connect iteration for the previous disconnect
Be stopped. This last, stopped disconnect causes a new, separate launch of the DPA process.

The present invention includes a failure notification message that contains information about software modifications and the contents of the hop count table that are assumed to be the same among all nodes. A node that receives such a message and finds that the content of the local software modification or hop count table does not match the content of the incoming notification message is not eligible for further DPA processing by itself. However, the node that knows the mismatch and locally disables the DPA continues to transmit the failure notification message.

The present invention provides a method of auditing recovery process data in a node, including protecting and verifying the contents of data tables in all nodes of the recovery protection network. In particular, such data may include setting values such as node ID, WAN address, hop count sequence table, defect threshold. This method includes disabling the recovery process node in the operation support system, writing out and verifying the setting data contents in each node, and reactivating the recovery process when all nodes have accurate data. Be done.

In data transport networks that use the distributed recovery approach, fault simulation can be performed in the network without interrupting normal traffic. This process is an initial broadcast of a description of the failure scenario, a modified DRA message indicating it is a "training" message, training can be terminated if an actual failure occurs during the simulation, node Including the logic inside.

Another aspect of the invention is the ability to coordinate with other recovery processes such as, for example, the RTR recovery system. This is the subject of the present invention, since in the present invention the ports protected by the inventive restoration process are often also protected by other network restoration algorithms.

Yet another aspect of the invention is a training function. The training function of the recovery process of the present invention has two purposes. The first is a sanity check to ensure that the recovery process is working properly. The other is capacity design training that determines what the recovery process does in the event of a link failure. In the present invention, the training function runs the same software as the recovery process during subnetwork recovery, with one exception. The training function does not perform the connection. Therefore, when the connection is executed, the connection remains unexecuted.

The training function basically produces the same report as in the case of a link failure. Unfortunately, due to the limitations of in-band signaling, there are some messages that are actually exchanged, but not during training. Therefore, the training function needs to provide some information in this untransmitted message. However, this enables the desired training function.

Another aspect of the invention is the function of an expiration timer or emergency closure. Termination timers and emergency closures provide protection from software bugs and flaws. If the recovery process of the present invention malfunctions due to a software problem causing an instruction to stall or jump somewhere, the recovery sub-network needs to be released. Termination timers and emergency closures provide this function. The termination timer is activated if the recovery process causes a certain maximum allowed time. By setting the maximum operating time, the recovery network will only run for 30 seconds, for example. After 30 seconds, the recovery process turns off.

Emergency closure is similar to the end timer, but is initiated manually. For example, in the present invention, the TL1 command can be entered to close the recovery process. Thus, the emergency closure feature provides another level of protection that supplements the expiration timer.

Out-of-band signaling allows the message to be delivered on any available communication channel. To this end, the present invention uses a recovery process wide area network. Due to the invention, some messages are sent out of band.
This includes survey messages, return messages, connect messages, step completion messages, as well as practice messages related to the training features of the present invention. The wide area network of the present invention operates under the TCP / IP protocol, although other protocols and other wide area networks can be used. In order to use a wide area network in the practice of the present invention, it is necessary to obtain access rights to the network. In the present invention, access to the wide area network is via the Ethernet ports of the two local area networks. Two Ethernet ports allow communication with wide area networks. In an embodiment of the invention, Ethernet is half-duplex in the sense that the recovery sub-network sends data in one direction on one Ethernet and information flows to the recovery sub-network in the other Ethernet port in the opposite direction. The wide area network of the present invention includes a backbone that provides the broadband portion of the wide area network. The backbone contains the same networks that the restoration subnetwork protects. Therefore, a failure in the recovery subnetwork
It may also disconnect the wide area network. This can also make wide area networks fragile.

Therefore, there may be better wide area networks used in the present invention. For example, reserve capacity can be used as a wide area network.
In other words, there may be reserve capacity in the network that is used to build the wide area network itself. This provides the necessary signal flow for the types of messages mentioned above. In the present invention, the connection via the wide area network is made automatically.

In the case of the cross-connect of the present invention, there is a control system including a large number of computers in the cross-connect switch. Cross-connects can also have hundreds of computers. These computers are hierarchically connected at three levels in this example. Computers that are processing-centric operate at the lowest level or layer 3. Another layer of computers controls the shelves of cards, for example. This computer occupies layer 2. The layer 1 computer controls the layer 2 computer.

The layer 1 computer executes the instructions of the recovery process of the present invention. This computer is centralized on a particular shelf where all Layer 1 computers are gathered in one place with the computer executing the recovery process instructions. Since the computer performing the recovery process of the present invention is a layer 1 computer, the computer itself cannot send in-band messages. If you want to send an in-band message, the message goes through the layer 3 computer. this is,
This is because the layer 3 computer controls the local card including the connecting cable. Thus, in-band messages are typically sent and received by layer 2 and / or layer 3 computers and not by layer 1 computers that execute recovery instructions for the process of the present invention.

Fault isolation is also performed by the layer 2 and layer 3 computers in the cross connect. This is because fault isolation requires a change in signal in the optical fiber. This must be done by the lower layer machines. In addition, a port that is a DS3 port or SONET port has a lower layer state tracked by the processor. Therefore, there is basically a division of labor between the layer 2 and layer 3 computers and the layer 1 computers that execute the instructions for the recovery process of the present invention.

Referring to FIG. 45, an end-to-end path in a communication network equipped with a distributed restoration algorithm is shown, including the origin node and destination node denoted by O and D respectively. . The origin node and the destination node are connected to each other via a large number of nodes, for example, intermediate nodes N1-N5. Each of these nodes was actually manufactured by Alcatel Company1.
It is a digital cross connect switch such as a 633-SX switch. Each of these switches has a number of ports to which a number of links are connected to connect each switch with other switches in the network. For simplicity of explanation and example, each neighbor pair of nodes is represented by 302, 304, 306, 3 as shown in FIG.
They are connected by spans or links such as 08, 310, 312. As is known, each span can have many links, and each neighboring pair of nodes can actually have many interconnected spans. For simplicity of explanation, no other node of the communication network is connected to the path of FIG.

As is clear from FIG. 45, the fault has occurred in the span 306. Such a failure is due, for example, to the disconnection of one or more links in the span. Alternatively, in the worst case, the entire span may have been cut. As we all know,
When a fault occurs, the nodes containing or sandwiching the fault 314 are the first nodes to receive the alarm signal, and the alarm signal is sent by these nodes to the downstream nodes. Thus, as soon as fault 314 occurs in span 306, management nodes N2 and N3 respectively receive an alarm. Then, this time, the alarms received by the nodes N2 and N3 are transmitted to, for example, the origin node in the node N1 and the management node N2, and the downstream nodes such as the nodes N4 and N5 and the destination node in the management node N3.

