JPH04100343A - Atm link system - Google Patents

Abstract

PURPOSE:To allow the system to cope with a fault of a transmission line quickly by setting a virtual path VP for the occurrence of a fault of the transmission line in advance so as to have only to implement loopback at a bus handling system adjacent to a fault point on the occurrence of a fault of the transmission line. CONSTITUTION:An active system VP (VPphi) is set in a direction of a shortest route toward a bus handling system 212 of an object point from a bus handling system 214 being a starting point and a VP 1 of a standby system in the opposite direction is set simultaneously. A bus handling system 213 detecting a fault loops back a cell flow going to a fault occurrence point forcibly and a cell of a VP of an active system is decoded as a cell of a VP of a standby system automatically on a looped back transmission line and carried up to a point B and looped back and carried up to an incoming bus handling system 212 while being restored to a cell of the VP of the active system. Thus, the bus handling systems at both sides close to a transmission line fault point have only to be simply looped back on the occurrence of the fault of the transmission line and the system quickly copes with a 7 fault of the transmission line.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention (Industrial Application Field) The present invention relates to an ATM (Asynchronous
The present invention relates to an ATM link system for providing a transmission line network using so-called VP (Virtual Path) to TM switching facilities, users/ATM switching facilities, etc. in a broadband communication network using nsfer mode technology.

(Prior art) In recent years, STM (Synchronous Transf.
ATM (＾5ynchron
There is growing interest and expectation in ousTransfer Mode).

In ATM, all information is broken down into fixed-length blocks called cells and an identification header is added to each block for unified information transmission, and each communication terminal side passes the cells to the communication network as necessary. This is a transfer mode in which the communication terminal uses the information transmission capability of the communication network when necessary.

At ATM, VC (Virtual C1rcuit)
Communication services are provided using two types of theoretical connections: virtual path and VP (Virtual Path).

A VC is a connection that is established for each call. When a certain terminal A communicates with a certain terminal B, the terminal A issues a call setup request to the terminal B to the communication network. The communication network (specifically, the ATM switch) searches for a route from terminal A to terminal B and at the same time checks the presence or absence of communication resources. If communication resources exist and terminal B is also in a state where communication is possible, the communication network is connected to terminal A.
A VC from Terminal B to Terminal B is set up on the above route. After that, terminal A is allowed to communicate with terminal B, but terminal B
All addressed cells are sent to terminal B through the VC. After the communication ends, the VC is released.

On the other hand, VP (Llrtual Path) is a semi-fixed connection that provides a theoretical transmission network that does not take the form of a physical transmission network between an ATM switch and a user. It is an object. The above-mentioned VC is a transmission line network theoretically established in this way.
It will be set through P. In the present invention, a system that provides a theoretical transmission network using VP is called an ATM link system. The ATM link system consists of a physical transmission path and a path handling system that handles VPs.

Now, various ATM link systems have already been proposed, one of which is a system in which the physical transmission path is composed of two rings in opposite directions, as shown in Figure 2.
S, 0hta, K, 5ato, 1. Tokjzav
a.

”A Dynamically Contro
e ATM Transport Network b
ased on the Virtual Path
Coneept, - Figure 5 of GLOBECOM'88 (
b)). This system has the advantage that the physical transmission path requires only two rings and is easier to install than other systems.

However, it is still insufficient for practical use.

(Problems to be Solved by the Invention) The first problem with the conventional technology is how to deal with transmission line failures. For ring-shaped ATM link systems, a conventionally proposed method for dealing with transmission path failures is loopback as shown in FIG. For example, when a failure occurs in the transmission line at point A, the path handling system adjacent to point A loops back as shown in the figure to deal with the transmission line failure. However, simply physically looping back like this will not solve the problem. This will be explained using FIG. The VPO cells are looped back as shown. However, since no VPO is defined on the transmission path in the reverse direction, the path handling system on the transmission path in the reverse direction does not know where to transfer the cells of this VPO.

In the end, it is determined that the VP is undefined and the VPO cell is discarded. This means that it is not enough to physically loop back as shown in FIG. 3, and it is necessary to rewire the VPO as shown in FIG. 4. However, in order to replace the VP, it is necessary to rewrite the information regarding the VP in almost all path handling systems. Originally, physical loopback was
VP replacement contradicts this idea because it attempts to quickly respond to a transmission line failure by changing only the configuration of the path handling systems on both sides of the transmission line failure point. Further, even if a method of replacing the VP is adopted instead of quickly responding to transmission path failures, problems with the transmission band still occur. That is, when looping back as shown in FIG. 4, the looped back traffic is carried on the reverse direction transmission path in addition to the traffic that was originally accommodated. This may cause traffic congestion on the reverse transmission path, and communication quality such as delay quality and cell discard quality may deteriorate rapidly. ATM',
It is unacceptable for communication quality to deteriorate below the required characteristics in a transmission line network such as a link system. In addition, there was a problem in that methods for detecting transmission line failures and methods for detecting transmission line failure recovery were not planned.

