JP3447232B2 - Redundant line control method and 4-wire bidirectional multiplex transmission device using the same - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a 4-wire bidirectional loopback switch ring (BLSR).
itch ring) relates to a method of redundancy control in a multiplex transmission device. In particular, the present invention relates to a control method for performing line switching by redundant control capable of quickly and stably avoiding signal interruption in repairing a plurality of line failures.

[0002]

2. Description of the Related Art A BLSR multiplex transmission device is used as a transmission device for transmitting data such as voice and images in a network. When a network is configured with BLSR multiplexed transmission equipment, the line capacity per optical fiber is theoretically 1 on average as compared with the capacity required corresponding to the transmission equipment used in the network up to now. It has a capacity of 5 times.

In particular, the 4-line BLSR multiplex transmission device has 2
Compared with line BLSR multiplex transmission equipment, in addition to ring switch (Ring-switch), span switch (Span-switch)
It has the function of. Therefore, since there is a high possibility that the line can be rescued, the demand for the 4-wire BLSR multiplex transmission device is increasing.

Here, a ring-switch
Is a function of relieving a line failure through a long path by looping back the line of the ring network and looping back.

On the other hand, the span-switch is a line failure caused by switching a line that has a line failure between spans, that is, between the stations provided in the opposite BLSR multiplex transmission apparatus, to a protection line. It is a function of relief.

Here, in each of the 4-wire BLSR multiplex transmission devices, line failure is detected based on the state of the signal input to the device itself. Further, the level of the detected fault is determined and the relief function of the corresponding line is executed.

Therefore, in the 4-wire BLSR multiplex transmission apparatus between spans, there may be cases where the detected failure levels are different. That is, the line redundancy control to be executed for the failure relief may be different depending on the difference in the detected failure level between the own apparatus and the opposite apparatus.
In such a case, if a fault occurs in another span, the fault may not be relieved.

[0008]

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to determine the redundant control to be executed based on the fault detected by the self-device and the fault detected by the opposite device, thereby performing the conventional BLSR multiplexing. An object of the present invention is to provide a line control method that increases the probability of repairing a line as compared with the case of using a transmission device.

Another object of the present invention is to provide a line control method which ensures a high stability of a redundant function by monitoring an unexpected abnormal control state and defining a protocol for exiting from the abnormal state. is there.

[0010]

In a network configured by connecting a plurality of nodes each provided with a four-wire bidirectional multiplexing transmission device by a working line and a protection line, one of the nodes adjacent in one span Node detects the fault level of its own node on the working line and the protection line,
And the level of the failure of the own node and the level of the failure detected by the other node of the adjacent node are received by receiving the level of the failure of the working line and the protection line detected by the other node of the adjacent node. The priority is added to determine the priority of line switching.

By doing so, it is possible to collect the span failures and perform appropriate switching control according to the combination of the failures.

As a concrete mode, the addition of the fault level of the local node and the fault level detected by the other node of the adjacent node is effective when the line switching priority corresponding to the fault level is high. And

A working line detected by the other node of the adjacent nodes received by the one node, and
The level of failure of the protection line is sent in K bytes of the SONET format overhead.

In another aspect, the remote defect indication (RDI: Remote Defect Indiction) of the working line and the protection line carried in the K-byte is carried out.
cation) to the local node's signal fail (SF: Signal
Failure) Trigger the circuit switching of the equivalent failure.

Ring segmentation
As a countermeasure against the above, the present invention causes a failure in both the working line and the protection line in a span of 2 or more, and the failure in at least one of the working line and the protection line exceeds a predetermined error rate. When the ring connection is a failure and the failure of the working line is more than the error rate lower than the predetermined error rate, the ring connection is released and the line switching is important. Exception processing is performed with the priority of the switching request having a higher frequency.

As a specific mode, the predetermined error rate is 10 ⁻³ to 10 ⁻⁵ (SF: Signal Failure), and an error rate lower than the predetermined error rate is
It is characterized by being 10 ⁻⁶ to 10 ⁻⁹ (SD: Signal Degrade).

In still another mode, a failure occurs in both the working line and the protection line in two or more spans, and at least one of the working line and the protection line has an error rate of 10 ^-3. -10 ^-9 (SF: Signal Failure) or more, and the ring connection is broken.
failure), the failure of the protection line causes an error rate of 1
When it is 0 ^-6 to 10 ^-9 (SD: Signal Degrade), span switching is performed.

As a switching with a high probability of signal saving, a ring topology (Ring topology) indicating the number of nodes existing on the ring and the positional relationship of each node on the ring when the ring connection falls into the condition of being disconnected. Topology) and the far-end node where signals on the ring topology can reach, the position of the span switched by the far-end node is counted from the neighbor of the span switched by its own node (total number of nodes /
If it is farther than 4), the ring switch release determination is valid. If it is closer than (total number of nodes / 4), the ring switch is always performed. This enables more line relief.

4-wire BL corresponding to the control method of the present invention.
The basic configuration of the SR multiplex transmission apparatus is a four-wire bidirectional multiplex transmission apparatus arranged in each of the plurality of nodes in a network configured by connecting with a protection line,
A line fault collecting unit that detects the fault level of the working line and the protection line of the own node, a K byte collecting unit that collects K bytes of SONET overhead bytes, and a self node that is collected by the line fault collecting unit Add the failure and the failure of the opposite node collected by the K byte collection unit,
A switching determination unit that determines the switching priority is provided, and line switching is performed according to the switching priority determined by the switching determination unit. Further features of the invention will be apparent from the embodiments of the invention described with reference to the drawings.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. In the drawings, the same or similar parts will be described with the same reference symbols.

FIG. 1 is a schematic configuration diagram showing a configuration example of a network using a BLSR multiplex transmission device which is a transmission device, which is a target of the present invention. In FIG. 1, BLSR
This is an example of a network in which multiplex transmission devices are arranged in stations A to D (hereinafter, referred to as nodes A to D) and four optical fiber transmission lines are connected in a ring shape.