At the failure 314 of the span 306, the communication path of FIG. 45 becomes invalid. This happens even if there is only one failure, that is, there is only failure 314 in the entire path.
In other words, there is only one malfunctioning span (span 306). However, the data from the origin node is no longer sent to the destination node of the path shown in FIG. This happens even if the spans or links 302, 304, 308, 310, 312 are in motion.

As indicated above, one object of the present invention is to utilize the intact spans or links of a failed path in order to efficiently use the resources of the communication network.

Therefore, the network of the present invention is equipped with the function of the management node in failure in the distributed restoration algorithm to send a message informing the downstream node of the part of the path where the failure is still alive. .

The first step of the inventive scheme is shown in FIG. As is clear from the figure, a "reuse" message is sent from each node N2 and N3 to its neighbor node N1.
And sent to N4. The reuse message coming from N3 is shown as 318 and the reuse message coming from N2 is shown as 316. In particular, as shown in FIG. 49, a reuse message is shown that includes an identifier field 320, with the identifier represented by R added to designate the message as a reuse message. Furthermore, the message of FIG. 49 includes a variable length route information field to which an identification ID for each node can be added. Other fields of the message of FIG. 49, which are not particularly necessary for the description of the present invention, remain blank and are not included in the message shown in FIGS. 46 to 48.

Returning to FIG. 46, it can be seen that the reuse message 316 has the node ID of the node N2 in the route information field. On the other hand, the reuse message 31
8 has the node ID of the node N3 in the route information field.

Upon receiving the reuse message 316, the node N1 adds its node ID to the route information field before sending the reuse message to the originating node using the link 302. Similarly, when the reuse message 318 is received, the link 31
Before sending a reuse message 318 to node N5 using 0, node N4 adds its node ID to the route information field of reuse message 318 (FIG. 4).
See 7). Thus, the reuse message, whether 316 or 318, is sent from one node to another node downstream, and additional node IDs are added to the message until the message reaches the last node in the path, eg, the origin node in FIG. Will be added. At that point, the originating node reads the route information field of the reuse message 316 to know which part of the failed path is intact. As shown in FIG. 47, the origin node can easily know from the reuse message 316 that the nodes N2 and N1 have transferred the reuse message. Therefore, not only the span or link connecting the node N1 and itself, but also the span or link connecting these nodes to each other are functioning normally.
Therefore, as far as the origin node 0 is concerned, the fault that caused the path fault has occurred somewhere across node N2, and the element in front of node N2 is available and traffic interrupted by fault 314. Can only be used as an alternative route for rerouting.

Further, as shown in FIG. 47, the reuse message 318 is the management node N
It is also the message sent by 3. As is apparent from the figure, the reuse message 318 was transmitted by the node N4 to the node N5. As seen by node N5, reuse message 318 has its route information field set to node N3.
And node N4. Therefore, the node N5 knows that the path connecting itself and the node N3 is normal. Node N5 adds its node ID to the route information field of reuse message 318 before sending it to the destination node via link 312.

As shown in FIG. 48, the destination node now receives the reuse message 318. From the route information fold of the reuse message 318, the destination node has a fault 314 that occurs beyond node N3 and links 30 connecting N3, N4 and N5.
It can be seen that 8, 310 and 312 are intact. For the purposes of the present invention, the communication of information between intermediate nodes, such as between nodes N3, N4, N5 and the destination node, is done by in-band messaging between these nodes. Similarly, the transmission of the reuse message between the nodes N2 and N1 and the origin node is performed by in-band messaging.

Upon receipt of the reuse message 318, the destination node repacks the reuse message into another “reuse” message 320 and originates the repacked message using the wide area network (WAN) messaging connection 322. Transmit to node. Upon receipt of the reuse message 320, the originating node knows all the intact parts of the failed path shown in FIG. Therefore, for example, the backup link 324 connecting the nodes N1 to N5 can be used to create an alternative path. In the implementation example of FIG. 48, the alternative recovery paths are links 302 and 3 of the failed path for rerouting traffic from the source node to the destination node.
You can use twelve. Of course, other alternative routes connecting the source and destination nodes can also be used.

To facilitate recovery of the failed path after failure 314 has been recovered, the useful links of the failed path of FIG. 48 are most often restricted to routing information from the source node to the destination node. In other words, in addition to links 302 and 312, intact links 304, 308, 310 are reserved for use by the source and destination nodes for routing data.

Although FIG. 48 shows that another reuse message 320 is sent by the destination node to the origin node, the reuse message indicates that the link or span is working within the failed path to the destination node. From the source node to the destination node 3
It should be understood that it may be sent to 22 as well.

As shown in FIG. 50, the communication network embodiment of the present invention includes a number of nodes 6302 to 324 each connected to a neighbor node by at least one working link and one backup link. Is included. For example, the node 6302 is connected to the node 6304 by an active link 2-4W and a backup link 2-4S. Similarly, the node 6304 has an active link 4-6.
W and the spare link 4-6S are connected to the node 6306. For simplicity, nodes 6302-6304, 6304-6306, 6302 are shown in FIG.
Only the specific link connecting −6310 is numbered. However, it should be noted that the working and protection links connecting neighboring links can be specified as well.

In the communication network of FIG. 50, all nodes of the network are equipped with a distributed recovery algorithm (DRA), although in practice only one or more parts of the communication network are provided for distributed recovery. I assume that. That part of the network is then called the Dynamic Transport Network Recovery (DTNR) domain.

Further, FIG. 50 shows an operation support system (OSS) 6326. OSS 6326 is where network management monitors the operation of the entire network. In other words, it is at OSS 6326 that an overall view or map of the layout of each node in the network is provided. The OSS 6326 has a central processor 6328 and a memory 6330 in which data retrieved from various nodes is stored. Memory 6330 may include running memory and a database store. Although not shown in the figure,
OS for the interface unit to interface with various nodes
S 6326. For simplicity, as shown in FIG.
Only nodes 6302, 6304, 6306, 6308 connected to SS 6326 are shown in the figure. If there is an interconnection between the OSS 6326 and a node of the network, then what happens within each node of the network is monitored by the OSS 6326.

Each node 6302 to 6324 of the network is connected to Alcatel Network System Company.
Digital cross-connect switches, such as the 1633-SX Broadband Cross-Connect Switch manufactured by The Company. Two such adjacently connected switches are shown in FIG. The switches in FIG. 51 are, for example, nodes 6304 and 6
50 represents any two neighboring switches in the network of FIG. 50, such as 306. As shown, each switch has a number of access / output ports 6332, 6334, which are line terminator (LTE) 6 ports.
336 and 6338 are shown as being multiplexed. LTE 6336 and 6338 are SONET devices with detectors for detecting link failures between various digital cross-connect switches. Again, for the sake of simplicity, the detection circuit that interprets whether or not a communication failure has occurred is the node 6304.
Running cards 6340a, 6340b and nodes 6306 6342a and 6340
LTE as incorporated in 342b is not shown in the figure sandwiched between nodes 6334 and 6336.