The second problem with the current technology is that a one-way branch connection service has not been provided. The link system may be used for conventional broadcasting services. To this end, it would be convenient if the cells input from one path handling system were branched and connected in each path handling system, as shown in Figure 5, and the ATM link system as a whole could provide a one-way branch connection service. . Although the need for such a service has been voiced, no practical implementation method has been provided to date.

A third problem with the techniques up to now is that no means have been provided for confirming that the VP connection has been properly established. The setting of the VP connection is performed in a distributed manner by the related path handling system based on instructions from the operation center within the TM link system.

As described above, since the VP connection is a logical transmission path, extremely high reliability is required. In the unlikely event that a wrong VP connection is set up, cells should not be transmitted to the wrong path handling system or cells should be lost. Therefore, after setting up a VP connection, it is important to provide a means for checking whether the correct VP connection has been set up. However, such a means (herein referred to as a VP connection connection test) has not been provided in the past.

In view of the above points, the present invention provides a method for quickly responding to a transmission line failure by changing only the configuration of the path handling system adjacent to the transmission line failure point, and even in the event of a transmission line failure. It is an object of the present invention to provide a method that does not cause traffic congestion, a method for realizing a one-way branch connection, and a method for testing a VP connection connection.

[Structure of the Invention] (Means for Solving the Problems) In order to achieve the above-mentioned objects, the present invention provides a clockwise and counterclockwise rotation system that combines the multi-pass handling system and the above-mentioned pass handling system. The path handling system is composed of a ring-shaped transmission path, and includes: a branch/insertion section connected to the clockwise ring-shaped transmission path; and a branch/insertion section connected to the counterclockwise ring-shaped transmission path. An ATM link system consisting of a branch/insert section, an ATM switch section connected to both the branch/insert sections, an in-office JNF section connected to the ATM switch section, and a loopback line connecting both the branch/insert sections. In order to solve the above problem, the present invention is characterized by having the following means.

(1) When setting a working VP between path handling systems, be sure to also set a backup VP. Set the VP so that it goes around the path handling system that is the destination point where you want to set the VP, and ends at the path handling system that is the starting point, and the above-mentioned backup system VP accommodates the above-mentioned active system VP. It is set to circulate on a ring-shaped transmission path facing in the opposite direction to the ring-shaped transmission path, and the VPI of the working VP and the protection VP are set to be the same at the same location.

(2) In the path handling system, which is the original destination point, the user cells carried through the working VP are branched and only the OAM cells are relayed, thereby establishing the starting point for setting the VP. It is possible to return from the path handling system of the path handling system through the VP to the starting point path handling system where you want to set the VP. Check.

(3) When a fault occurs at one point on the ring-shaped transmission path, the path handling systems on both sides adjacent to the fault point use the loopback line to completely control the cell flow toward the fault point. Control is performed to loop back to the ring-shaped transmission path in the opposite direction, and on the ring-shaped transmission path in the opposite direction, the cell flow is carried to the path handling system on the opposite side of the failure point using the backup VP. As described above, by looping back again to the path handling system on the opposite side of the fault point, the cell flow carried by the protection VP is transferred to the working VP again, thereby dealing with the transmission path fault. do.

(4) The bandwidth reservation method for the backup VP is based on the above-mentioned active VP.
The sum of the bandwidths of all active VPs accommodated in the ring-shaped transmission path in which the above-mentioned protection VP is accommodated is calculated for each path handling system, and the maximum value is calculated for the bandwidth of all the working VPs accommodated in the ring-shaped transmission line in which the above-mentioned protection VP is accommodated. The bandwidth is reserved as a shared protection band on the ring-shaped transmission path, thereby minimizing the reservation of the protection band.

(5) In addition to the three basic functions of dropping, inserting, and relaying, the branching/inserting unit has a function of branching to the ATM switch while relaying cells input from the incoming transmission line to the outgoing transmission line. By doing so, it is possible to perform one-way branch connection for the entire ATM link system.

(Function) According to the present invention, since a VP for use in the event of a transmission path failure is set in advance, when a transmission path failure occurs, the transmission can be quickly transmitted by simply physically looping back in the path handling system adjacent to the failure point. Can deal with road obstacles. Furthermore, since the minimum required bandwidth is secured in advance for the VP for use in the event of a transmission path failure, communication quality will not deteriorate even in the event of a transmission path failure.

Also, by setting the VP to go around the VP
Connection tests can be easily performed. Furthermore, by providing a function of simultaneously copying and branching while relaying to the branch/insertion unit in the path handling system, one-way branch connection is possible for the TM link system as a whole.

