JPH10224389A - Multiring communication device and its node device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To omit the management of a data base, a path selection tool, etc., and to attain an economical communication network by sorting the grades of paths according to the reliability and quality of each path and setting a specific path in each ring network for both paths of high and low grades respectively. SOLUTION: The node devices 11 to 19 are connected together in the ring networks 21 and 22, and the node devices 13 and 24 are connected to each other as the bridge nodes. The paths to be set are sorted in classes A to C according to the reliability and quality of each path. Then two paths running opposite to each other are set in each ring network for the path of class A of the highest grade, and a non-short break switching device 31 placed at a receiving terminal selects with no short break one of paths passing through two routes that has higher quality. Meanwhile, the paths running in the single direction are set in each ring network for the paths of classes B and C respectively.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for a backbone network using an ultra-high-speed transmission line. In particular, the present invention relates to path setting in a multi-ring network in which a plurality of ring networks are connected to each other, and collection of information on the node arrangement and node status.

[0002] The present invention relates to an SDH (Synchronous Digital Hie).
rarchy) Suitable for use on nets.

[0003]

2. Description of the Related Art In recent years, ultra-high-speed communication networks utilizing the broadband characteristics of optical fibers have been introduced. Especially for backbone network links, as shown in Reference 1,
Transmission systems with a capacity of 10 Gbit / s have been introduced. Reference 1: Y. Kobayashi, Y. Sato, K. Aida, K. Hagimoto an
d K. Nakagawa, "SDH-Based 10 Gbit / s Opotical Transm
ission Systems ", Proc.IEEE GLOBECOM '94 (SanFrancis
co, CA), pp. 1166-1170, 1994 As a signal processing method in a network node, an asynchronous transfer mode (ATM) supporting various services has been recommended by the ITU-T and the ATM forum. In the ATM layer located between the physical layer and the application layer, signals are processed in units of cells. However, there is a problem in performing all the processing of signals on the backbone network link exceeding 10 Gbit / s in units of cells. The number of virtual paths (VPs) that must be processed in each node device is as many as thousands as shown in Reference 2, which causes an increase in the size of the node device and complicated network management. Reference 2: S. Matuoka, "Classified Path Restoration Sc
heme with Hitless Protection Switching for Large-C
apacity Trunk Transmission Networks ", IEEE GLOBECO
M '95, p.941 From such a viewpoint, the inventors of the present application conclude that although signal processing in the ATM layer is used for service node devices, various functions of the backbone network node device, path setting, failure recovery, etc. It is assumed that the process is performed in units of large bundles in the physical layer. The capacity of this large bundle path varies. By processing in large bundles in this way,
Path network management can be simplified. In addition, time division multiplexing (Time Division Multiplexing) in the physical layer is used for multiplexing.
xing: TDM) may be used. In this specification, mainly the Synchronous Digital Hierarchy (Synchronous D
igital Hierarchy (SDH).

On the other hand, as shown in Reference 3, a network having a very large capacity link is required to have high reliability and survivability. In ultra-high-speed networks, a single fiber failure can damage thousands and tens of thousands of users. Reference 3: T.-H.Wu, "Fiber Network Service Survivabil
ity ", Artech House, Boston and London The self-healing function has been studied and introduced. Self-healing is a fast recovery function from network failure, and the best-known example is the path S with switch or line switch
An ONET (Synch-ronous Optical Network) ring network. A self-healing network is advantageous in its simple device configuration and high reliability. In order to construct a backbone network, a multi-ring that combines a plurality of rings is promising because of delay problems.
However, a multi-ring network having a self-healing function has not been realized, and a path setting function such as routing and slot assignment cannot be said to be complete.

[0005] Network monitoring and control will now be described. For a single ring network, TMN (Telecommunication Management Network)
Models have been standardized. FIG. 24 shows the management network architecture. In this architecture, a network element NE (Networ
k Element) is a message converter module MD (Medi
ation Device or MCM: Message Converter Modu
le) via the operating system OpS, which is connected to the packet transfer network DCN.
Is connected. The figure also shows a workstation WS for operating the operating system OpS. Each of the network elements NE includes a control communication unit, performs control communication with the operation system OpS via the message conversion module MD and the packet network DCN, and delivers monitoring control information.