Now, let us consider the definition of a line between two transmission devices arranged in a network. As shown in FIG. 2, between two transmission devices placed in the node A and the node B (this is called a span), a working line WK (Work) and a protection line PT (Pro) are provided for bidirectional communication.
connection).

Considering the node A as a reference, a case where there is a failure in the working line and the protection line to the transmission device placed in the node A is called a self-node failure. Conversely, Node B
When there is a failure in the working line and the protection line directed to the transmission device placed in the node, and this is notified from the node B to the node A, it is called an opposite node failure.

Returning to FIG. 1, for example, a line pair LP1 of a working line WK and a protection line PT which is input to the waist side of the node A and in which a signal flows in the waist direction.
And a line pair LP2 of a working line WK and a protection line PT which are input to the east side of the node B and through which a signal flows in the east direction.

In the following description, for example, the protection line in the West direction of the node B indicates the protection line PT of the line pair L1 shown in FIG.

BLSRs arranged in nodes A to D
Since the configuration of the multiplex transmission device is common, FIG. 1 shows the configuration of the BLSR multiplex transmission device 100 arranged in the node C below as a representative.

The BLSR multiplex transmission apparatus 100 has a BLSR redundancy control unit 1 as a central function in relation to the present invention. The BLSR redundancy control unit 1 controls the signal switching unit 2 based on the fault information to control the line switching.

The BLSR multiplex transmission apparatus 100 further includes multiplexing circuits 20 and 21 for sending out a multiplexed signal in the West direction and the East direction, and the East (West) direction and the West (West) direction. West) direction multiplexed signals are received, and demultiplexing circuits 30 and 31 for demultiplexing the signals are included.

The demultiplexing circuits 30 and 31 have a function of demultiplexing a multiplexed signal and detecting a line fault from a received signal. Then, the detected line fault information is transferred to another B
It has a function of sending to the LSR multiplexing transmission apparatus 100.

FIG. 3 shows the SONET (Sync
hronous Optical NETwork) is the frame format of the signal. As shown in FIG. 3A, the frame format has a header part I and a payload part II.

Further, LOH (Line Over Hea
The section d) has K1 and K2 byte sections as shown in FIG. 3B, and the line failure information transmitted / received by the BLSR multiplex transmission apparatus 100 is added to the section according to the present invention.

FIG. 4 shows the contents of K1 and K2 bytes.
The structure of K1 and K2 bytes is shown in FIG. K1
As shown in FIG. 4B, bits 1 to 4 of the byte indicate the priority of the line switching request corresponding to the importance of the line failure. Furthermore, K
Bits 5 to 8 of 1 byte are the destination ID of K1 byte
Can be placed.

On the other hand, in bits 1 to 4 of the K2 byte, as shown in FIG. 4C, the ID of the own node which is sending the priority request corresponding to the line failure is put. Bit 5 of the K2 byte indicates whether the bridge request of the K1 byte is a short path request (0) or a long path request (1). Here, the short path is a case of sending a line switching request between opposite nodes, and the long path is a case of sending a line switching request via another node.

The BLSR multiplex transmission apparatus 10 shown in FIG.
An embodiment of the present invention which is premised on the configuration of 0 will be described below.

FIG. 5 is a diagram showing a first embodiment network of the present invention. Similar to the configuration shown in FIG. 1, BLSR
Nodes A to D having the multiplex transmission device 100 are
They are connected by a transmission line to form a ring-shaped network.

FIG. 5 (A) is a diagram for explaining the problems assumed in the first embodiment. Now, assume that in the span between the node A and the node B, the fault 1 occurs in the working line WK in the east direction and the fault 2 occurs in the protection line PT in the transmission line in the west direction.

In the redundancy control in the conventional BLSR multiplex transmission apparatus, when a failure is detected only in the working line WK, the span switch for passing the protection line PT of the span in which the failure has occurred is executed to rescue the line.

Therefore, in the example of FIG. 5A, when the BLSR multiplex transmission apparatus of the node B detects the failure 1 of its own working line WK, the span passing through the backup line PT of the span in which the failure has occurred. Run the switch.

Then, the information on the fault 1 is notified from the BLSR multiplex transmission apparatus of the node B to the BLSR multiplex transmission apparatus of the node A through the protection line PT in the waist direction. However, since the fault 2 has occurred in the protection line PT, the BLSR multiplex transmission device of the node A cannot receive the fault information from the BLSR multiplex transmission device of the node B.

On the other hand, when the BLSR multiplex transmission apparatus of the normal node A receives a failure notification from the BLSR multiplex transmission apparatus of the node B, a request for confirming this and performing span switching is issued to the BLSR multiplex of the node B. The transmission device will be notified.

However, the BLSR multiplex transmission apparatus of the node A cannot receive the failure information from the BLSR multiplex transmission apparatus of the node B because the failure 2 occurs in the protection line PT.

After all, the BLSR multiplex transmission apparatus of the node B cannot receive the span switching request from the BLSR multiplex transmission apparatus of the node A for a predetermined time (for example, 50 ms), and it times out. As a result, Span-switch does not succeed.

In such a case, in the conventional control method, ring-switching is performed according to the fault level of the own node. In the example of FIG. 5A, in the BLSR multiplexing transmission device of the node B, the fault 1 is at the SD (Signal-degrade) level, which is the error rate of 10 ⁻⁶ to 10 ⁻⁹ , and the BLSR multiplexing of the node A is performed. The transmission device detects that the fault 2 is at the SF (Signal-failure) level, which is an error rate of 10 ⁻³ to 10 ⁻⁵ . Then, as ring switching corresponding to these levels, SD ring (SD-Ring), SF
A ring (SF-Ring) is performed by each BLSR multiplexing transmission device.

As described above, in FIG. 5A, the next operation is determined only by the failure detected in the own node. In such a case, it may not be possible to rescue the line. Depending on the fault (SF or SD) detected at the local node, SF-
Even if Ring or SD-Ring is executed, if different faults occur in the opposite node, there is no means for confirming the state of the opposite node.