(Example) FIG. 6 is a diagram showing the overall configuration of an ATM link system. 311 to 31n are branch/add circuits for the clockwise ring-shaped transmission path, and 321 to 32n are branch/add circuits for the counterclockwise ring-shaped transmission path. Lines 351 to 35n connecting the two add/drop circuits are for loopback in the event of a ring-shaped transmission path failure. In the drop/add circuit, when a cell arrives from the input transmission path, the cell's VPI is used as an index to search the header conversion table in the drop/add circuit to instruct the routing of that cell (Zuna). Either relay the cell to the next path handling system, branch it at the branch/add circuit, branch to the next path handling system and at the same time branch its copy to the branch/add circuit, or branch to the branch/add circuit.・If branching is done with an insertion circuit, there are multiple I
A routing tag for determining which NF to output from) is added to the cell, and VPI conversion is performed at the same time. Thereafter, within the path handling system, the cell is transferred based on the suiting tag. If the cell is a cell to be transferred to the next path handling system, the cell is preferentially output to the outgoing transmission path (without delay). When the cell is branched by the branch/add circuit, it is first branched to the ATM switches 331 to 33n, and then transferred to the target intra-office INF 341 by the ^TM switches 331 to 33n.
~34n. The intra-office INF removes the routing tag and outputs the cell. On the other hand, in-station I
When placing cells from the NF onto a ring-shaped transmission path, the following measures are taken. First, the in-office INF adds a routing tag to the cell by searching the header conversion table in the in-office INF using the cell's VPI as an index, and simultaneously performs VPJ conversion. Thereafter, within the path handling system, the cell is transferred based on the routing tag. That is, either a clockwise ring transmission path or a counterclockwise transmission path is selected depending on the routing tag. Thereafter, cells are inserted into the transmission path in the drop/add circuit, and this is done during time slots in which there are no cells to be relayed from the input transmission path to the output transmission path. Furthermore, the branch/add circuit removes the routing tag before outputting the cell to the output transmission path.

Although the speed of the intra-office INF, the number of intra-office INFs, and the speed of the transmission path vary depending on the application, in one embodiment, the intra-office INF
The speed may be −155.52 Mb/s, the intra-office INF −32, and the transmission line speed −2.4 Gb/s. Further, it is not necessary that there be one ring-shaped transmission line in each direction. For example, a configuration in which there are 16 transmission lines of 155.52 Mb/s in each direction is also possible. In this case, there will be 16 branch/add circuits in each direction. Furthermore, in this case, 16 pieces of 155.521
The 7s transmission line can also be physically multiplexed into a 2.4Gb/s transmission line for transmission.

Next, the VP setting method that is characteristic of the present invention will be explained in accordance with the seventh zu for the case of point-to-point connection. A working VP (VPO) is set in the direction of the shortest route from the intra-office INF of the path handling system 211 serving as the starting point to the intra-office INF of the path handling system 214 serving as the destination. At the same time, the path handling system 214 at the destination
is branched and extended so that it goes around all the way to the originating path handling system 211. This branch condition is set as follows. That is, the branching condition of the destination path handling system 214 is set so that the user's cell is dropped by the calling path handling system 214, but the TM layer OAM cell is returned to the originating path handling system 211. In this way, after setting VPO ^TM
By transmitting the layer 0^M cell and confirming that it returns to the originating path handling system 211, it is possible to easily confirm whether the VP connection has been set up normally. Since only OAM cells for testing will flow in the expanded portion, only a very small amount of bandwidth will be secured. Of course, the originally required bandwidth from the originating path handling system 211 to the receiving path handling system 214 is set to the amount originally required for the VP. Therefore, even for the same VP, the bandwidth secured will be different between the originally necessary part and the extended part. Furthermore, in addition to the VPO, a loop-shaped reverse backup system VP (VPI) is set. As will be described later, the backup VP is used for loopback when the transmission line used by the active VP fails. VP1 of the active system VP and standby system VP is set at each point 221~ as shown in the figure.
225 so that they are the same.

Next, a VP setting method in the case of one-capable branch connection will be explained with reference to FIG. Working system V that goes around the ring-shaped transmission line
P is set, and each path handling system 211 to 21
In step 5, the active VP is branched as necessary. At the branch point, a copy of the cell of the active VP is created, relayed to the next path handling system, and simultaneously dropped to the local INF. In addition, the current VP
In addition, a backup system VP in a loop-shaped reverse direction is set. As will be described later, the backup VP is used for loopback when the transmission line used by the active VP fails.

The VPI of the active VP and the backup VP are set to be the same at each point 221 to 225 as shown in the figure.

By setting the active VP as described above, a VP connection connection test can be easily performed as in the point-to-point case. This will be explained using FIG. 9. The originating pass handling system 211 is A
Insert the TM layer OAM cell from the local INF. This OAM cell is copied to each path handling system 212 to 215 as configured in the VP.
It is branched into NF. At the second time, a copy ID is written in the information field of the copied OAM cell. In the intra-office INF, if it is a TM layer OAM cell, it is looped back and placed on the transmission path again. The originating path handling system 211 can confirm the VP connection by returning OAM cells with copy number + 1#. It is possible to estimate whether the system is abnormal.Please note that the copy ID is
A circuit called D-RTA in the drop/add circuit writes into the information field of the M cell. Details will be discussed later.

Next, we will show how to deal with transmission line failures. First, a method for detecting a transmission path failure will be explained with reference to FIG.