[0006] However, as the transmission system and the network become more sophisticated, the model shown in FIG.
The cost of the pS system, especially the development cost of software, has increased, which has raised the cost of the network. Further, in a centralized control network such as the model shown in FIG. 24, a system down in the control node device leads to a complete network down.

[0007] In view of the above, many studies have been made on distributed control in which network control is performed in a distributed manner by each network node device. FIG. 25 shows a distributed management network architecture in a single ring network. In this architecture, each of the network elements NE is provided with a small-scale operating system OpS. Such distributed control is disclosed in, for example, Reference 4. Reference 4: I. Cidon, I. Gopal, M. Kaplan and S. Kutten, "
A Distributed ControlArchitecture of High-Speed Ne
tworks ", IEEE Transaction on Communications, Vol.4
3, No. 5, pp. 1950-1960, 1995 As shown in Reference 5, the distributed control network requires only a small-scale operating system deployed in each network element and requires several control node devices. It is more reliable with respect to node device failure than a centralized control network. Further, another control network is not required, and there is an advantage that the network database memory held by each node device can be reduced and high-speed control can be performed. Reference 5: AEBaratz, JPGray, PEGreen Jr., M.Jaf
fe and DPPozefsky, "Sna Networks of Small System
s ", IEEE Journal on Selected Areas in Communicatio
ns, Vol.SAC-3, pp.416-426, 1985 Future multimedia services are diverse, and it is considered that the required signal quality or reliability differs for each service type. Therefore, the function required for the backbone network is to operate and manage the path multiplexed for each service in accordance with various required quality and at low cost.

[0008]

However, in the conventional network technology, the quality and reliability of all operated paths are the same. Thus, the quality and reliability of the network were dragged by the highest demanding path, which made the entire network expensive. To deal with this problem, Reference 6 states that a line (VC:
A virtual circuit) approach is being considered. Reference 6: N. Yamanaka, E. Oki, F. Pitcho and H. Sato, "F
ullNet: a Flexible Multi-QoS ATM Network Based on a
Logically Configured VC-Network ", IEICE Trans.Com
mun., E78-B, No. 7, pp. 1016-1024, 1995 However, this proposal also required a high-quality path network, and lacked flexibility at the level of path operation.

[0009] In the future multimedia network, it is considered that flexibility is required also in the path layer. In other words, there is a path that requires high cost but does not want to drop one bit and a path that requires quality or reliability to be reduced but cost is required at the same time.

An object of the present invention is to solve such a problem and specifically realize the operation of paths classified according to self-healing reliability by a multi-ring architecture under a distributed control environment. .

[0011]

A multi-ring communication apparatus according to the present invention includes a plurality of ring networks in each of which a plurality of node devices are connected in a ring by a transmission line. In a multi-ring communication device in which the node devices are connected to each other as a bridge node, each node device performs control communication with another node device using a control communication channel to transmit information to and from another node device. Distributed path setting means for setting paths as a unit of the path. The distributed path setting means divides paths according to reliability and quality. Two reverse paths are set, and each ring network is used for low-grade paths. Respectively, characterized in that it comprises means for setting the one-way around the path.

The path setting is performed by using a token ring protocol configured on a data communication channel (DCC) in a section overhead (SOH) embedded in a signal, using other nodes necessary for determining a route for distributed control. It is preferable to realize communication with the device and secure a band.

That is, when the own node device is a transmitting node, the means for setting a path acquires a token circulating around the ring network to which the node device belongs and sends a packet for requesting a path setting in two opposite directions. Means for sending in the direction, and when the own node device is a bridge node,
Means for transferring a packet arriving in one direction around the same direction in the next ring network, and when the own node device is a receiving node, a packet received from two directions requests setting of a high-grade path. And a means for returning a response to the response in two directions opposite to each other, and returning a response to one of the packets received from the two directions in one direction when the packet received from the two directions requests the setting of a low-grade path. Good.

For the purpose of self-healing, each node device is preferably provided with a means for instantaneously selecting a high-quality path of a high-grade path in which a path having the node device as a receiving end is duplicated. . The path setting means may preferably include means for automatically relieving a relatively high-grade path among low-grade paths by rerouting when a failure occurs. The rescue unit may include a unit that loops back a token included in the control communication channel when a failure is detected in an adjacent link or node device in each node device.