As shown in FIG. 51, each digital cross-connect switch is connected to each operating interface card 6340a, 6340b, 66342a, 6342b.
Connect the nodes 6304 and 6306 so that they can communicate with each other.
It has two working links 6344a and 6344b. Also, the node 630
4 and node 6406 are a pair of backup links 6346a and 6346b connected to backup link interface cards 6348a, 6348b, 6350a, 6350b of nodes 6304 and 6306, respectively. In the case of the embodiment of FIG. 51,
Each working link 6344a and 6344b and spare link 6356a and 634
6b is assumed to be part of logical span 6352. In addition, node 630
Although only four links connecting 4 to 6306 are shown, in practice neighboring nodes are more or less connected by links. Similarly, although only four links are shown as part of span 6352, it should be noted that a span that actually connects two neighbors may have more links. I have to. In the present invention, operating links 6344a and 6344b are shown in FIG.
0 corresponds to the operating link 4-6W, and the spare links 6346a and 6346b in FIG. 51 correspond to the spare link 4-6S in FIG. Because of this aspect of the invention, each link shown in FIG. 51 is a conventional fiber optic transport system OC.
It is assumed to be -12 or a link embedded in a higher order (ie OC-48 or OC-192) fiber.

If focusing on node 6304, then 6340a, 6340b, 6348a,
Each interface card or card of a digital cross-connect switch, such as the 6348b, has multiple STSs for transmission to SONET LTE 6336.
Note that it is connected to the -1 port 6352. Although not shown, intelligence, such as a processor in each digital cross-connect switch, controls the routing and operation of various interface boards and cards. Although not shown in the figure, a database storage device that stores maps for identifying various sender nodes, selection nodes, and addresses is present in each digital cross-connect switch and will be discussed later. . Similarly, operating boards 6342a, 6342b and spare boards 6350a, 63
50b is connected to the access / output port 6354 of node 6306.
Further, the figure shows a non-DRA between neighboring nodes 6304 and 6306.

In the present invention, access / output ports such as 6332 and 6334 transmit their port numbers to neighboring nodes through the matrix of each digital cross-connect switch. Therefore, in the case of neighboring nodes 6304 and 6306 interconnected in the illustrated embodiment, ports 6352a and 6352b of node 6304 are connected to ports 6354a and 6354c of node 6 by an operating link 6344a.
It is connected to the. Similarly, ports 6352e and 6352f are coupled to ports 6354e and 63 of node 6306 by backup links 6346a and 6346b, respectively.
It is connected to 54f. Thus, if node 6304 transmits a signal to node 6306 using backup link 6346a, the node will send a message to port 6
352e to the spare card 6348a and then to the spare link 6346a. As a result, the message is received on a spare fiber-optic card OC- 12 or on a link spare card 6350a embedded in a higher (ie OC-48 or OC-192) fiber.

If focusing on node 6304, then 6340a, 6340b, 6348a,
Each interface card or board of a digital cross-connect switch, such as the 6348b, has multiple STs for transmission to the SONET LTE 6336.
Note that it is connected to the S-1 port 6352. Although not shown in the figure, intelligence, such as a processor in each digital cross-connect switch, controls the routing and operation of various interface boards or cards. Although not shown in the figure, a data storage device for storing maps for identifying various sender nodes, selection nodes, and addresses is actually present in each digital cross-connect switch. Featured in. Similarly, operating boards 6342a, 6342b and spare boards 6350a, 63
50b is connected to the access / output port 6354 of node 6306.
Furthermore, the figure also shows a non-DRA between neighboring nodes 6304 and 6306.

In the present invention, the access / output ports such as 6332 and 6334 transmit their port numbers to neighboring nodes through the matrix in each digital cross-connect switch. Therefore, in the neighboring nodes 6304 and 6306 connected to each other in the illustrated embodiment, the ports 6352a and 6352b of the node 6304 are connected.
Is running on a link 6344a, which causes ports 6354a and 6 of node 6306 to
It is connected to 354c. Similarly, ports 6352e and 6352f are coupled to ports 635 of node 6306 by backup links 6346a and 6346b, respectively.
4e and 6354f. Thus, when node 6304 uses spare link 6346a to send a signal to node 6306, it sends the message from port 6352e to spare card 6348a and then to spare link 6346a. As a result, the message is received at the spare card 6350a of node 6306 and routed to the receive port 6354e of node 6306. Therefore, as long as each working link or spare link connecting a pair of neighboring nodes, such as nodes 6304 and 6306, is operational, when a message is sent between these nodes, information about each send / receive port is returned: OSS 6326 (Fig. 50)
, And records of various ports connecting two neighboring nodes can be collected.

The inventor of the present invention has found that the topology or map of the available spare capacity of the network, in the form of the available spare links connecting the nodes, is different for different nodes to which the spare links are connected. We came up with the idea that it could be created from stored data representing port numbers. In other words, a message transmitted by a node to a neighboring node includes in OSS 6326, for example, the ID of the transmitting node, the IP (internal protocol) addresses of the transmitting and receiving ports of the node and the port number on which the message is transmitted from the node. OS if it can provide a large number of parameters
An S can get an overall picture of the reserve capacity of the network from similar messages exchanged between neighboring nodes of the protection link connecting the neighboring nodes.

Briefly, if each digital cross-connect switch in a DRA equipped network knows the port number and node connected by the backup link, that node has detected a failure on any working link. In that case, you know how to reroute traffic. Then, by gathering information about each node in the network, OSS 6326 can obtain a complete picture of all the spare links interconnecting the various nodes. As a result, when a working link fails, the OSS 6326 can send a network reserve capacity map to the management node of the failed link. It allows any management node, designated as the sender or origin node, to initiate the recovery process by finding an alternative route to reroute interrupted traffic using the network's reserve capacity map. can do.

The structure of the special message used to continuously monitor the available reserve capacity of the network is shown in FIG. In the case of the present invention, this message is called a live message. As can be seen from the figure, this survival message has many fields. Field 6356 has an 8-bit message field. In the message of FIG. 52, 8-bit data can be configured to represent a live message. As a result, each node that receives the message recognizes that it is a live message for updating the availability state of the standby link that received the message. On the other hand, when the OSS 6326 receives a live message, it combines it with all other live messages received from various nodes to map the reserve capacity of the network.

The next field in the message of FIG. 52 is field 6358. This is an 8-bit field containing the software modification number of the DRA used in the network. The next field 6360 is an 8-bit field containing the node identifier of the transmitting node. Field 6362 is a 16-bit field containing the port number of the transmission node to which the live message was sent.

The next field of the message is field 6364. 63 containing the IP address of the node's DS3 port, used for incoming half-duplex messages
It is a 2-bit field. Node D, used for outgoing half-duplex messages
The IP address of the S3 port is contained in the 632-bit field 66.

Field 6368 is a 1-bit field which, when set, informs the receiving node that the message came from the managing node for the failure. In other words, when a failure occurs, the management node of the failed link will
An alive message is sent from the management node to the downstream node because a failure has occurred, and an alive message is sent to inform that the recovery process is proceeding.