The add/drop circuit monitors the incoming transmission line for reception signal interruption. In addition, each path handling system 211 to 21
5 periodically sends physical layer OAM cells to the adjacent path handling system and confirms that they are looped back. If an abnormality occurs in any of these, it is determined that a transmission path failure has occurred.

When a transmission path failure is detected by the method described above, the path handling system that has detected the failure forcibly loops back all cell flows heading toward the point where the failure has occurred. Through this operation, the path handling systems on both sides of the transmission line fault occurrence point can detect the transmission line fault, and a loopback is performed as shown in FIG. 11. Current system VP<VP
Cells belonging to point A) are also looped back at point A, but as mentioned above, the active VP and backup VP (VPI
) is set to be the same at each point, so on the looped-back transmission path, cells of the working VP are automatically interpreted as cells of the protection VP, and the path handling of point B It will be carried to system 212. By looping back at point B, the cells of the protection VP return to the cells of the active VP and are carried to the destination path handling system 212. In the conventional system, when a transmission path failure occurs, it is necessary to rewire the VP, but according to the method provided by the present invention, there is no need to rewire the VP at all. Simply fold back the path handling systems on both sides closest to the transmission line failure point. Thereby, transmission path failures can be dealt with quickly. However, when a failure occurs and a loopback is necessary, the path handling system 212 at point A sends an OAM cell to the backup path that declares that cells will begin flowing to the backup path from now on.
It is assumed that a backup VP connection is activated. Normally, the backup VP link is deactivated so that if a cell enters, the cell is immediately discarded.

This is because the backup VP connection goes around the ring once, so if a cell gets mixed into that VP connection (due to a header error, etc.), the cell will rotate around the ring forever and will not be set on the same ring. This is because there is a possibility that the traffic of the currently active VP may be compressed.

As mentioned in the "Prior Art" section, when looping back, the transmission path becomes congested, and there is a possibility that the transmission quality will deteriorate below the required characteristics. Since transmission quality is very important in a link system, it is unacceptable for the transmission quality to deteriorate below the required characteristics even in the event of a failure. In order to prevent transmission quality deterioration, a method can be considered in which the same band as the active VP is allocated to the protection VP. However, if bands are allocated to all backup VPs, there is a possibility that the effective transmission path usage rate under normal conditions will be extremely reduced.

Therefore, the first method of efficient bandwidth allocation for standby VPs is
This will be explained according to Figure 2. In (1) of the figure, VPO
Assume that I-VPO5 is the active VP. (The portion extended for the VP connection test is omitted.) As described above, backup VPs, VPII to VP15, are set for each VP.

The bandwidth of the standby VPs is not secured for each of VPII to VP15, but for VP11 to VP], all 5 11 to VP].

First, for each transmission path between path handling systems,
The total bandwidth of the active VPs accommodated in the transmission path is calculated. For example, in the transmission line at point A, this sum is V
The POI band + the VPO2 band + the VPO5 band. As the VP band of the protection system, the maximum value BM of the total of the working band Bl, B2, ・Bn determined for each transmission path is used.
Assign AX. (2) in the figure shows the case when there is a transmission path failure to the point. In this case VPOI, VPO2, VP
It is necessary to turn back O5, and VPI among the backup VPs
1, VPI 2, and VPI 5 are actually used. Therefore, in this case, all that is required as the loopback band is the Bl-VPOI band+VPO2 band+VPO5 band. However, as a backup band, BMA
It is sufficient to ensure X. It can be seen that even for failures other than point A, it is sufficient to secure BMAX as a backup band in the same way.

Next, the detailed configuration of the ass handling system will be explained with reference to FIG. On the upstream side of the station INF100,
OMD I (Opera) is used to load and unload OAM cells.
tion and management cell
Drop and In5erNon) 11 and VP1
RTA (R
There is an outing Tag Adder) 21.

On the downstream side of the intra-office INF 100, there is an RTD (Routing Tag Delete) for removing a routing tag.
r) OM old 12 that loads and unloads 31 and 0^M cells
There is. OMDI) 1.12 of the intra-office INF 100 is connected by a loopback line 300. Branch/insertion circuit 2
00.201 has OMD I 13.15 on both sides, and D-RTA (Routi ng T
ag Adder with cell Drop)2
2 and I-RTD (RoutingTag Delete
The RTA 22 performs VPI conversion and adds a routing tag to the cell on the transmission path, and then, according to the instruction of the routing tag, outputs the cell to the I-RTD 32 or sends the cell to the ATMSW (ATM 5w1tch). 50 or both 1-RTD 32 and ATMSW 50. Here, the VPI of the maximum two cells output from the D-RTA 22 is
22. ATMSW5
The cell output on the route to ATMSlf performs switching according to the routing tag attached to the cell.
50 and guided to the desired intra-office INF 100. On the other hand, in I-RTD32, from ATMSW50 to I-
The cells coming up to the RTD 32 are inserted into the cell stream from the D-RTA 22. At this time, as mentioned above, D-R
TA22's cell stream is preferentially transmitted without delay and
The cell stream from the Sll 150 is inserted into an empty time slot of the cell stream from the D-RTA 22. After this insert operation, the routing tag is removed and the OMD
The cells are sent out onto the transmission line via I15. A loopback line 400 connects the OMD I 13.15 of the bidirectional drop/add circuits. The bottom right of the figure is OMD
The block configuration of I is shown. D is a cell branching circuit which branches a cell when it meets a certain predetermined condition (for example, a physical layer OAM cell).