That is, three classes of different grades are prepared according to the degree of reliability, and the reception with high reliability is classified into “Class A”, “Class B”, and “Class C”. For class A, the transmitting node device performs two branches to accommodate paths in two different routes, and performs non-instantaneous switching at the receiving end.
The path is relieved without losing bit information. This technology is described in Reference 7: Kawase et al., "Investigation of Non-stop Frame Switching Method in SDH Network", IEICE BI, Vol.J78-BI, No.12, pp.
764-772. Next, with respect to the highly reliable class B, the failure is recovered by resetting the path in the single ring in the same manner as the path setting. For class C, signal reconnection is not performed until equipment restoration is completed. Self-healing other than class B is almost the same as that shown in Reference 2.

Each ring network is connected to another ring network via two or more bridge node devices, and at least one node device of any one of the ring networks has a node device in the ring network and another ring network. Means for transmitting, in one direction, a node device information collection packet for collecting information relating to the arrangement and operation status of the node device, and terminating the node device information collection packet transmitted and returned by itself. Means for accumulating information collected by the packet, and writing the number and status of the own node device in the received node device information collection packet in each node device of each ring network, and With means to transfer to the bridge Each of the node devices used as the node devices includes, in addition to the transfer unit, a unit for temporarily storing a node device information collection packet received from one of two ring networks connected to each other by the bridge node device. Means for transferring the node device information collection packet stored in the means for temporarily storing when the transmission right to one of the ring networks is obtained to the other ring network, and another bridge node device without obtaining the transmission right Means for erasing the node device information collection packet stored in the temporarily storing means when receiving the node device information collection packet from the server, and terminating the node device information collection packet transferred and returned by itself. Means to write-protect the packet and return it to the original ring network It can be provided with.

Preferably, each of the node devices and the bridge node device is provided with a means for deleting the same node device information collection packet when it is received within a predetermined time. It is preferable that the transmitting node device further includes means for distributing the information collected by the node device information collection packet to at least the bridge node devices of the plurality of ring networks and, if necessary, each node device. It is preferable that each of the bridge node devices is provided with a unit for restricting a path setting via the bridge node device based on information distributed from the distribution unit.

By using a ring network configuration,
High reliability and survivability can be secured as a communication network, such as easy recovery at the time of failure recovery. Furthermore, by limiting the communication network to a ring type,
The route setting is limited to clockwise or counterclockwise, which simplifies the route setting and resetting functions when opening a path or restoring a path in the event of a failure, and the hardware scale and algorithm for route editing Can be reduced,
An economical communication network becomes possible. Also, by arranging a plurality of ring networks in a plane and connecting these ring networks with two or more node devices,
High scalability while ensuring high reliability, survivability, and economy for large-scale communication networks.

Techniques for providing a difference in path priority include:
There are those disclosed in JP-A-3-276937 and JP-A-3-217140. In the former technique, if a high-priority path has a failure, the quality of the high-priority path is ensured at the expense of a low-priority path without a failure. On the other hand, according to the present invention, paths are set with priorities set in advance. In the technique disclosed in Japanese Patent Application Laid-Open No. 3-276937, sharing switching is performed to recover from a failure, and the route is not duplicated like the class A path in the present invention. Further, in the technique disclosed in Japanese Patent Application Laid-Open No. 3-217140, a parent node device exists and performs centralized control.
On the other hand, the present invention performs distributed control. The technology disclosed in Japanese Patent Application Laid-Open No. 3-217140 is a technology relating to a packet network in which transmission is performed only when data occurs, and priority is not provided for a failure, and the parent device performs centralized control. This is a completely different technology from the present invention.

[0020]

FIG. 1 is a diagram showing a first embodiment of the present invention, in which a ring network 21 in which a plurality of node devices 11 to 15 are connected in a ring by a transmission line,
Similarly, the node devices 16 to 19 and 13 and 14 include a ring network 22 connected in a ring shape by a transmission line. The two ring networks 21 and 22 connect the node devices 13 and 14 in the respective networks to bridge nodes. Connected to each other. The ring network 22 is further connected to another ring network using the node devices 17 and 18 as bridge nodes. Each of the node devices 11 to 19 can perform control communication with another node device using the control communication channel, thereby setting a path that is a unit of information transmission with the other node device.