The last field of the live message is field 6370. A 7-bit field, reserved for future use.

In operation, before a failure is detected, live messages as shown in FIG. 52 are continuously exchanged on the protection links between neighboring nodes. This exchange of survivor messages causes the network to monitor the various available actual spare links, and each spare link to receive a survivor message as well as the neighbor node's port number on which the spare link is connected and the survivor message is received. , The port number of each node can be identified. By collecting the data contained in each live message, a record is kept of the inbound and outbound IP addresses of the various nodes, port numbers, and various backup links available on the network. Then, from this collected data, a topology of the available spare capacity of the network can be created by OSS 6326 or each node that can have the collected data downloaded for storage. In any case, a map of the available spare links in the network is available so that in the event of a failure, the failing management node will look for an up-to-date map of network spare capacity and reroute interrupted traffic accordingly. You can find the most efficient alternative route for.

Since the present invention is concerned with the distributed restoration process, since each digital cross-connect switch of the network knows the port number and the node connected by the backup link, the OSS stores the topology of the backup capacity of the network. Note that you do not have to. Thus, in the event of a failure, the originating node in charge of recovery can construct an overall topology of available spare links by retrieving different surviving messages from various nodes, so that each node can retrieve the surviving message. Keep sending. In other words, in an attempt to determine the available backup links, the originating node only needs to put all the alive messages together, because each node that has at least one standby link sends an alive message to the originating node. Then, the spare capacity of the network can be confirmed by searching the ID of the node and the port number of the node to which the spare link is connected. As a result, the backup link topology map can be used in a distributed manner in the origin nodes in the DRA compatible network of the present invention.

As mentioned previously, in the DRA network, the C bit is sometimes used on the link carrying the payload, but is used to exchange alive (KA) messages on the protection link. The link has two ports, one at each end of the link.

The KA message provides each port with information about the other ports. Nodes have one or more ports to which links between nodes are connected. In conventional arrangements, the node does not need to know, or need to know, which other nodes are connected to the link. For example, as shown in FIG. 61 (see also FIG. 51), node 10
0 has ports 10, 15, 20 and node 200 has ports 30, 35, 40. The node 100 is connected from the port 20 to the port 30 of the node 200 via a link. The nature of such links can be represented by the following shorthand notation (node 100, port 20) to (node 200, port 30). In the conventional configuration, node 100 and node 200 do not know that they are connected to each other.

In the present invention, as mentioned previously, the KA message provides each port and each node with information about other ports. KA messages are exchanged during normal operation. Other segments in the DS3 signal can also carry information, but the information contained in the KA message, especially carried in or embedded in the C bits, is used during the restoration of a failed link or span. To be done.

Generally, in communication, a so-called performance monitor is carried out, in which a port or a node watches an incoming signal, monitors the quality of the signal, and reports the quality of the signal to a higher authority. Corresponding ports at the other end of the link perform similar monitoring. The information in the KA message can be extended to include far end quality of service (QoS) information.

As mentioned above, the QoS information is a measure of the signal quality received at each port of the link. This information can be received in error (seconds), very erroneous (seconds)
, And signal loss (LoS), but is not limited thereto. QoS information can be reported over a period of time. The present invention defines this interval as either the last 15 minutes, the last hour, or the last day. The time when the KA message and the QoS information are sent is continuous.

The additional information is used to assign a quality value to the link, usually the backup link. The higher the QoS, the higher the quality of the link. Since both ports have the same information, both ports assign the same quality value to the link. The quality value is
Coupled with the transmission of data from one port to another in both directions, the quality value is usually higher in one direction than in the other.

For example, in FIG. 61, quality values are calculated from (node 100, port 20) to (node
Transmission of information to node 200, port 30) equals 3, but (node 200,
The transmission of information from port 30) to (node 100, port 20) shows a quality value equal to 5, here using a scale from 1 to 10, 1 being the best 1
0 represents the worst. Therefore, the link between these ports is assigned an average quality value of 4. Alternatively, when it is determined that the lower one of the two QoS values should be assigned to the link, (node 100, port 20
) To (node 200, port 30) are assigned an average quality value of 5.

In this additional information, the assignment of data to a particular link may be modified based on multiple criteria. For example, during a recovery event, the distributed recovery algorithm requires that the data with the highest priority be sent on the link with the highest QoS, and the data with the second highest priority receives the second QoS. You may decide that you have to be sent on the link you have.

For example, payload banking information and stock information may be given the highest priority, and system information called overhead may be given the lowest priority. Depending on the type of interruption that has occurred, the system information can be significant,
Sometimes it receives the highest priority. There are countless such combinations and we will not discuss them further here.

If returning to the time when the failure occurs, referring to FIG. 53 showing the prior art, the digital service 3 (DS3) path connecting the node 7301 to the node 7306 in the distributed restoration domain of the communication network. It is shown. For simplicity, other nodes or domains of the network are not shown.

As is well known, in DRA networks, when a link connecting two neighboring nodes fails, an alert is created and sent to each neighboring node. Such a failure or malfunctioning link is shown in FIG. 53 as a failure that occurred between nodes 7303 and 7304. This fault is due to node 7303
Caused by a loss of signal (LOS), loss of frame (LOF), loss of pointer (LOP), etc. For purposes of describing the present invention, it is assumed that such alarm signals are alarm indication signals (AIS).

In the prior art, when each node of the distributed restoration network or domain receives the AIS signal, each downstream node of the management node, such as node 7303 or node 7304, further propagates the signal to its own downstream node. Requesting Bell
It is set to comply with the standard defined in core document TR-NWT-00170. Thus, as shown in FIG. 53, upon receipt of the AIS signal, node 7303 conveys the AIS signal to node 7302, which in turn node 7302 to node 7301 and node 7301 further downstream in the DS3 path. Tell. The same flow of AIS signals received by node 7304
It also occurs at nodes 7304, 7405, 7406. In the embodiment of FIG. 53, the nodes 7301 and 7306 allow the distributed recovery domain to communicate with other parts of the communication network, or to other networks if the distributed recovery domain is not part of any network. It is assumed to be the connecting access / output port.

The problem with the prior art distributed recovery domain is that most, if not all, of the nodes in the domain receive the AIS signal so that the node is pinpointing which node is the true management node. It is very difficult, if not impossible, to decide whether a node is a node. Therefore, even if the network management easily recognizes that a failure has occurred in a specific path, it is not possible to isolate the exact location of the failure.