2 is a cell insertion circuit that seizes an empty time slot and inserts a cell into the transmission path.

Next, details of the VP link setting method for each path handling system will be explained. FIG. 13 shows a VP link setting method for the path handling system that is the starting point for the active VP. The active VP enters from the ATMSW50, goes around the ring, and then ends at the D-RTA22. In this D-RTA 22, user cells among the cells that have gone around the ring are discarded, and OAM cells undergo header conversion and are output.

Although it is possible to guide this user cell to another location using the ATMSW 50 without discarding it, the length of the VP connection becomes long, so this is considered undesirable for boat fishing. OAM cells that are not discarded by D-RTA22 are
is guided to the desired output port of ATMS150, and its V
It is guided to INF, which is the end of the road on the entry side of P, and is dropped there. Further, the backup VP is set on the ring opposite to the ring in which the active VP is set. Here, in order to facilitate switching in the event of a failure in OMDs 113 to 16, as described above, the active VP (7) R
After TA21 day and preliminary system (7) VIV) D-RTA
It is assumed that the VPI of the active VP before D-RTA 23 and after D-RTA 23 of the backup VP are the same.

In addition, for OAM cell routing, the VPI of the active VP after DRTA22 and the VPI before passing through RTA21.
Assume that PI is also the same.

FIG. 14 shows a method for setting VP links in the path handling system for relay points and destinations. A working VP1 and a backup VP are set on both rings 100 and 101. Here, in order to facilitate switching in the event of a failure in OMDs 113 to 16, the D-R
D-RTA23 of the VP bus in the backup system before passing through TA22
After passing, and D-RTA22 of the active VP bus
The VPIs after passing and before the D-PTA 23 of the standby VP bus are set to be the same. Furthermore, in the D-RTA 22 through which the active VP passes, the header conversion table HTT
(Header Transformation
ble) is set. In addition, in the original final point path handling system, the user's cell is
The header translation table HTT is set so that the ATM layer OAM cells are dropped to the NF, but are also transferred to the next stage path handling system.

The operation of the path handling system when avoiding obstacles is explained in Chapter 15.
As shown in the figure. The OMD 116 that is outputting cells toward the failure point forcibly lubebacks all cells. Further, the OMD 113, which has been receiving cells from the failure point, unconditionally drops all cells.

User cells on the transmission path at the time of system switching are discarded.

Physical layer OAM to confirm recovery again toward the failure point in order to automatically recover when a failure occurs or recovers.
If cells are sent and returned after being looped back, it may be decided to stop the forced loopback in OMD I 18. Once the worker has repaired the failure point, the recovery confirmation physical layer OAM cell will return stably. After that, the two pass handling systems sandwiching the point where the failure occurred were OM.
We used D113.16 to discuss some kind of protocol, and first stopped the unconditional cell drop at OMD 113, which was receiving and receiving cells from the point where the failure occurred, and both parties took that action. After checking the O that was outputting the cell to the point where the failure occurred,
Stop unconditional cell dropping in MD 116. The system then returns to normal operation. At this time, the cells on the standby VP are discarded. For this reason, the OMD I 13, which is receiving cells from the point where the failure occurred, sends a 0^M cell that declares the deactivation of the backup VP connection. Each D-RTA
recognizes this OAM cell and starts discarding the cell that is input and carries the VPI of the protection VP. It is assumed that the OAM cell that declares VP connection deactivation is discarded by the D-RTA 22, which discards the VP cell of both Sakae and Sakae. As a result, it is possible to realize that 200 AM cells can be removed from the ring by going around the ring exactly one time.

However, to confirm that the VP connection has been properly set, either confirm it with the VP connection connection test as described above, or depending on the application, make sure that VPI conversion is performed correctly at the level of each path handling system before the VP connection connection test. will need to be confirmed. Such a test method at the level of an individual path handling system (referred to as a VP link connection test) will be explained in the case of a copy connection using FIG. 16. ATM layer OAM with VPI that identifies the copy connection you want to test
Insert the cell before D-RTA22 where copying takes place, after 1-RTD32 and OMD 1 after RTD31.
Recognizes that it will be dropped from 12 onwards.

OMD I 11-15, RTA21 in Figure 13
5 RTD 31 and ATM SW 5[1 can be constructed using conventionally known technology, and are not characteristic parts of the present invention, so they will not be described in further detail. On the other hand, D-1? TA22 and I-
Since the RTD 32 has a characteristic configuration of the present invention, it will be described in more detail. First, according to Figure 17, D-1?
In the figure, SP-]?AM411 is a RAM having one write port and two read ports, each of which can operate asynchronously.