Here, high-speed TDM is used for data transmission, and for distributed operation for setting a path,
A token ring protocol configured on a DCC in the section overhead or on a channel in another signal frame by wavelength multiplexing is used. The reason for this is that the connection of the path of the relay network is relatively permanent, the frequency of connection change is small, and the connection delay does not substantially matter.

The paths to be set are classified into classes A, B, and C according to reliability and quality. Then, for the class A path having the highest grade, two paths that are opposite to each other are set in each ring network through which the path passes. FIG. 1 shows a path between the node devices 11 and 16. One (clockwise) path is connected via a node device 12 to a node device 13 which is a bridge node, dropped by an add / drop module ADM in the node device 13, passes through a switch SW, and passes through a ring network. Add 22 on the side
Drop module ADM to ring network 2
2 and connected to the node device 16 which is the receiving end. The other (counterclockwise) path is connected to the node device 14 which is a bridge node via the node device 15, and the ring network 2 is connected similarly to the node device 13.
2 and is connected to the node device 16 which is the receiving end via the node devices 19, 18 and 17. For a path passing through two routes, a hitless switching unit (31) at the receiving end selects a higher quality one without interruption. Although the instantaneous interruption switching device 31 is provided in each of the node devices 11 to 19, only the node device 16 is shown here.

With respect to the paths of the classes B and C, a path in one direction is set in each ring network through which the path passes. FIG. 1 shows a class B path between the node devices 19 and 15.

In such a multi-ring topology, when a full mesh connection of paths is configured, the number of working paths required for each ring is represented by the following equation.

[0025]

(Equation 1) Here, N is the number of node devices in the ring, n is the number of rings, j
Is the ring number counted from the left. In the equation (1), the first term is a closed path in the ring, the second term is a path passing through the j-th ring, and the third term is the number of paths exiting from the j-th ring to another ring. It is. Of course, it can be seen that the number of paths in the ring of the ring located at the center is the largest.

FIG. 2 is a diagram for explaining a recovery method when a link failure occurs. The path of class A is remedied by selecting a route without any failure in the instantaneous interruption switching device 31 at the receiving end. Class B paths are rescued by closed rerouting in the ring network 21.

FIG. 3 shows a frame structure of the token ring formed on the DCC. This frame is composed of a frame control FC (Frame Control) including a delimiter and a token.
), A receiving node D-ID, a transmitting node S-ID, a control field C indicating routing or response, a data field, and a frame check sequence FCS. The data field contains
Grade indicating the class of the path, c / cc indicating the direction in which the clockwise or counterclockwise packet flows, the capacity of the path to be connected, the area to be written by the relay node device between the transmission, reception or bridge, and the bridge or reception node device The identifier includes an area indicating whether or not it extends between rings.

The present invention is not limited to the token protocol shown in FIG. 3, but can be similarly implemented by the IBM token ring protocol or the Mitsubishi Electric-Loop1 protocol. Further, as long as the token protocol is used, the multi-token method or the early token method can be similarly implemented. A bidirectional two-fiber ring is assumed as each ring, but the same can be implemented with a four-fiber ring. Although the token circulates only in one direction, the packet can be transmitted in both directions.

FIG. 4 shows a configuration example of the node device. This node device includes a main signal processing unit 41, a path management unit 42, and a control communication unit 43. The main signal processing unit 41 includes an intra-office transmission unit 44, a low-speed frame signal termination processing unit 45, a path selection processing unit 46, Connection processing unit 47, demultiplexing processing unit 48,
Multiplexed frame signal termination processing unit 49 and inter-station transmission unit 5
0, and the path management unit 42 includes a path setting unit 51.

The intra-office transmission section 44 transmits and receives intra-office signals input from other devices. The low-speed frame signal termination processing unit 45 performs frame synchronization of the transmitted / received signal and demultiplexes the signal for each path. The path selection processing unit 46 performs duplexing of the class A path in the case of transmission and performs instantaneous interruption switching of the class A path in the case of reception. For classes B and C,
Go through one pass. The path connection processing unit 47 is a so-called switch unit, and rearranges a path from the inter-station transmission path and a path from the inside of the station. The demultiplexing processing unit 48 demultiplexes a path that is output between stations or that enters from between stations. The multiplexed frame signal termination processing unit 49 performs section overhead processing such as frame synchronization on the multiplexed signal. The inter-station transmission unit 50 transmits and receives the multiplexed signal. The control communication unit 43 performs control communication between the main signal processing unit 41 and the path management unit 51, extracts a portion defined for control communication from the section overhead, assembles a packet, and decomposes the packet to obtain a section. Insert into overhead bytes. Path setting unit 51
Sets a path according to the algorithm described below.