Therefore, the activated ring switch may not be appropriate. Moreover, I cannot confirm that it is not appropriate. Further, even when the switching system of the opposite node cannot operate, this cannot be determined and the operation of trying the ring switch is performed.

An example of the line control method of the present invention which solves such inconvenience is shown in FIG. 5 (B). Furthermore, FIG.
1 shows a configuration example of the BLSR redundancy control unit 1 according to the present invention in FIG.

The operation of FIG. 6 will be described with reference to the description of the embodiment of FIG. The K byte collection unit 11 of the BLSR redundancy control unit 1 of the BLSR multiplex transmission device of the node B receives the K byte (byte) sent from the opposite node (assumed to be the node A), and sends it to its own node and other nodes. Classify

K1 byte 1 to 1
The switch determination unit 14 and the opposite station failure collection unit 12 are notified of the corresponding priority (Priority) of 4. The opposite station failure collection unit 12 separates the priority transmitted by the K bytes into failures of the opposite node according to Table 1, and notifies the failure addition unit 13 of the failure status.

[0049]

[Table 1]

For example, 1 to 4 bits of the K1 byte of the opposite node notified to the opposite node failure collection unit 12 are "1".
111 ", and therefore the priority LP-S (SF
-P), from Table 1, the working line of the opposite node, node A, is normal (NoALM), and the protection line is S.
It can be seen that it is an F level disorder.

The fault collecting unit 10 collects the faults detected by its own node, that is, the node B, and notifies the fault adding unit 13 when there is a fault. For example, in the case where the own node failure detected by the node B is an SD level failure in the working line and the protection line is normal, according to Table 1, the priority (SD-S) is notified to the failure adding unit 13. .

The fault adder 13 adds the fault of the node A and the fault of the node B, and notifies the switching determination unit 14 of the addition result. The switching determination unit 14 knows that the working line WK side of the node B is SD based on Table 1 above. It is added to the SF of the own node protection line PT side that has already been notified from the failure collection unit 10.

Therefore, the fault condition of working line = SD and protection line = SF is recognized, the above table 1 is inversely converted, and the priority is determined to be SF-Ring (SF-R). According to the newly determined switching priority, S
F-Protection (SF-P) is released and SF-R is started. Also in the node B, the SD-Span (SD-S) is released and the SF-R is started by the same procedure as the node A, and the ring switch (Ring-switc) is started as shown in FIG.
h)

The switching execution unit 15 includes the switching determination unit 1
According to the priority determined in 4, the span switching (Span-switch) S or the ring switching (Ring-switch) R is controlled, and K bytes are transmitted K to notify the opposite node A of the priority of the own node. .

In this way, it becomes possible to collect the faults in the own node and the faults in the opposite node, and perform appropriate switching control according to the combination of faults. At the same time, it is possible to match the control of switching of the opposite BLSR multiplexing transmission devices connected to the span.

As described above, according to the redundancy control method in the conventional 4-wire BLSR multiplex transmission apparatus, switching is performed by using a failure detected at the own node or an external command for the own node as a switching control trigger. It has started and is not aware of the failure detected at the opposite node. A node that has not detected a failure or received an external command
It always depends on the node that initiated the switch.

On the other hand, as described with reference to FIG. 5, according to the present invention, switching is performed based on the fault detected by the own node and the fault condition detected by the opposite node which can be judged from the K bytes received from the opposite node. When the operation is decided, it is possible to rescue the line in one span.

Here, the meaning of addition of the failure of the node A and the failure of the node B will be examined.

The fault adding section 13 individually performs addition processing on the working line and the protection line. As shown in FIG. 2, each line physically has a line for transmitting from the own node to the opposite node and a line for receiving from the opposite node to the own node. The fault adder 13 performs addition on each of the working line and the protection line.

The fault of the local station (local node) input to the fault adder 13 of the BLSR redundancy controller 1 shown in FIG. 6 is a fault of the receiving line of the local node detected by the fault collector 10.

On the other hand, the opposite station (opposite node) failure is a failure of the transmission line and is a failure detected by the opposite node which the opposite station failure collection unit 12 obtains from the received K bytes.

The importance of the failure is in the order of No ALM <SD <SF, and the addition result in the failure adding section 13 is as shown in Table 2. That is, a failure of high importance among the failures of the own node or the opposite node is obtained as the addition result as shown in Table 2.

[0063]

[Table 2]

For example, when the fault of the own node is SD and the fault of the opposite node is SF, the addition result is set to SF. This is the case in FIG. 5 (B).

The switching determination unit 14 selects the highest priority from the priority of the failure resulting from the addition, the priority of the own node and the priority of the other station, and determines the priority of the switching executed by the own node.

In the case of FIG. 5B, the priority for the own node is SD and the priority for the opposite node is SF. Therefore, according to Table 2 above, the priority SF
Is the priority for switching the local node.

Here, as a concrete method of addition, there are a method of calculating a logical sum and a method of performing a judgment process by a conditional branch.

In the former method, as shown in FIG. 7, three 2-bit registers aa, bb, and cc for respectively holding the own node fault, the opposite node fault, and the addition result are used. The first bit is assigned to SD and the second bit is assigned to SF. 1 is set in each failure state.

The logical sum of the own node fault register aa and the opposite node fault register bb without carry is taken and stored in the addition result register cc. If the value of the addition result is "00", it is No ALM, and if it is "01", it is SD and "1".
If it is 1 "or" 10 ", it is considered as SF. Therefore, the result according to Table 2 can be obtained.

When the judgment processing is performed by the latter conditional branch, the processing of the flow of FIG. 8 is performed. That is, in step S01, if the own station (node) failure is SF, SD, or No A
Processing is distributed by LM.