A cell input from OMD I in, for example, 8-bit parallel form is converted into 32-bit parallel form by 8-323P412, passed through two 32-bit length registers, and then written into 3P-RAM 411. Here, input of an empty cell is recognized by referring to the header pattern when the two 32-bit registers 413 contain the header portions of the cell. Empty cells are 3P-RAM
It is not written to 411.

After that, the cell's VP accumulated from the 3P-RAM 411
(old VPI) is read and the header conversion table HT
T is referenced using the read VPI as a key, and the additional information (routing tag, etc.) required for conversion to the internal cell format (called SW nacelle) and the VPI newly attached to that cell (new vp+> are read). CC is output and stored in the rewrite information register 414.
SW nacelle format conversion is performed from the cell format determined by ITT. Specifically, the additional information for creating the SIF cell format and the new VPI are rewritten from the information register 414, and the other parts (excluding the last octet of the SW nacelle) are read from the 3P-RAM 414, respectively. The SW nacelle is created by selecting.

The S cell read out from the 3P-RAM 411 is )H
A 2×2 crossbar switch 415, which is written for each VP in the TT and is controlled by a flag indicating whether to branch or pass a cell belonging to that VP, selects a branch output port or a passing output port. will be forwarded to. The branch output port is connected to the input of the ATMSW, and the pass output port is connected to the I-PTD. At this time, empty cells generated by empty cell generation are output from output ports that have no cells to output.

The SW nacelle output from the 2×2 crossbar switch 415 is converted to 8-bit parallel data by a 32-8P8416 independently at two output ports, and then a parity bit is added to the final octet of the SW cell and output.

Here, since format conversion is performed from the CCITT format cell to the SW nacelle, the frequency of the data CK at the output port is higher than that of the pig CK at the input port. Therefore, it is necessary to change the clock. The asynchronous operation of the 3P-RAM 411 is used for clock switching here. It is assumed that operation is performed using data CKI (input port clock) up to the write port, and thereafter using data CK2.

The address space of the 3P-RAM 411 is divided into blocks each large enough to write one cell. A valid cell is written to one of the blocks held by the W address creation 4Jg to which no cell has been written. The W address creation 41g creates an address to write a cell from the selected block number, and selects gP-1? A.M.
Give to 411. Further, the number of the block into which the cell has been written is passed to the R address creation 417 by handshaking.

When the R address creation 417 receives a block number from the W address creation 418, it first reads the old VPI from that block for HTT access and transfers it to the VPI register 419. At the same time, the PT field is read and transferred to the PT register 420. This is because in the case of an ATM layer OAM cell, the D-RT^ needs to perform a different operation than a user cell.

When the additional information from the HTT and the new VPI have been transferred to the rewrite information register, the R address creation 417 reads the cell from the 3P-RAM 411 and selects the cell from the selector 421.
, and performs self-automatic conversion.

The number of the block from which the cell has been read is passed to the W address generation 418 by handshaking as a block in which no cell has been written.

The HTT is composed of a selector for arbitrating HTT access between a processor that manages data in the HTT and a D-RTA, and a RAM.

An example of HTT addressumab is shown in Figure 18.

An entry of 32 bits x 2 words is allocated to the HTT for each old VPI. ) D-R for ITT
A control flag for controlling the operation of the TA and a field for holding the new VPI for passing cells are set. The meaning of the information written in each field in an entry is summarized below. a) Entry valid flag 511 (1 bit) Indicates that valid information is set in the entry. It is assumed that an input cell carrying the VPI of an entry invalidated by this flag is discarded by D-RTA.

b) Routing tag (for branch route) 512 (30
(bit) The routing tag attached to the SW cell that is branched and goes to the ATMSW is held.

C) New VPI (for branching) 528 (12 bits) The new VPI attached to the 8w cell that is branched and goes to TMS4 is held.

d) New VPI (for passing) 529 (12 bits)
The new VP1 attached to the 81 cells passing through and heading to the I-RTD is shown.

e) Backup system/active system flag 521 (1 bit) Its V
Indicates whether the PI is reserved for the backup system or for the active system. D-1? There is a standby system activation flag in the TA, which holds the status of whether or not the standby system is currently activated to avoid failures. If the backup system activation flag indicates that the backup system is activated, the input cell carrying the VPI that this flag indicates is reserved for the backup system will be formatted as usual. It is output after the conversion operation. If the standby system activation flag indicates that the standby system is not activated, the input cell carrying the VPI indicated by this flag as being reserved for the standby system is discarded.

f) Pass 0N10FF flag 522 (1 bit) This is a flag indicating whether or not to pass the input cell carrying the VPI. If the cell is not allowed to pass, an empty cell is output from the pass output port.