FIGS. 5 to 9 are diagrams showing the flow of path setting control in each node device. FIG. 5 shows the overall flow, FIG. 6 shows a transmission node, FIG. 7 shows a relay node, and FIG. As a bridge node, FIG. 9 shows each operation as a receiving node.

Each node device executes the control shown in FIG. 6 as a transmitting node when there is a path connection request. In other cases, the own node device receives the connection request packet from another node device as a relay node when it is necessary to relay the packet, and as a bridge node when it needs to operate as a bridge node, the own node device receives the connection request packet. , As receiving nodes, respectively, as shown in FIG.
Controls 8 and 9 are executed.

FIGS. 5 to 9 show the control of one node device. At the time of path setting, a plurality of related node devices are respectively connected to a transmission node, a relay node, a bridge node,
Act as a receiving node. These operations will be described by taking as an example a case where a path is set between two ring networks of rings #a and #b.

When a node device (transmitting node) having a path connection request grasps a token circulating in only one direction in the ring #a, the node device (counterclockwise and counterclockwise) transmits a connection request packet in which the grade and capacity of the path are written. In both directions to perform a double route search. If the relay node has enough bandwidth to transfer the corresponding path as it is in the packet transmission direction, the relay node stamps its own identifier (ID) in the relay node area of the packet and relays the packet. If the path cannot be passed to the next node device, write No Good (NG). The bridge node writes whether or not the path can be connected to the next ring #b in a packet, and transmits it as a response packet in the ring #a. The response packet that has returned to the transmitting node after making a round of the ring #a is terminated, and the token is released.
Release of the token may be done after terminating the packet,
The control may be performed after terminating the token, after transmitting several packets, or by multi-token. Here, for simplicity, a single token packet termination type will be described.

Since the packet is transferred clockwise and counterclockwise, the two node devices become bridge nodes in ring #a. These two bridge nodes grab the token of ring #b and transmit the packet to the receiving node. Here, the bridge node realizes the route diversity of the class A path by transmitting the packet received in the clockwise (left) direction from the ring #a in the clockwise (left) direction of the ring #b. The same applies to the routing in the ring #b.

After confirming that the path is connected, the receiving node of ring #b transmits a routing completion notification packet. The route through which the route information and the routing completion notification packet notified here pass are two routes through which the routing packet has passed in the class A path, and of the routes through which the two routing packets have passed in the class B and C paths. This is the route that has reached the receiving node earlier. That is, in the class B and C paths,
The route completion notification packet is transferred from the receiving node to the transmitting node by only one route. However, each bridge node transfers the routing completion notification packet to the next ring and, at the same time, transmits a response packet of the completion notification packet to another bridge node or receiving node in the ring by the opposite route. In the class B path, the relay node device on the opposite side of the working route in the ring writes the ID if the protection path in the ring can be secured by looking at the contents of the response packet, and NG if the protection path cannot be secured.
Write. The NG of the backup path reservation is transmitted from the bridge node or the receiving node that terminates the response packet.
It is later transferred to the transmitting node as backup path NG information.

The average delay time from when a path connection request is generated until connection is established depends on the frequency of connection requests. The average delay time can be expressed by the following equation, assuming that the connection request generation process is random.

[0038]

(Equation 2) Here, N is the number of node devices in the ring, τ _f is SDH (synchronous digital hierarchy) one frame transfer time, and d is 1
The number of DCC channel bits per frame, n is the number of rings, t is the propagation delay, L is the link distance, τ _p is the packet processing time per node device, and L _h is the packet overhead.

FIG. 10 shows the relationship between the frequency of the path connection request and the average connection delay time obtained by the equation (2). The horizontal axis indicates how many connection requests are required until the token returns after making one round of the ring. Here, the total number of node devices is 5
0, τ _f = 128 μs, d = 12 bytes, t = 5 ns /
m, L = 100 km for each link, τ _p = 1 ms,
L _h = 8 bytes. The connection delay decreases as the number of rings increases (the number of node devices in the ring decreases). This is because the probability that a certain node device grabs a token increases as the total number of tokens increases. If the total number of node devices is about 50,
It can be seen that the average connection delay is on the order of a few seconds.