The present invention is applicable to networks that generally incorporate digital cross-connect switches, and especially to networks that incorporate broadband 1633-SX digital cross-connect switches. FIG. 54 illustrates an embodiment of the present invention in which faulty or malfunctioning links can be easily isolated. It should be noted that in FIG. 54, each node 7301 to 7306 is connected to an operation support system (OSS) 7310 that monitors the operating state of each node. Similar to the scenario of FIG. 53, the failure is assumed to have occurred between nodes 7303 and 7304. In the event of a link failure, as shown in FIG. 59, nodes 7303 and 7 are digital cross-connect switches, respectively.
304 detects LOS, LOF, or AIS defect signals, where each of these defect signals is defined by American National Bureau of Standards (ANSI) standard T1.231, respectively. For simplicity of explanation, assume that an AIS signal is detected. When a switch, such as node 7303 or node 7304, detects an AIS signal, it typically passes that AIS signal to the next downstream switch on the path, such as node 7302 or node 7305.

In the embodiment of the invention shown in FIG. 54, each switch or node converts or modifies the received AIS signal into a modified AIS signal before transmitting such a non-alarm signal to a downstream node. It has the appropriate hardware. In the embodiment of FIG. 54, such non-alarm signal is a DS3 idle signal. Therefore, with respect to nodes 7303 and 7304, note that when each of these nodes receives an AIS signal, these nodes convert the received AIS signal to an idle signal and pass the idle signal from the output port to a downstream node. There must be. In an embodiment of the invention, the DS3 idle signal contains a built-in message in the C-bit maintenance channel that identifies the presence of a fault within the distributed recovery domain.

Upon receiving the idle signal converted from the AIS signal, each downstream node transmits or propagates the idle signal to the downstream node. Therefore, the node 7302 outputs the idle signal received at the input port from the output port to the node 7302.
Hand it over to 301. Similarly, node 7305, upon receiving an idle signal from node 7304 at the input port, transmits the idle signal from the output port to node 7306. This process is repeated indefinitely until the idle signal reaches the access / output port connecting the distributed restoration network to the rest of the network.

Therefore, in the embodiment of FIG. 54, only the nodes 7303 and 7304 are A
It receives the IS signal, so it is monitoring the domain by OSS 7310,
It is easy to see that management of the distributed recovery domain has failed between node 7303 and node 7304 and that traffic must be routed out of the malfunctioning link connecting node 7303 and node 7304. be able to.

For a network outside the distributed recovery domain, the network cannot recover the interrupted traffic in a distributed manner, and in many cases cannot be controlled by the domain administrator, so the idle signal must be returned to the AIS signal. As a result, as far as equipment is located on the path of the external network, the alert has occurred somewhere within the distributed recovery domain and appropriate action must be taken. To that end, each access / output node of the domain has the ability to reconvert the idle signal received at the input port into an AIS signal so that it can be sent from the output port to a downstream node of the network outside the distributed recovery domain. ing.
Due to the reverse conversion of idle signals to AIS signals at the access / output nodes, customers and equipment outside the distributed recovery domain continue to receive the standard compliant AIS signals.

The process by which an alarm signal is converted to a non-alarm signal, ie an AIS signal is converted to an idle signal, is illustrated in FIG. FIG. 55 shows a DS3 frame structure according to the format published in ANSI standard T1.107-95, for example. In particular, the DS3 signal is divided into M frames of 4760 bits each. Each M frame is divided into 7 M subframes each having 680 bits. Each M subframe is divided into eight blocks of 85 bits each, and 84 bits of 85 can be used for the payload, and 1 bit is used for frame overhead. Therefore, there are 56 frame overhead bits in an M frame. These are divided into a number of channels such as: M
Frame allocation channel (M1, M2, M3), M subframe allocation channel (F1, F2, F3, F4), P-bit channel (P1, P2), X-bit channel (X1, X2), C-bit channel (C1, C2, C3).

The M frame allocation channel signal is used to accommodate all 7 M frames. The M subframe allocation channel signal is used to identify all frame overhead bit positions. In the P-bit channel, bits P1 and P2 are set to 11 or 00 and used for performance monitoring. The C-bit channel bit (C1, C2, C3) position is reserved for application-specific use. According to ANSI standard T1.107-95, the C bit can be used to indicate the presence or absence of stuff bits. Therefore, 3 of M subframes
If two C bits (C1, C2, C3) are set to 1, stuffing is done. If these C bits are set to 0, no stuffing is done. Further, it is determined by the one having the most three stuffing bits in each M subframe. For additional description of various bits, M subframes and M frames of DS3 signals, please refer to the above standard T1.107-95.

In order to convert the AIS signal into the idle signal of the embedded message, the inventor has used 3 C bits of M subframe 5 and set them to 1
We are aware of the fact that it is permitted by IT 1.107 to be used as a data link. Therefore, by changing at least one C bit of M subframe 5, a digital cross-connect switch, or node, can carry a message embedded in a standard idle signal otherwise. For example, a failure or malfunction that occurred on a link near a node
When an AIS signal is detected at a node, according to the invention, the node receives the AI
Block S to convert the AIS signal to an idle signal so that the AIS signal does not pass through the node, instead the node carries the OS3 idle signal as defined by ANSI T1.107-95. At the same time, the node starts the transmission of the embedded message by changing the three C bits of the M subframe 5 of the idle signal. Therefore, what is output from the node is an idle signal in which the C bit is changed.

For downstream nodes of the management node, such as node 7302 and node 7305, does the detected incoming idle signal have embedded messages, even if it contains all the usual attributes of a standard idle AIS signal? , At least 1
It has a change embedded message due to the change of the state of one C bit. As a result, these nodes are informed that the idle signal contains messages not found in the standard signal. Then, in the detection of this unusual idle signal, when this idle signal whose C bit is changed is transmitted to an access / output port such as the node 7301, the node 7301 propagates to a node outside the distributed recovery domain. Then, the idle signal is converted back to the AIS signal. This is the opposite behavior of an access / output node that received a normal AIS or idle signal. In the normal case, the same AIS or idle signal is propagated to downstream nodes outside the distributed recovery domain.

Therefore, according to the present invention, when an AIS signal is received, any node in the distributed recovery domain converts the AIS signal into an idle signal with the C bit changed, so that at most two nodes in the distributed recovery domain. Detect the AIS signal. The two nodes must be neighbors that are connected by a malfunctioning link that apparently alerted. If you take these things into consideration, OS
Since all nodes in the distributed recovery domain according to S 7310 are constantly monitored, failures in the distributed recovery domain are easily isolated.

When a failure occurs in a network outside the distributed recovery domain, the AIS signal created as a result of the failure enters the distributed recovery domain at either the access / output node. These nodes connecting the domain to the external network are configured so that they can convert incoming alarm signals into non-alarm signals and pass them on to other nodes in the distributed recovery domain. As before, for AIS signals,
The access / output node converts the AIS signal into an idle signal with at least one C-bit embedded embedded message. As a result, this converted idle signal is routed in the distributed recovery domain until it reaches another access / output node connecting the distributed recovery domain to an external network at the other end of the domain. At that time, the second access / output node that detects the idle signal with the C bit changed returns the signal to the AIS signal and transmits it to the downstream node outside the dispersion recovery domain. By such conversion and reconversion of the AIS signal in the distributed recovery domain, the domain management knows that the failure has occurred outside the domain, in the communication network.