If the own station (node) failure is SF, the addition result is unconditionally SF (step S02). If the own node failure is SD, the addition result is SF when the opposite node is SF (step S03). Further, when the own node failure is No ALM, the addition result is determined by the importance of the failure of the opposite node (step S04). The same result as in Table 2 can be obtained by the branching process of FIG.

However, even in the method of the first embodiment of the present invention, it may not be possible to rescue the line. That is, this is the case where a failure has occurred in another span.

FIG. 9 is a diagram for explaining a case where such a line cannot be repaired. In the network, faults 1, 2 and 3 occur in two spans.

The node B detects the fault 1 of the working line WK at its own node, and from the K byte (SF-P) sent from the node A, the working line WK and the protection line in the span between the node A and the node B. Both PTs have a failure and priority S
It can be determined that the switching corresponding to F-P should be performed.

However, the K bytes sent by node B are
The node A cannot be reached due to a failure between the node B and the node A and between the node C and the node D. To this end, node A
Determines its operation only by the fault 2 of the protection line PT detected at its own node, and the priority remains SF-P.

With respect to the failure between the node A and the node B, the nodes A and B should be paired and have the same switching state. However, since they do not have sufficient information transmission means, the switching states do not match. Get up.

Therefore, according to the present invention shown in FIG.
In the embodiment of the above, in the BLSR redundancy control unit 1 of FIG.
The K byte collection unit 11 separates bits 6-8 of the K2 byte from the K bytes received from the working line WK and the protection line PT.

Bit 6 of the separated K2 byte
-8 is notified to the opposite station failure collection unit 12 together with the K1 byte. In the opposite station failure collection unit 12, the K byte collection unit 11
Bits 6-8 of the K2 byte notified by the
If it is a code indicating the RDI shown in FIG. This determination is based on the working line WK and the protection line P.
Do T individually.

Here, RDI (Remote Defect Indicati
on) is a game alarm, and SF or more (LOF,
LOS unit missing, unit failure included) when a failure occurs, code of bits 6-8 of K2 byte
It is a game alarm notified by "110".

[0080]

[Table 3]

That is, the fault of the opposite node is added according to Table 3 by the priority of bits 1 to 4 of the K1 byte obtained in the embodiment of FIG. 5B. In the example of FIG. 10, since the addition result is SF, this is sent to the switching determination unit 14.

By doing so, it is possible to determine all the faults of the opposite node that could not be determined only by the priority of bits 1 to 4 of the K1 byte, and it becomes possible for the own node B and the opposite node A to completely match. .

As a concrete example, in FIG. 9, node A-
In the span between the nodes B, the protection line PT for the node B (the direction of the node A, that is, the west direction ) , the working line WK for the node A (the direction of the node B, that is the east direction ) , and the SF for all the lines between the nodes C and D Suppose a failure occurs.

In this case, as shown in FIG. 9, in accordance with Table 1, node A starts SF-P switching control and node B
Starts switching control of SF-S. The node C and the node D start SF-R. Next, in the node B, as described in the first embodiment, according to Table 1, the SF of the working line WK detected in the own node and the K1 byte from the node A
(SF-P) is added to the SF of the protection line, and SF-
Start control of R.

At the same time, in the node A, the opposite node failure collection unit 12 detects the RDI code in the K2 byte of the K bytes on the working line WK from the node B. Since the K1 byte from node B is communicated through the protection line, it cannot reach node A in this case. Therefore,
The failure status of the opposite node B cannot be judged from K bytes.

Therefore, the fault status of the opposite node B obtained from the K1 byte is No ALM. Then, according to Table 3 above, the RDI code is added and the SF of the working line WK is added.
Is notified to the switching determination unit 14.

The switching determination unit 14 of the node A adds the SF of the protection line detected by the own node and the RDI of the working line of the opposite node B, and starts SF-R according to Table 3. Through the above operation, all of the nodes A to D select SF-R as the priority and control the switching as shown in FIG.

Further, considering the redundancy control method in the above-mentioned 4-wire BLSR multiplex transmission apparatus, the failure status (SF, SD, No ALM) of the working line WK and the protection line PT.
, The switching request is fixedly determined to be n: 1 or 1: 1 for each span. Then, the highest-priority request is executed by a simple comparison of the priority orders assigned to each switching request.

However, the network is divided into rings.
If the switching operation is determined based only on the fault information that has occurred on the span when the condition for segmentation) is met, the signal that can be saved may be cut off.

That is, according to the technique described above with reference to FIG. 10, when failures 1, 2 and 3 as shown in FIG. 11 occur, both node A and node B and node C and node D are
Performing SF-R, the ring is divided into two. Therefore, a ring segmentation state in which neither signal is rescued is set.

In the case of FIG. 11, the working line WT of node A
Since the signal is at the SD level, the signal is not completely cut off, but the signal is cut off by executing the ring switch.

Therefore, FIG. 12 shows an embodiment method according to the present invention for eliminating such inconvenience. In the embodiment of FIG. 12, the switching determination unit 14 adds the faults of its own node and the opposite node, and the obtained priority is S.
Bits 1-4 of the K1 byte that are F-R and are addressed to other stations
Ring priority is SF-R, ring division (R
ing Segmentation) conditions apply. FIG. 11 shows an example thereof.

As described with reference to FIG. 9, since the switching judgment unit 14 knows the failure status of the working line WK and the protection line PT, the ring division (Ring) is performed based on this information.
Segmentation) conditions are met and working line WK
If the failure condition is SD, the exception process of the priority determination for raising the priority to SF-P is performed according to the flow shown in FIG.

The switching execution unit 15 releases the ring switch (Ring-switch) R of its own node, puts the SF-P request on the transmission K bytes, and releases the ring switch of the other station. As a result, ring segmentation
It is possible to pass the signal of its own node through the working line WK in the SD state by avoiding the above.

By doing so, it is possible to avoid all disconnection of the signal between the own node and the opposite node according to the failure situation.