Next, self-healing will be described.

A class A path is a highly reliable path that does not always lose any bit of information due to duplication in the transmitting node device, accommodation in two different routes, and non-instantaneous switching at the receiving end.

FIG. 11 shows an example of a configuration for instantaneous interruption switching. The class A path passes through two separate paths (path 0 and path 1) from the two flip-flops TTF at the transmitting end,
It is terminated by two flip-flops TTF at the receiving end. Here, the delay of path 0 is τ ₀ , and the delay of path 1 is τ ₁
And Two transmission line termination function circuits TTF (Tr
ansport Terminating Function) is output via a memory or other delay circuit DLY to the instantaneous interruption switch SW.
, And the amount of delay by the delay circuit DLY and the operation of the instantaneous interruption switch SW are controlled by the switching control circuit SWC. The switching control circuit SWC compares the two paths for each frame using the B3 byte of the VC frame, and selects a higher quality fume. Therefore, even in the case of a sudden failure, not only instantaneous interruption switching can be performed, but also in normal operation, the output error rate is improved over the error rate of the transmitted path. Assuming that the error rate of the working path is ε ¹ and the error rate of the protection path is ε ² , the output error rate ε ^out of the switch is expressed by the following equation.

[0043]

(Equation 3) Here, Q (ε) is B3 (BIP−
8) is the probability of detecting an error, and is expressed as follows.

[0044]

(Equation 4) P (ε) is a probability that the parity per rail of BIP-8 detects an error, and is expressed as follows.

[0045]

(Equation 5) Assuming that the path size is VC-4 (156 Mbit / s), the effect of improving the error rate is as shown in FIG. Assuming that the error rate of both paths is 10 ^-10 , the output error rate of the hitless switching circuit is about 10 ^-16 .

Next, self-healing for a class B path will be described. Class B is not required until instantaneous interruption switching, but is positioned in a medium reliability rank in which backup route switching is required in the event of a network failure.
While class A is a dedicated switch for standby system, class B is a shared switch for standby system.
hing). In the conventional sharing switching method, completely different methods and apparatuses are used at the time of path setting and at the time of failure recovery. On the other hand, in the present invention, the same method as the path setting is used at the time of failure, thereby realizing device sharing. Less than,
Parts different from the path setting will be described.

When a failure occurs due to a ring failure or the like, the node devices at both ends of the failed link detect the LOS and notify a section alarm. At this time, the two failed-end node devices loop back the token that was transferred only in one direction in normal operation on firmware or software, and also transfers the token to the ring in the opposite direction.
Thereby, even if there is a link failure, any node device can grab the token. What is important is that only the token included in the protocol (FC byte) is looped back, and the signal is saved at the path layer. When a failure occurs, all the class Bs in the failed link request the connection, not the random connection request generation process as described in the description of the path setting. The node device that has grasped the circulating token executes reconnection with respect to one failed path related to its own node device in the same manner as in the case of path setting, and releases the token. Next, the same applies to the node device that has acquired the token. In this way, the paths are sequentially reconnected one by one.

FIGS. 13 to 15 are diagrams for explaining the self-healing characteristics of the class B path. FIG. 13 shows that a failure has occurred in the j-th ring network including the node devices A, B, C, D and Z. FIG. 14 shows the position of the node device and the number of paths to be recovered when all the failed paths are of class B, and FIG. 15 shows how the token is transferred to each node device.

Here, assuming the worst case in the ring, it is assumed that a failure has occurred in the link between the node device A which is a bridge node and the adjacent node device Z. At this time,
The path affected by the fault includes an inter-ring path # 1 ((1) equation 1 second
Term), an inter-ring path # 2 (third term in equation (1)) connecting the j-th ring network to another ring network, and a path closed in the ring (first term in equation (1))
You could think so. J-1 for the ring-to-ring path # 1
The node devices A and D, which are the respective bridge nodes with the (j) -th and (j + 1) -th ring networks, must be shared and restored. In the figure, m is a sharing coefficient between the node devices A and D, and is represented by the following equation.

[0050]

(Equation 6) Since the node devices A and Z are faulty end nodes, the token is looped back by these node devices A and Z and transferred to each node device. Here, it is assumed that a processing time for restoring a path in a certain node device is τ _r1 , and a time for simply relaying to a neighboring node device is τ _r2 . Both are represented as follows.