g) Branch 0N10FF flag 523 (1 bit) This is a flag indicating whether or not to branch the input cell that carries the VPI. If it is determined that the cell is not allowed to pass, an empty cell is output from the branch output port. When the pass 0N10FF flag is set to pass and the branch 0N10FF flag is set to branch, a copy of the cell is created with a 2x2 crossbar switch. However, here, in the case of a VP connection test OAM cell, the original flag in the information field is referenced,
Those with that flag set perform both branching and passing. At 22, the original flag of the OAM cell to be branched off is removed. (The original flag is set where the 0^M cell is input from the originating path handling system.) The cell for which the original flag is set only performs a passing operation and does not perform a branching operation. This prevents the OAM cells that have been looped back at the VP connection termination point from being copied. In addition, if the pass 0N10FF flag is set not to pass and the branch 0N10FF flag is set not to branch, the input user cell carrying that VPI will be discarded, but the ATM layer OAM cell will be discarded. It is passed through format conversion (
D-RTD).

h) Access node/transit node flag 524
(1 bit) If the VPI is reserved in the active system, the access node (the path handling system that is the starting point of the VP)
User cells/ATM layer 0^M cells are handled differently depending on whether they are a transit node (a VP relay point or a destination path handling system).

Specifically, in the case of transit nodes, these two types of cells undergo format conversion equally (transit 0N10
FF, if necessary according to branch 0N10FF flag)
Output. At the access node, the user cell is discarded and the ATM layer OAM cell undergoes VPI translation and is routed tagged and branched to ATMS III.

i) User cell valid/invalid flag 525 (1 bit)
If the VPI is for the active system, the VP link on the ring starting from d-RTA may or may not actually have capacity reserved for user cells. If the capacity for user cells is not secured, the D-RTA cannot output user cells. This flag is a flag indicating whether or not user cell capacity is secured in the VP link starting from D-RTA for that VP. If this flag indicates that the user cell capacity is not reserved, the user cell is discarded and
Only ATM layer 0^M cells are passed through after format conversion.

j) Backup System Activation Flag 526 In order to activate the standby system, a special VPI will be used for OAM cells to be sent to the ring when a failure occurs.

This flag indicates that the VPI is reserved for the standby activation OAM cell. When the D-RTA's standby system activation flag indicates that the standby system is activated, cells input carrying this VPI are discarded.

k) Backup System Deactivation Flag 527 In order to deactivate the backup system, a special VPI will be used for OAM cells to be sent to the ring upon recovery from a failure. This flag indicates whether the VPI or the standby system is deactivated.
Indicates that it is reserved for AM cells. When the protection system activation flag held in the D-RT indicates that the protection system is inactivated, the cells input carrying this VPI are discarded.

FIG. 19 shows the configuration of the I-RTD.

A passing input port 611 is an input port into which the cell flow from D-RTA is input, and an insertion input port 12 is an ATM port.
This is an input port into which the cell flow from SW is input.

The insertion input port 12 is provided with a dual buffer 621. If this dual buffer 821 holds a cell, it captures the empty cell input from the passing input port 11 and replaces it with that cell. Flow control is applied between the dual buffer 621 and ATMSIF to prevent cell discard from occurring.

After the cell stream from the insertion input port 12 is inserted, S
The parity bit, the last octet of the W cell, is checked and cells in which a parity error is found are discarded.

2P-1? The address space of the AMI 322 is divided into blocks in which one cell can be stored. Input control 23
receives a block with no cells written from the output control, and if a cell is input, writes a cell to the empty block. On the other hand, the output control 624 receives from the input control 623 the number of the block in which the cell is written, and reads the cell from that block. Conversion to CCITT self-format is performed by skipping unnecessary octets when reading cells from 2P-RAM 622. The number of the block from which the cell has been read is passed to the input control 623 as an empty block number.

It is assumed that when no cells to be output are stored in the 2P-RAM 622, empty cells are output from the output boat.

2P-1 of SW cell? When writing to the AM822, the parity of the last octet of the SW cell is checked at the same time. Cells in which a parity error is found are discarded. This can be done by input control 623 not passing to output control 624 the number of the block into which the cell in which the parity error was found was written, but instead overwriting that block with the next input cell.

An example address map of the RTD-RAM 631 is shown in FIG.

The ring system with a copy function of the present invention can also be used in combination with an ATM switch to configure an ATM switch system with a copy function, as shown in FIG. 711 is an ATM switch without a copy function. 721 to 724 are input INF, 731 to
734 is an output INF. These three functions are functions commonly used to configure an ATM switch system. Here, a plurality of branch/insert units 701 to 704 and branch/insert unit 701 having a copy function described in the present invention are further illustrated.
A transmission line 777 on the ring connecting the lines 704 to 704 is provided as shown in the figure. Then, the branch/insertion units 701 to 704 and the output IN
Combine F731-734. Assume that cells to be copied and output come in from a plurality of input 1NPs 721 to 724. This cell is transferred by the ATM switch to one of the output INFs 731-734 to be copied. At this point, cells have not yet been copied. Then,
This cell is connected from the output INFs 731 to 734 to the branch/insertion unit 70 connected to the output INFs 731 to 734.
1 to 704, and further onto a ring transmission line 777. While this cell goes around the ring transmission path 777, it
~Create one copy if necessary in 704 and output INF
It is output to 731-734. In this way, cells can be copied and transferred to multiple desired output ports.