FIG. 56 illustrates another aspect of the present invention. Here, the nodes in the distributed recovery domain are connected by optical fibers. The description of the embodiment of FIG. 55 is equally applicable here, but with a different type of signal: SONET STS.
-N type signals are transmitted between the nodes of the domain and in the communication network to which the domain is connected. For this type of STS-n signal, if a link failure occurs, the management cross-connect switch on either side of the malfunctioning link will
Detect one of the following conditions. Loss of Signal (LOS), Loss of Frame (LOF)
), Signal line warning instruction (AIS-L), pointer path loss (LOP-P), path AIS (AIS-P). See ANSI Standard T1.231 for these various defect signals.

Similar to the previous asynchronous scenario with reference to FIGS. 54 and 55, the STA-path AIS signal is propagated throughout the network when a SONET link failure occurs. The same conversion and re-conversion as before are performed as shown in the embodiment of FIG. However, instead of the format shown in FIG. 55, the ST shown in FIG.
An STS-n frame format such as S-3 frame is used. STS-
Instead of C bits in the n-type format, the inventor has discovered that the manipulable bit is the Z5 bit of the payload section of the STS-n format, eg the STS-3 format of FIG. DS3 shown in FIG. 55
As in the format, by changing one state of the Z5 bit, the management node can convert the detected AIS signal into an idle signal with the changed Z5 bit. When this altered idle signal reaches the access / output node of the distributed recovery domain, it is now converted back to the AIS signal and passed on to nodes outside the domain.

FIG. 58 is a logic diagram showing various processing layers performed in the digital cross connect switch. For example, an AIS signal is provided to an input port, specifically a layer 3 or level 3 port card, where basic lower level functions such as detecting the input signal and determining if the signal is an alarm signal are performed. To be executed. The signal is then used by the layer 2 shelf processor 73 to perform other processing functions.
14a. When an idle signal with a C-bit message, that is, an idle signal in which one of the C bits of any subframe is changed, enters the port card 12b, the contents of the C-bit message are removed at the port card level, and the message is Sent to Level 2 shelf processor 7314b for DR
Routed to A processor 7316. As a result, it becomes possible to determine to which port the message should be output. Also, the C-bit message is routed to the management processor 7318, which manages the DRA processor 73.
Together with 16, the appropriate database (not shown) is accessed to obtain information about the particular port or card of the cross-connect switch, which is the final destination of the signal. As a result, the signal is routed to the appropriate input port of the node's downstream cross-connect switch. In the logic diagram of FIG. 58, the reverse operation is
This is performed with respect to the input of the AIS signal.

A specific node of the present invention is shown in FIG. As can be seen, a number of transceiver units 7320 are connected to the digital cross connect switch 7322 by demultiplexers 7324 and multiplexers 7326. Each transceiver unit 7320 is connected to an alarm processor 7328, a restoration signal sensor 7330, and a restoration signal generator 7332. Alarm processor 7328, recovery signal sensor 7330, and recovery signal generator 7332 are connected to the processor at node 7334. Cross connect switch 73
The operation of 22 is, of course, controlled by node processor 7334.

In each transceiver unit 7320, a signal transmission / reception detector 7336 that communicates with the frame reception unit 7338 and the frame transmitter unit 7340.
There is. The frame receiving unit 7338 is connected to the converter 7342, and the frame transmitter unit is connected to the converter 7344. Since each transceiver unit 7320 has the same components and operates in the same way, only the operation of one transceiver unit 7320 will be discussed below.

Upon receiving the signal, the signal transmit / receive detector 7336 determines whether the signal is an alarm signal or another type of signal. If the input signal is indeed an alarm signal, then the signal transmit / receive detector 7336 first detects the interface 73
The alarm processor 7338 is notified via 36 and the signal is routed to the frame receiving unit 7338. The signal is then parsed and transferred to converter 7342 for conversion of the alarm signal into a non-alarm signal, and, if in a DS3 system, with the appropriate C bit changed. next,
The signal is transferred to the recovery signal sensor 7330 and further transmitted to the node processor 7334.

The non-alarm signal is then forwarded to the demultiplexer 7324 and then to the digital cross connect switch 7322. With appropriate determination from node processor 7334, the non-alarm signal is output to a downstream node so that the non-alarm signal is provided to multiplexer 7326 and to converter 7344 if further conversion is required. Appropriate port is selected. The non-alarm signal is then provided to the frame transmission unit 7340 and the signal transmit / receive detector 7336 and provided to the appropriate port for output to the downstream node.

The node shown in FIG. 59 is converted to an idle signal having a message in which the C bit is changed when the signal input is from outside the dispersion recovery domain and if the signal is surely an AIS signal, It should be noted that it is configured as an access / output node so that it can be propagated to downstream nodes in the distributed recovery domain until it reaches the access / output node at the other end of the domain. At that time, an idle signal having a message whose C bit has changed is converted back to an AIS signal and transmitted to a node outside the dispersion recovery domain. See US Pat. No. 5,495,471 for more information on nodes and their various units for transmitting and receiving signals in a SONET environment. Its content is included in the present application.

FIG. 60 is a diagram showing the status of various signals within and outside the distributed recovery domain of a DS3 system. In particular, it should be noted that outside the distributed recovery domain, the type of port used is non-DRA, and as far as the management of the distributed recovery domain is concerned, there is no need to pay attention to the input and output signals.
Looking first at the 7350 block row, at the DRA access port for access / output nodes where the input signal is the AIS signal, the output signal is an idle signal with a C-bit modified message. Then see line 7352a. If the signal to the DRA access input port is an idle signal, the idle signal is output from the node. See row 7352b. At the same access node, there is no idle signal with a C-bit message coming in from outside the distributed recovery domain. Therefore, there is no output signal from the access node shown in row q7352c of FIG.

Here, the access / access of the distributed recovery domain shown in the row 7354 of FIG.
Pay attention to the output side of the output node. When receiving the AIS signal from within the domain,
The AIS signal must be provided to the node outside the domain, as shown in row 7354a. That is, the nodes in the distributed recovery domain that are close to the access / output node receive the AIS signal and the failure must be specific to the link connecting the access / output node to a neighboring node in the domain. It should be noted that in row 7354b, like row 7354a, the idle signal is output from the access / output node to a node outside the domain if the idle signal is provided to the access / output from within the domain.
In line 7354c, if an idle signal with a C-bit changed message is received from within the recovery domain, this access / output node returns the previously converted alarm signal to the AIS signal and restores the AIS signal in a distributed manner. Communicate to downstream nodes outside the domain.