[0051]

(Equation 7) The recovery characteristics of a path at any time are characterized by the number of token repetitions k as shown in the following table.

[0052]

[Table 1] Since the position of the token when the failure occurs cannot be specified, the recovery characteristics are averaged in the end-to-end one-way section of the token. If the recovery rate R (T) is defined as "the ratio of the recovered path at any time to all the failed paths", it is expressed as follows.

[0053]

(Equation 8) However,

[0054]

(Equation 9)

[0055]

(Equation 10) And

[0056]

[Equation 11] It is. Also,

[0057]

(Equation 12) It is.

FIG. 16 shows the recovery rate calculated using the above equation. Here, assuming that the total number of node devices is 50, the case of a single ring, two rings, and five rings will be described. Also,
Other parameters are the same as in FIG. In FIG.
A class A and a facility restoration class C in which no bit is lost are also shown. For the class B path, complete recovery is performed in 40 seconds or more for a single ring, about 20 seconds for two rings, and about 10 seconds for five rings.

An important point of this embodiment is that a path can be set without a network configuration information in a multi-ring configuration. Further, by modifying such a technique, it becomes possible to grasp information on the arrangement of the node devices.

A description will be given below of how to grasp the information on the arrangement of node devices and to recognize the state for operation and management of the distributed multi-ring network.

As a state recognition method in a single ring of centralized control, there are methods disclosed in JP-A-8-191318 and JP-A-7-58765. These methods require a parent node device that controls the network and collects information. On the other hand, as a state recognition method applicable to distributed control, there is a method disclosed in Japanese Patent Application Laid-Open No. H8-23200. In this method, a remote device connected to an arbitrary node device makes a control message having a pointer indicating a device number and an address value setting area make a round on a loop to grasp the state. This method is
In that control from an arbitrary node device is possible, it can be executed even under distributed control.

However, if this method is executed on a multi-ring as it is, two bridge nodes connecting any two rings independently transfer control messages to the next ring. Therefore, assuming that the number of rings is ^N , the number of control messages as a whole network is 2 ^N , causing a problem that the control messages are congested and the network performance is deteriorated.

An embodiment that solves such a problem and enables the state grasp and state recognition of the node device arrangement in a distributed multi-ring network in which two bridge node devices are connected between ring networks will be described.

FIG. 17 is a diagram showing a second embodiment of the present invention, in which six ring networks 61 to 6 are connected in a ring shape by a plurality of node devices, respectively.
6, ring networks 61 and 62 are connected via bridge node devices 71 and 72, ring networks 62 and 63 are connected via bridge node devices 73 and 74, and ring networks 61 and 64 are bridge node devices 75 , 76, the ring networks 62, 65 are connected via bridge node devices 77, 78, the ring networks 63, 66 are connected via bridge node devices 79, 80, and the ring networks 64, 65 The multi-ring communication devices are connected via bridge node devices 81 and 82, and ring networks 65 and 66 are connected via bridge node devices 83 and 84. In the ring networks 61 to 66, node devices other than the bridge node devices are also provided. Is omitted. FIG. 17 also shows an example of a transfer route of the node device information collection packet.

In this embodiment, the ring network 6
Connections among the node devices 1 to 66 are distributedly controlled, and information on the node device arrangement and the node device state of each ring is efficiently collected from any node device for the control. Here, a description will be given assuming that the transmission node device 91 in the ring network 61 collects information. The transmission of the node device information collection packet from the transmitting node device 91 is performed when the network is laid or when a new node device is added. The node device serving as the transmission node device 91 is:
A node device having an operator at the time of laying the network or a node device added at the time of operation may be used.

First, as in the first embodiment, the transmitting node device 91 transmits the section overhead (SO
H) The node device information collection packet is transmitted in one direction of the ring network 61 using the token protocol configured above. In the node device information collection packet, a signal for node device information collection is placed on the data field portion. Here, only one token exists in one ring network, and only the node device that has grasped the token can transmit the packet.

The node device information collection packet transmitted from the transmitting node device 91 goes around the ring network 61 on the SOH. Each node device sequentially stamps its own ID. If the node device has failed, NG is written. The transmission node device 61 accumulates the packet returned after making a round in the memory and terminates the packet.