As a concrete example, as shown in FIG. 12, the SD fault 1 occurs in the working line WK in the node B (node A direction, that is, the waist direction ) in the span between the node A and the node B. B (node A direction ) backup line W
An explanation will be given assuming that SF faults 2 and 3 occur in all the lines between K and the nodes C and D.

[0097] Figure 13 is flow describing the exception processing priority in the processing of FIG. 12 - is. Description will be made with reference to the description of FIG.

At the node A, in the failure collection unit 10,
The SD of the working line WK and the SF of the protection line are detected, and the opposite node failure collection unit 12 detects the absence of RDI of both lines.

Therefore, the switching determination unit 14 sets the priority addressed to the own node to SF-R according to Table 1 (step S1). The K byte collection unit 11 detects that SF-R is being executed by another node (step S
2) Furthermore, ring segmentation (Ring Segmentatio) is performed based on the priority for the local node and the priority for other nodes.
n) It is determined that the condition is met (step S3).

Furthermore, the priority of failure is SF-R.
If the ring segmentation condition is met, it is determined whether the working line WK is SD or SF (step S4).

When there is no RDI on the working line, from Table 1, neither node A (node B direction ) nor node B (node A direction ) is in SF state. In other words, it is SD or No ALM, and if both are added, it can be determined that the fault on the working line WK is SD. At this time, the priority of the own node is SF-
Increase to P (step S5).

The switching execution unit 15 of the node A releases SF-R and starts SF-P according to the priority notified from the switching determination unit 14. In the node B, since both the working and protection lines of the own node are No ALM, SF-R is first executed according to the switching request from the node A by K bytes.

When the switching request from the node A changes to SF-P, the request is followed and the ring switch (Ring-switc
Release h) and execute SF-P. The node C executes SF-R due to the failure of its own node, and when the switching request from the node A changes to SF-P, follows the request and releases the ring switch.

Node D receives SF-R due to the failure of its own node.
Ring switch (Ring-switch)
To cancel.

In this way, the switching state finally becomes as shown in FIG. 12, and the signal disconnection between the node A and the node B is avoided.

Here, the switching judgment unit 14 is
As described with reference to FIG. 12, the Ring S
egmentation) conditions are met and the backup line PT
If the failure status of the node is SD (step S6), the switching priority of its own node is set to SF as shown in the flow of FIG.
-It is determined to increase to S (step S7).

The switching execution unit 15 releases the ring switch (Ring-switch) R of its own node, executes SF-S, puts an SF-S request on the transmission K byte, and sends the ring switch (Ring) of another station. -switch) is released. As a result, ring segmentation is avoided, and the signal of its own node is passed through the protection line PT in the SD state. By doing so, it is possible to avoid a complete disconnection of the signal between the own node and the opposite node depending on the failure situation.

As a specific example, as shown in FIG. 14, in the span between the node A and the node B, the SD fault 2 occurs in the node B (the protection line PT in the direction of the node A, and the node B (node A
A description will be given assuming that SF faults 1 and 3 occur in all the lines between the working line WK and the nodes C and D in the direction.

In this case, the node A detects the SD of the protection line PT and the SF of the working line WK in the failure collection unit 10. The opposite station failure collection unit 12 detects the absence of RDI on both lines, and the switching determination unit 14 sets the priority of its own node to SF-R, and the K byte collection unit 11 sets SF-R at other stations.
Detects that R is being executed.

Further, the switching judgment unit 14 judges that there is a ring segmentation condition based on the priority of the own node and the priority of the other station, and FIG.
Exception handling of 3 is performed.

If the priority of the failure is SF-R (step S6: FIG. 3), the protection line PT is SD, and there is no RDI in the protection line PT, from Table 1 node A
(Neither node B direction nor node B (node A direction is in SF state.

That is, it is SD or No ALM, and when both are added, it can be determined that the failure of the protection line PT is SD, and the priority of the own node is raised to SF-S (step S7: FIG. 13). The switching execution unit 15 of the node A releases SF-R and starts SF-S according to the priority notified from the switching determination unit 14.

At the node B, since both the working and protection lines of the own node are No ALM, SF-R is first executed according to the switching request from the node A by K bytes, and the switching request from the node A is issued. When it changes to SF-S, it follows it, the ring switch (Ring-switch) R is released, and the span switch (Span-switch) S is executed.

Node C receives SF-R due to the failure of its own node.
When the switching request from node A changes to SF-S, the ring switch (Ring-switc
h) Release R. Node D is S due to its own failure
When the F-R is executed and the switching request from the node B changes to SF-S, the F-R is followed and the ring switch (Ring-switc
h) Release R.

In this way, the switching state finally becomes as shown in FIG. 14, and the signal disconnection between the node A and the node B is avoided.

Here, in the ring network using the BLSR multiplex transmission device, cross-connection is usually performed by selecting a line having a short distance. In the large-scale ring network as shown in FIG. 13, when connecting the signal connected to the node 3 to the nodes 4 to 6, a clockwise circuit is used, and the nodes 8 to 10, the node 1, and the node 2 are connected to each other. When connecting, use the counterclockwise line.

As shown in FIG. 15, node 3 and node 4
Between the nodes 6 and 7, a failure occurs in the span between the node 6 and the node 7, and the ring switch R is performed to enter the ring segmentation state. in this case,
Node 4 as represented on the ring configuration shown in FIG.
~ The signal connected to node 6 is not rescued.

Assuming that signals are evenly connected from the node 3 to the nodes 4 to 7, the more signals that can be rescued, the more signals are rescued.

The K byte collection unit 11 receives K bytes from K bytes.
The originating station ID shown in bits 1-4 of the 2 bytes is extracted. When the originator node determines the originator node from the originator ID, if the priority indicated in bits 1-4 of the K1 byte is SF-R or FS-R, the far end node (Far End Node) is K. Suppose it is the source node of the byte.

Then, the switching judgment unit 14 performs ring segmentation (Ring Segment Ratio) according to the judgment flow shown in FIG.
n) When it is determined that the condition is satisfied (step S3: FIG. 13), the number of spans from the span switched by the own node to the span switched by the far end node (Far End Node) in the east direction and the waist direction is determined. Count around both directions.