Row 7356 of FIG. 60 shows input / output signals of nodes other than the access / output nodes in the dispersion recovery domain. As shown in row 7356a, if the DRA enabled node receives the AIS signal, the processing described above is performed. That is, this AIS signal is converted into an idle signal having a message in which the C bit is changed. On the other hand, when a node receives an idle signal, nothing is done because the same idle signal is output from the node and passed down to downstream nodes. Row 7356
See b. Similarly, in row 3s56c, it should be noted that if a node receives an idle signal with a C-bit altered message, the same idle signal with a C-bit altered message is passed to the downstream node. Therefore, the nodes in the distributed recovery domain perform the conversion process only once when they receive an alarm signal such as an AIS signal. In other words, the AIS signal is converted into the idle signal only once when the connection link fails. Therefore, faults in the domain are easily identified by identifying the neighboring pairs of nodes in the distributed recovery domain that each detected the alarm signal.

So far, the connection of the nodes of the distributed recovery domain by the link has been taken up, but it should be understood that the present invention is similarly applicable to the distributed recovery domain in which the node is connected without the link. I have to. For example, for recovery domains that operate using microwave transmission, no physical link is used. Instead, each node is connected by microwave transmission, which has a well-defined format. Also, packet messages for microwave transmission include specific bits that can be altered using the same principles of the invention. As a result, if a given domain of a wireless network supports distributed recovery to quickly identify and isolate failures in the event of malfunction, alter the unused bits of microwave messages similarly. Allows the status of transmitted signals to be changed without affecting the operation of the network.

If the malfunctioning node is a node rather than a link, the malfunctioning node creates an alarm signal that the neighboring node receives, so that the neighboring node conveys the alarm signal to a node further downstream in the distributed recovery domain. The present invention is also applicable as long as it is converted to a non-alarm signal with a busy message. Therefore, it is possible to isolate the site where the failure has occurred whether it is a link failure or a node failure.

While the preferred embodiment of the invention has been disclosed for purposes of illustration, many changes, modifications, variations, replacements, equivalents, etc., in part or in whole, may be made by those skilled in the art to which the present invention pertains. Is clear. Accordingly, the invention is limited only by the scope and spirit of the claims appended hereto.

Where the invention is subject to many variations, modifications and changes in detail, it is understood that all that has been described herein and shown in the accompanying drawings is to be interpreted as illustrative rather than restrictive. ing. Accordingly, the invention is limited only by the spirit and scope of the appended claims.

[Brief description of drawings]

FIG. 1 briefly illustrates the concept of a communication recovery network and provides some definitions that apply to the present invention.

FIG. 2 shows a restoration sub-network for explaining the concept applied to the present invention.

[Figure 3]   FIG. 3 shows a failure in the recovery sub-network.

FIG. 4 shows two origin / destination node pairs to illustrate the scope of the invention.

FIG. 5A   FIG. 5A illustrates the loose synchronization feature of the present invention.

FIG. 5B   FIG. 5B illustrates the loose synchronization feature of the present invention.

[Figure 6]   FIG. 6 shows a fault notification message flow applied to the present invention.

[Figure 7]   FIG. 7 shows the flow of a live message according to the present invention.

[Figure 8]   FIG. 8 shows the flow of a path validation message according to the present invention.

FIG. 9 shows a time chart applied to the fault notification and fault isolation process of the present invention.

[Figure 10]   FIG. 10 illustrates the AIS signal flow within the restoration subnet of the present invention.

FIG. 11   FIG. 11 illustrates the AIS signal flow within the restoration subnet of the present invention.

FIG. 12 details a failure notification message in the recovery subnet according to the present invention.

[Fig. 13]   FIG. 13 illustrates the repeated initiation of the recovery process of the present invention.

FIG. 14 shows a time chart applied to the first iteration probe, return, maxflow, connect phase of the recovery process of the present invention.

FIG. 15   FIG. 15 shows a time chart of the investigation phase of the process of the present invention.

FIG. 16 shows a possible configuration of multiple origin / destination node pairs as seen by a given origin node.

FIG. 17 illustrates the two steps of the first iterative survey phase of the recovery process.

FIG. 18 shows a time chart applied to the return phase of the recovery process of the present invention.

FIG. 19   FIG. 19 shows the steps associated with the return phase of the recovery process.

FIG. 20   FIG. 20 illustrates link allocation according to the return phase of the present invention.

FIG. 21   FIG. 21 illustrates link allocation according to the return phase of the present invention.

FIG. 22   FIG. 22 illustrates link allocation according to the return phase of the present invention.

FIG. 23 shows an exemplary return message received by the originating node of the restoration sub-network.

FIG. 24 is a time chart for explaining a correction map obtained from a return message received by an origin node.

FIG. 25 shows that a part of the restoration sub-network mode in the origin node is the origin node /
It has been assigned to the destination node pair.

FIG. 26 shows the Max Flow output for the Max Flow phase of the recovery process.

FIG. 27 shows the optimal routing applied to the Maxflow output of the present invention.

FIG. 28 provides a time chart to show the sequence of the connection phase of the first iteration of the process of the present invention.

FIG. 29 shows a connection message for providing an alternative path route between a source node and a destination node of the restoration sub-network.

FIG. 30 illustrates how the present invention handles hybrid restoration sub-networks.

FIG. 31 illustrates how the present invention handles hybrid restoration sub-networks.

FIG. 32 shows the investigation phase and the return phase, respectively, applied to the hybrid network.

FIG. 33 shows an investigation phase and a return phase applied to a hybrid network, respectively.

FIG. 34 shows a time chart including additional iterations for processing a hybrid network in accordance with the subject matter of the present invention.

FIG. 35   FIG. 35 illustrates a low quality spare in accordance with the teachings of the present invention.

FIG. 36   FIG. 36 illustrates a low quality spare in accordance with the teachings of the present invention.

FIG. 37   FIG. 37 illustrates the usage of the management node message of the present invention.

FIG. 38   FIG. 38 illustrates the restricted reuse feature of the present invention.

FIG. 39   FIG. 39 illustrates the restricted reuse feature of the present invention.

FIG. 40   FIG. 40 illustrates the restricted reuse feature of the present invention.

FIG. 41   FIG. 41 illustrates the restricted reuse feature of the present invention.

FIG. 42   FIG. 42 illustrates the restricted reuse feature of the present invention.

FIG. 43   FIG. 43 illustrates the path prohibition function of the present invention.

FIG. 44   FIG. 44 further illustrates the path prohibition function of the present invention.

FIG. 45   FIG. 45 is a communication network path for explaining the present invention.

FIG. 46 illustrates a message sent from the management node of the failed path of FIG.

FIG. 47 provides another view of the failure path of FIG. 45 in which messages are sent from intermediate nodes to their respective downstream nodes.

FIG. 48 is the faulty path of FIG. 45 showing the messages arriving at the destination node, the connection between the origin node and the destination node, and the use of the backup link to bypass the fault. There is.

FIG. 49   FIG. 49 illustrates the reuse message of the present invention.

FIG. 50   FIG. 50 illustrates the communication network of the present invention.

FIG. 51 is a block diagram showing two proximity cross-connect switches and the physical connections between them.

FIG. 52   FIG. 52 illustrates the structure of an example live message of the present invention.