[Effects of the Invention] According to the present invention, since a VP for use in the event of a transmission path failure is set in advance, in the event of a transmission path failure, the process can be quickly performed simply by physically looping back using the path handling system adjacent to the failure point. Can deal with transmission path failures.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example of a path handling system configuration according to the present invention, Fig. 2 is a block diagram showing a conventional example of a TM link system configuration, and Fig. 3 is a block diagram showing a conventional example of a TM link system configuration. Figure 4 is a diagram for explaining the countermeasure method, Figure 4 is a diagram showing VP replacement required when a transmission line failure occurs, Figure 5 is a diagram showing a branch connection for one response, Figure 6 is a TM link system according to the present invention. 7 is a diagram for explaining the VP setting method when connecting point-to-point, FIG. 8 is a diagram for explaining the YP setting method for 1-handling branch connection, and FIG. 9 is a diagram showing the overall configuration of . 1 is a diagram for explaining a VP connection connection test when connecting a branch for response 1, Figure 10 is a diagram showing a method for detecting a transmission line fault, and Figure 11 is a diagram for explaining a method of handling the path handling system after detecting a transmission line fault. FIG. 12 is a diagram for explaining the bandwidth allocation method for the protection VP, and FIG. 13 is a diagram showing the VP link setting method in the path handling system that is the starting point for the active VP.
Fig. 14 is a diagram showing the VP link setting method in the path handling system for relay points and destinations, Fig. 15 is a diagram showing the operation of the path handling system during failure avoidance, and Fig. 1
Figure 6 is a diagram for explaining the VP link connection test, Figure 17
The figure shows the configuration diagram of D-RT^, Figure 18 shows an example address map of BTT (header translation table), Figure 19 shows the configuration diagram of 1-RTD, and Figure 20 shows an example address map of RTD-RAM. FIG. 21 is a diagram showing a configuration example of an ATM switch equipped with a copy function to which the present invention is applied. 11-16...OMDI (Operation xn
d Mxncgemenj ceDrop and
In5ertion)21-RT^(ROL口ng
Tag Axd+! er) 22, 23 ・D-
RT ＾(Routing Tag Adder
with celDIop) 31=RTD(Routing Tag Dele
jer) 32.3L= I-RTD (Rou mouth ng
Tag Delejer vijhcell I
n5ertion) 50...ATMSW (^TM 5w1tch)20
0.201... Branch/insertion section 300.400... Loopback line 211.215... Path handling system 221
．． 225...Points with equal VPI 311.31n
... Branching/inserting section 321.32n for the clockwise ring transmission path... Branching/inserting section 331.33n for the counterclockwise ring transmission path -'ATM switch 34n ... Intra-office INF ~35n ・・Loop back line

Claims (1)

[Claims] Consisting of a plurality of path handling systems and clockwise and counterclockwise ring-shaped transmission lines that connect the above-mentioned path handling systems, the above-mentioned path handling system consists of clockwise ring-shaped transmission lines. A branch/insert section connected to interpose the branch/insert section, an intervening branch/insert section connected to the counterclockwise ring-shaped transmission path, an ATM switch section connected to both of the branch/insert sections, and the ATM switch. In an ATM link system consisting of an in-office INF section connected to the INF section, and a loopback line that connects both drop/add sections, when setting a working VP between path handling systems, a backup VP must also be set. The above-mentioned active VP starts from the path handling system, which is the starting point where you originally want to set the VP, passes through the path handling system, which is the destination point where you originally want to set the VP, and then returns to the path handling system, which is the starting point. The backup VP is set to circulate on a ring-shaped transmission path facing in the opposite direction to the ring-shaped transmission path in which the working VP is accommodated, and The VPI of the active VP and the backup VP are set to be the same at the same location, and in the path handling system that is the original destination, the user cells carried through the active VP are By branching and relaying only the OAM cells, the OAM cells flowing from the path handling system at the starting point where you want to set the VP, through that VP, go around and connect to the path at the starting point where you want to set the VP. It is possible to return to the handling system, and the VP connection connection is confirmed when the OAM cell goes around normally. When a failure occurs in one part of the ring-shaped transmission path, both sides adjacent to the failure point are The path handling system uses the above-mentioned loopback line to control all the cell flows heading towards the failure point to loop back to the ring-shaped transmission line in the opposite direction, and on the ring-shaped transmission line in the opposite direction, the above-mentioned backup system Using VP, the cell flow is carried to the path handling system on the opposite side of the failure point, and is placed in the path handling system on the opposite side of the failure point and looped back again.
Transmission path failures are dealt with by transferring the cell flow carried by P to the working VP again, and the bandwidth reservation method for the protection VP is based on ring-shaped transmission in which the working VP is accommodated. The sum of the bandwidths of all active VPs accommodated in the path is calculated for each path handling system, and the maximum value is shared on the ring-shaped transmission path in the opposite direction where the protection VP is accommodated. In addition, in addition to the three basic functions of drop, add, and relay, the drop and add section has the following functions: An ATM link system characterized in that it is equipped with a function of branching to the ATM switch while relaying cells along an outgoing transmission path, thereby making it possible to perform a one-to-N branch connection as a whole ATM link system.