From this count, both

[0122]

[Equation 1]

In the case of, the ring segmentation cancellation condition described with reference to FIGS. 10 and 12 is determined, and the ring switch (Ring-switch) or other switching is selected.

Also, either one of

[0125]

[Equation 2]

In the case of, by executing the ring switch without judging the ring segmentation cancellation condition described with reference to FIG. 12 and FIG. Change the switching control. By doing so, it becomes possible to select the switching with a high possibility of signal relief.

Explaining in a concrete example, when a failure as shown in FIG. 15 occurs, the nodes 3, 4, 6 and 7 perform a ring switch.

Since the failure occurs at two places, the K byte communication on the long path side is performed between the node 3 and the node 7 and between the node 4 and the node 6, and the K byte of each node is used. The collection unit 11 determines that the far end node (Far End Node) of the node 3 is the node 7 and the far end node (Far End Node) of the node 4 is the node 6.

According to the determination procedure flow shown in FIG. 17, the node 3 is switching in the east direction. Therefore, when the number of nodes is counted up to the far end node (Far End Node) in the east direction and 1 is subtracted, the number of spans is 3 (step SI). Farthest end node in the waist direction (Far E
number of nodes up to nd Node = 6 plus 1 and the number of spans is 7
(Step SII).

In both cases

[0131]

[Equation 3]

Is exceeded. Since the node 4 is switching in the waist direction, the number of nodes is counted up to the far end node (Far End Node) in the east direction and 1 is added, and the number of spans is 3.

Far end node in the waist direction (Far End Nod
The number of nodes up to e) = 8 minus 1, and the number of spans is 7. both

[0134]

[Equation 4]

Is exceeded. Here, when the failure of the working line WK between the node 3 and the node 4 is SD, the switching determination units 14 of the node 3 and the node 4 cause the ring switch (Ring
-switch) is released and the priority of the own node is set to SF-
Set to P (step SIII).

By these operations, the nodes 3, 4, 6,
Since No. 7 releases the ring switch, signals connected from the node 3 to the nodes 4, 5 and 6 are relieved through the working line WK between the nodes 3-4. Further, the signal connected from the node 3 to the nodes 7 and 8 breaks the relief route and becomes a signal disconnection. In this way, the switching control with the higher probability of relieving the signal is performed.

Here, the error correction function in the redundancy control in the conventional BLSR multiplex transmission apparatus will be considered. K-byte continuity received, defined code (Co
de) Consistency with Ring Topology
If it is incorrect, the received K bytes are discarded, and the last received K bytes is regarded as correct and B
The LSR multiplexing transmission device operates.

The line switching operation in the BLSR multiplex transmission apparatus is required to be immediacy, and it is necessary to deal with a failure that is unknown when it will occur. Therefore, unlike ordinary data communication, it does not have an expected reception value, an acknowledgment (acknowledge), or a re-sending function, but only returns a predetermined response to the received value.

Therefore, if there is a problem in the protocol arrangement or the control device, the self-node and the opposite node react with each other and the switching state does not converge.
pong).

According to the conventional technique, although it is normal to see the minute time and the state of a single node as described above, the detection of an abnormal state and the escape from an abnormal state are not considered as a whole. It was

Therefore, as a first embodiment of the present invention which copes with such a problem, a FIFO memory is prepared, and data is stored in the FIFO memory each time the transmission / reception protocol data changes. At the same time, start the timer. The timer measures a time T1 for assuming that the transmission / reception protocol data is stable if the data does not change within a certain time, and a time T2 for assuming that the data becomes unstable if the data continues to change for a certain time or more. However, T1 <T2.

If the protocol becomes unstable, FI
The stack (stacking) in the FO memory is stopped, and the data matching the oldest data indicated by the stack pointer (SP) is searched from the FIFO memory.

If there is no matching data, the data is not circulated, and therefore the stacking into the FIFO memory is restarted. If there is data that matches the oldest data, the number of data n between the data that matches the oldest data is stored, and a match search is also performed for data that is one newer than the oldest data. When this is repeated n times and all match, the transmission / reception protocol data is circulated and an alarm is issued.

By doing so, it becomes possible to detect the circularity of the transmission / reception protocol data and detect the abnormal state of the switching control.

Specifically, the configuration of FIG. 18 is shown as the protocol abnormality detection unit 18 of FIG. As shown in FIG. 18, a transmission / reception protocol change detection unit 20, a stability detection unit 21, a FIFO memory 22, and a circularity detection unit 23 are prepared. The transmission / reception protocol change detection unit 20 transmits the transmission K.
Monitor for changes in bytes and received K bytes.

Each time a change in the transmitted K bytes and the received K bytes is detected, the change is stacked in the FIFO memory 22, and at the same time, the first timer 210 of the stability detector 21 is reset,
The first and second timers 210 and 111 are set to the counting state.

The count time of the first timer 210 is set to T
1, and the count time of the second timer 211 is T2. When T1 <T2, if the transmission / reception protocol change detection unit 20 does not detect a change for a period longer than T1, the first timer 210 overflows and the second
The timer 211 of is reset.

If the transmission / reception protocol continues to change within a time period shorter than T1, the first timer 210 repeats counting from the initial value, the second timer 211 continues counting, overflows after T2 time, and an error occurs. Detect steady state.

Therefore, the above-mentioned first timer 210
The relationship between the count of the second timer 211 and the overflow is as shown in FIG.

The FIFO memory 22 stores the unstable transmission / reception protocol data so far. The cyclicity detector 23 detects whether the data stored in the FIFO memory 22 is circulating.

The stack pointer of the FIFO memory 22 indicates the position of the oldest data d0. The circularity detection unit 23 searches the FIFO memory 22 for data d0 ′ that matches the data d0 stored here. If there is no matching data, it does not cycle.
If there is the matching data d0 ', the circularity is confirmed.