FIG. 53 shows multiple nodes in a distributed recovery domain through which an alarm signal created as a result of malfunction on a link connecting two nodes is propagated to downstream nodes. There is.

FIG. 54 shows the same node as shown in the DS3 environment of FIG. 53, but in this example the nodes in the distributed recovery domain are configured to convert the input alarm signal to a non-alarm signal. Configured, the access / output node of the distributed recovery domain is further configured to reconvert the received and modified alarm signal to an alarm signal.

FIG. 55 shows a DS3 signal frame structure to illustrate the conversion of an alarm signal to a non-alarm signal in DS3 format.

FIG. 56 is similar to the embodiment of FIG. 54, but the embodiment of FIG. 56 describes a SONET network.

FIG. 57 illustrates the format of the STS-3 frame to illustrate the conversion of alarm signals to non-alarm signals in SONET networks.

FIG. 58 is a simplified block diagram showing the flow of signals to nodes in the distributed recovery domain.

FIG. 59 is a block of nodes of a distributed recovery domain of the present invention configured to convert alarm signals to non-alarm signals and reconvert non-alarm signals to alarm signals at access / output nodes. It is a figure.

FIG. 60 is a diagram showing the states of various signals going in and out of a dispersion recovery domain.

FIG. 61 is a block diagram illustrating a communication network having a plurality of interconnection nodes connected via links and ports of the present invention.

─────────────────────────────────────────────────── ─── Continued front page    (81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, I T, LU, MC, NL, PT, SE, TR), OA (BF , BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, G M, KE, LS, MW, MZ, SD, SL, SZ, TZ , UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AG, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, B Z, CA, CH, CN, CR, CU, CZ, DE, DK , DM, DZ, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS, J P, KE, KG, KP, KR, KZ, LC, LK, LR , LS, LT, LU, LV, MA, MD, MG, MK, MN, MW, MX, MZ, NO, NZ, PL, PT, R O, RU, SD, SE, SG, SI, SK, SL, TJ , TM, TR, TT, TZ, UA, UG, US, UZ, VN, YU, ZA, ZW

Claims (28)

[Claims] 1. A method for identifying a neighboring node pair in a communication network having at least one distributed restoration sub-network, which constructs a first C-bit alive message at a first node of the neighboring node pair connected by a link. Embedding the first C-bit survival message in the C-bit of the first DS3 signal, determining the quality of service information of the link by seeing the neighbor node from the first node, embedding the quality-of-service information in the C-bit survival message Transmitting the first DS3 signal from the first node of the pair of neighboring nodes on the link to the second node, the first C-bit alive message looking at the second node from the first node, A method of identifying nodes and identifying quality of service information for links. 2. Receiving a first DS3 signal at a second node, processing the first DS3 signal to identify the first node from the first C-bit alive message to the second node, and viewing the first node to the second node. The method of claim 1, further comprising: identifying quality of service information for the link from the first C-bit alive message. 3. Constructing a second alive message in the second node of the pair of neighbor nodes, embedding a second C-bit alive message within the C bits of the second DS3 signal, looking at the first node from the second node and linking Determining the quality of service information of the C-bit survival message, embedding the quality of service information in the C-bit survival message, transmitting the second DS3 signal from the second node to the first node of the neighboring node pair on the link, and transmitting the second node to the first node And identifying the second node with respect to the first node and identifying quality of service information for the link from the C-bit alive message. 4. The first node has a first identification, the second node has a second node identification and quality of service information for the link, and the first C-bit alive message includes the first identification and the second C The method of claim 3, wherein the bit live message includes a second identification designation and quality of service information for the link. 5. The first identification designation includes a first node identifier, a first node port number, a first node wide area network address in the first C-bit alive message,
A first node "administrator" indicator, the second identification designation is a second node identifier,
The method of claim 4, wherein the second node port number, the second node wide area network address, and the second node "administrator" indicator are included in the second C-bit alive message. 6. The method of claim 1, wherein the network includes a plurality of neighbor node pairs and a C-bit embedded live message is sent between each neighbor node pair connected by a link. 7. Each node has an identification designation, and each C-bit embedded survival message identifies the source node of the C-bit embedded survival message and the quality of service of the link when one node sees the next node. The method of claim 1 including information. 8. The method of claim 7, wherein the link is a backup link. 9. The method of claim 3, wherein the link is a backup link. 10. The method of claim 2, wherein the link is a backup link. 11. The method of claim 1, wherein the link is a backup link. 12. The method of claim 7, wherein each identifying designation includes a node identifier, a port number, a wide area network address, an "administrator" indicator of the node that originated the C-bit embedded live message. 13. The method of claim 12, wherein each DS3 signal passes in-band. 14. The method of claim 12, wherein the link quality of service information comprises at least one erroneous time (seconds), a very erroneous time (seconds), and a signal loss time (seconds). 15. The method of claim 8, wherein each identifying designation further comprises a node identifier, a port number, a wide area network address, an "administrator" indicator of the node that originated the C-bit embedded live message. . 16. The method of claim 15, wherein each DS3 signal passes in-band. 17. The method of claim 16, wherein the link quality of service information comprises at least one erroneous time (seconds), a very erroneous time (seconds), and a signal loss time (seconds). 18. The method of claim 17, wherein erroneous time (sec), very erroneous time (sec), signal loss time (sec), quality of service are defined by the DS3 standard. 19. A plurality of nodes interconnected by a plurality of links,
A communication network including a distributed restoration subnetwork, the first node having a first unique identifier, the second node having a second unique identifier, a link connecting the first node to the second node, and a link. A DS with the quality of service information of the link embedded in the live message and C bits, including the DS3 signaling channel in
A communication network that can send three signals to each other. 20. The network of claim 19, wherein the DS3 signal from the first node comprises a first unique identifier and the DS3 signal from the second node comprises a second unique identifier. 21. The first node can send a first node DS3 signal to the second node on the link to identify the first node to the second node, and the second node can send the second node DS3 signal to the second node on the link. 21. The network of claim 20, capable of sending to one node to identify the second node to the first node. 22. The network of claim 19 further comprising a plurality of neighboring node pairs, each pair connected by a protection link capable of sending a plurality of DS3 signals with a C-bit embedded live message to each other. network. 23. The network of claim 22, wherein each node has a unique numeric identifier, and each live message includes the unique numeric identifier of the node from which the DS3 signal was created. 24. The network of claim 23, wherein each unique numeric identifier includes at least one node identifier, a port number, a wide area network address, an "administrator" indicator of a node that is a source of DS3 signals. 25. The network of claim 19, wherein each node has a unique numeric identifier, and each live message includes a numeric identifier of the node from which the live message was created. 26. The network of claim 19, wherein the alive message includes at least one node identifier, port number, wide area network address, "administrator" indicator. 27. The network of claim 19, wherein all DS3 signals pass in-band. 28. The network according to claim 27, wherein erroneous time (sec), very erroneous time (sec), signal loss time (sec), quality of service are defined by the DS3 standard.