As shown in FIG. 20, when the data is circulating, the data d1 to d3 between d0 and d0 '' is d
Matches 1'-d3 '. Therefore, the circulation detecting unit 23 detects that all of these data match, and issues an alarm.

Further, as another method for dealing with the possibility of falling into the oscillation state (Ping-pong) in which the switching state does not converge, a state transition counter, a transition rate measuring unit, and a comparator are prepared. The state transition counter is the switching execution unit 15
(See Fig. 6) Monitor the output of the ring switch (Ring-s
witch) R, span switch (Span-switch) S, or incrementing the count value for each change K of the transmission K byte.

Further, if the count value of the state transition counter is 1 or more, the transition rate measuring unit decreases the state transition counter value at regular time intervals, and the comparing unit constantly compares the state transition counter value with a constant threshold value. , The comparison result is an alarm.

By doing so, it becomes possible to detect an abnormal state of switching control which should normally be completed in several steps.

To explain the specific configuration, the protocol abnormality detection unit 17 of FIG. 6 is configured as shown in FIG. 21, and the state transition counter 30 monitors the output of the switching execution unit 15 and detects a change each time. Count up. The transition rate measuring unit 31 periodically gives a subtraction instruction to the state transition counter 30, and the comparing unit 32 causes the state transition counter 30 to operate.
And the threshold value 33 is compared.

In the case of a simple span switch, the output of the switching execution unit 15 makes a transition as shown in FIG. Number of transitions is node (Node) 1, node (Node)
Both 2 are twice, and the time required for the transition is specified to be 50 ms or less.

To provide a margin, the transition rate measuring unit 3
10 ms for the subtraction instruction to the state transition counter 30 of 1
Each count, one countdown per subtraction instruction,
FIG. 23 shows the transition of the count value of the state transition counter 30 of the node (Node) 1 of FIG. 22 when the threshold value is set to 4 times.

The state transition counter 30 counts up due to the state transition of the switching execution unit 15 at T1 and T2, and the subtraction instruction asynchronous with the switching execution unit 15 causes
Count down every 10 ms.

When the state of the switching execution unit 15 does not become a steady state due to some trouble and the state transition continues, the count value of the state transition counter 30 eventually exceeds the threshold value SL, and the comparison unit 32 issues an alarm.

Further, as an abnormal control state, as shown in FIG. 24, different requests are mixed between the own node and the opposite node, and there is circulation. In such a case, the configuration described above with reference to FIGS. 18 and 21 detects the circulating property and issues an alarm.

Upon receiving the alarm, the switching execution unit 15 enables the delay unit 16 to delay the switching control when the priority is higher than the previous value, and control so as to disturb the cycle of circulation. As a result, as shown in FIG. 25, the circulation is lost and the stable state S is reached. By doing so, it is possible to shift to a stable state regardless of the cause of the abnormal state.

As a specific example, when a protocol abnormal state as shown in FIG. 24 is encountered, an alarm is issued. The switching execution unit 15 enables the delay unit 16 (see FIG. 6) by this alarm.

Priority of request A and request B is request A
In the case of the relationship of request B, as shown in FIG. 25, when the output of the switching execution unit 15 of the node (Node) A and the node (Node) B changes from the request A to the request B, from the input of the request The time until the output is the same as usual, and when changing from the request B to the request A, it takes a long time from the input of the request to the output, and the request A is overwritten by the request B before long. It disappears from above and the stable state S is reached.

[0165]

As described above with reference to the drawings, according to the present invention, it is possible to select and execute an appropriate switching priority in any combination of obstacles and avoid a signal interruption as much as possible. Yes, 4-line BL
It is possible to improve the reliability of the SR multiplex transmission device.

The above embodiment of the present invention is for understanding the present invention, and the scope of protection of the present invention is not limited to this. What is equivalent to the composition described in the claim is also contained in the scope of protection of the present invention.

[Brief description of drawings]

FIG. 1 is a transmission apparatus BLSR which is a target of the present invention.
It is a schematic block diagram which shows the structural example of the network which used the multiplex transmission apparatus.

FIG. 2 is a diagram illustrating the definition of a line between two transmission devices arranged in a network.

[Fig. 3] Network for SONET (Synchronous Opti
cal NETwork) is the frame format of the signal.

FIG. 4 K1, K of SONET frame format
It is a figure explaining the content of 2 bytes.

FIG. 5 is a diagram showing a network configuration according to the first embodiment of the present invention.

FIG. 6 is a diagram showing a configuration example of a BLSR redundancy control unit 1 according to the present invention in FIG.

FIG. 7 is a diagram illustrating an example of priority addition using a register.

FIG. 8 is a processing flow when priority addition is performed by conditional branching.

FIG. 9 is a diagram illustrating a case where the line cannot be repaired by the method of FIG.

FIG. 10 is a diagram illustrating a second embodiment according to the present invention, which addresses the problem of FIG. 9;

FIG. 11: Ring segmentation by the method of FIG.
FIG. 3 is a diagram illustrating a problem of a (ntation) state.

FIG. 12: Ring segmentation of FIG.
It is a figure explaining one Example method according to this invention which eliminates the inconvenience of a state.

FIG. 13 is a flowchart illustrating an exception process of priority determination in the embodiment of FIG.

14 is a diagram illustrating an example of repairing a signal stage by exception processing of the priority determination of FIG.

FIG. 15 is a diagram illustrating a relief range for ring segmentation.

FIG. 16 shows an example of the ring topology (Ring topol) of FIG.
ogy) is a diagram expressed above.

FIG. 17 is a determination flow for determining a repair range for ring segmentation.

FIG. 18 Detects the circularity of transmission / reception protocol data,
It is a figure which shows 1 Example structure for detecting the abnormal state of switching control.

FIG. 19 is a diagram showing the relationship between the count and overflow of the first timer and the second timer in FIG.

20 is a diagram for explaining the contents of the FIFO memory in FIG.

FIG. 21 is a diagram illustrating a configuration example of circularity detection of transmission / reception protocol data using a count value of a state transition counter.

FIG. 22 is a diagram illustrating normal state transition in the case of a span switch (Span-switch).

FIG. 23 is a diagram for explaining the relationship between the transition of the count value of the state transition counter of the node 1 of FIG. 22 and the threshold.

FIG. 24 is a diagram illustrating a state in which line switching requests are circulated.

FIG. 25 is a diagram for explaining the multiple cases of the circulating state of FIG. 24.

[Explanation of symbols]

100 BLSR multiplexing transmission device 1 BLSR redundancy controller 2 signal switching section 20,21 Multiplexer 30,31 Demultiplexer 10 Obstacle Collection Department 11 Kbyte collection unit 12 Opposite station failure collection unit 13 Fault adder 14 Switching judgment unit 15 Switching execution unit 16 Delay section 17, 18 Protocol error detector

─────────────────────────────────────────────────── --- Continuation of the front page (56) References JP-A-10-117175 (JP, A) JP-A-10-290249 (JP, A) JP-A-2000-41056 (JP, A) JP-A-2000-134244 ( (58) Fields investigated (Int.Cl. ⁷ , DB name) H04L 12/437

Claims (11)

(57) [Claims] 1. In a network configured by connecting a plurality of nodes each provided with a four-wire bidirectional multiplexing transmission device with a working line and a protection line, one node of nodes adjacent in one span, said current <br/> for circuit and to detect the importance of the node failure spare lines, and wherein said current <br/> for lines is detected by the other node of the adjacent nodes and, a spare line the receive level failure, and severity of the local node failure, the of the importance of the fault detected by the other node of the adjacent nodes, the high importance
Redundant line control method characterized by performing addition processing to obtain the fault of one of them as the addition result and determining the priority of line switching 2. A method according to claim 1, wherein the level of the node failure, the addition level of fault detected by the other node of the neighboring node, the line corresponding to the level of the disabled <br/> harm A redundant line control method characterized in that one having a higher switching priority is validated. 3. The fault level of the working line and the protection line detected by the other node of the adjacent node received by the one node according to claim 1, wherein the level of the fault in the SONET format is K bytes. A redundant line control method, characterized in that the redundant line is transmitted by the. 4. The remote defect indication (RD) of the working line and the protection line carried in the K bytes according to claim 3.
A redundant line control method characterized in that I: Remote Defect Indication) is used as a trigger for line switching of a failure corresponding to a signal failure (SF: Signal Failure) of its own node. 5. The method according to claim 1, wherein a fault occurs in both the working line and the protection line in two or more spans, and at least one of the working line and the protection line has a predetermined error rate. a more failures, when the ring connection is interrupted (ring segmentation), the failure of the working line, when a failure of the lower error rate than the predetermined error rate, prioritization of the station
A redundant line control method characterized by performing exception processing in which a priority is given to a switching request with a higher degree of importance of line switching. 6. The predetermined error rate according to claim 5, wherein the predetermined error rate is 10 ⁻³ to 10 ⁻⁵ (SF: Si).
a gnal Failure), the predetermined lower error rate than the error ^{rate, 10 -6 ~10 -9 (SD:} Signal Deg
rade) is a redundant line control method. 7. Oite to claim 6, wherein when the ring connection falls into condition being separated, the number of nodes on the ring, the ring topology (Ring Topology showing the positional relationship of each node on the ring ) and from the signal capable end node reached on the <br/> ring topology, the position of the span in which the far-end node is switched,
When the own node is farther than (total number of nodes / 4) counted from the adjacent span, the ring switch release determination is enabled, and when it is closer than (total number of nodes / 4), ring switching is always performed. A redundant line control method characterized by the above. 8. A four-wire bidirectional multiplexing transmission device arranged in each of the plurality of nodes in a network configured by connecting a plurality of nodes with a working line and a protection line, A line failure collection unit that detects the importance of the failure of the working line and the protection line, a K byte collection unit that collects K bytes of the SONET overhead bytes, and a failure of the own node collected by the line failure collection unit . Heavy
And iodine, wherein among the severity of the disorder of the opposite node that is collected by the K bytes collecting unit, adding binding failures higher importance
Performs addition processing to obtain a result, both 4-wire, characterized in that it comprises a switching determination unit for determining a switching priorities, the line switching unit to perform a line switch in accordance with the switching priorities which are determined by the switching determination unit Directional multiplexing transmission equipment. 9. The protocol abnormal state detecting unit according to claim 8 , further comprising a protocol abnormal state detecting unit for detecting a ping-pong state of the protocol, wherein the protocol abnormal state detecting unit includes a FIFO memory and a change detecting unit for detecting a change in the K byte. the change detecting unit each time detected by, when a FIFO memory for storing data of K bytes, timer and the timer starts counting at the timing of the change detection of said change detection unit has counted the predetermined number, the FIFO A four-wire bidirectional multiplexing transmission device, characterized in that the memory is provided with a cyclicity detector for judging whether or not the same data is repeated. 10. The method of claim 8, further comprising an abnormal state detecting section for detecting an abnormality of the switching protocol, the abnormality detecting unit includes a counter for counting the number of transitions line switching state, from the count value of the counter a transition rate measuring unit that calculates the rate of occurrence of state transition in a predetermined period, a comparator unit for comparing the incidence of state transition is calculated by the transition rate measurement unit with a predetermined value, the incidence of the state transition A four-wire two-way multiplex transmission apparatus having means for issuing an alarm of a switching protocol abnormality when the value exceeds a predetermined value. 11. The method of claim 10, further wherein the abnormality detecting unit holds the previous value of the switching priorities, the new switching priority is lower than the destination switching priority, the switching control delay a delay unit for performing a delay operation of the delay unit, when detecting the switching protocol anomaly alarm from the means for issuing an alarm of the protocol anomaly, 4-wire bidirectional multiplex, characterized in that an effective Transmission equipment.