Bridge pairs connecting the ring networks 61 and 62, that is, bridge node devices 71 and 72
Accumulates in the memory the node device information collection packet to pass to the next ring network 62. Of the two bridge node devices 71 and 72, only the bridge node device 71 that has obtained the first token transfers the token to the next ring network 62. The bridge node device 72 that has received the node device information collection packet transferred by another bridge node device 71 without holding the token deletes the packet in the memory.

The bridge node device 71 that has transferred the node device information collection packet accumulates and terminates the node device information collection packet that has returned after making a round, and returns it to the original ring network 61. At this time, the write prohibition information is written in the packet.

The node device that has already processed the packet
If the same packet is received again within the determined time, the packet is deleted. FIG. 17 shows that the packet is deleted by the bridge node device for simplicity, but this may be performed by any node device.

By repeating the transfer and return of the node device information collection packet between the ring networks, one node device information collection packet is returned from each of the ring networks 61 to 66, and the transmission node device 91 transmits all the node device information collection packets. Can be efficiently collected without duplication.

FIGS. 18 to 21 show the flow of processing for collecting information and recognizing information on node device arrangement and node device status by each node device, and FIG. 18 wants to collect information on node device arrangement and node device status. FIG. 19 is a processing flow of the relay node device, FIG. 20 is a processing flow of the bridge node device,
FIG. 21 shows a flow of the packet processing during the processing shown in FIG. These processes are performed by the path setting unit 51 and the control communication unit 43 in the configuration shown in FIG.

When the transmitting node device grasps the circulating token in its own ring, it transmits a node device information collection packet. This node device information collection packet has a data field portion of a packet used for a normal token ring protocol, and an area to be stamped by each node device is provided. The address of the destination node device is set to the transmission node device address. After transmitting the node device information collection packet, the transmitting node device enters a state of storing all received packets.
However, when the first packet receives a complete round,
Release the token.

When the node device other than the transmission node device receives the node device information collection packet, it operates as a relay node device. Then, first, it is checked whether or not the received packet is write-protected. If the write is prohibited, the data is transferred to the adjacent node device. If the packet is the same as the one already received, delete the packet. If it is the first reception, write its own number (ID) in the stamp area if its own node device is normal, and write No Good (NG) in the stamp area if any node in the node device has failed. I do. Here, the own number includes a ring number and a node device number. For example, 1
Four bits of the byte are used as a ring number, and another four bits are used as a node device number in the ring.

The bridge node device roughly performs three processes. The first is a process as a relay node device, the second is a node device information collection packet process as a bridge, and the third is a ring configuration information packet process. Here, the ring configuration information packet is a write-inhibited information notification packet. In the figure, the node device information collection packet and the ring configuration information packet are collectively shown as “X packet”.

The bridge node device first performs processing as a relay node device, and then enters a node device information collection packet processing mode. Upon receiving the node device information collection packet from the transmitting node device, the packet is stored in a memory so as to be passed to the next ring. Then, it enters a mode of waiting for the token of the next ring. When the token is secured, a node device information collection packet is transmitted. If the same node device information collection packet is received from another bridge before securing the token, the packet stored in its own memory is deleted immediately and the process enters the processing mode of the relay node device. As a result, only one node device information collection packet is transmitted from one bridge in the ring, so that congestion of control packets can be avoided. Next, the bridge node device enters a mode for receiving a packet transferred (filled with information) by itself, and when receiving the packet, disables writing and returns all packets to the original ring by the same route. When a packet that has been write-protected is received, the packet is returned to the ring that transferred the ring, and the process ends.

By these operations, the transmitting node device can collect and recognize information on all node device arrangements and node device states. After all of this information is accumulated (when the time becomes equal to or longer than T), the transmitting node device similarly transmits only the ring configuration information. The relay node device receives them, and deletes them if they have already been received. The ring configuration information packet is write-protected. The bridge node device similarly transfers and stores this information. When both the relay node device and the bridge node device receive and transfer the ring information packet, the node device information collection processing ends.

Based on the node device information collected in this way, an optimum path is set between any node devices.

FIG. 22 shows an example in which a path is set in the multi-ring communication apparatus shown in FIG. Here, the transmitting node device 92 in the ring network 61 (the “transmitting node device” here means a transmission source in path setting, and needs to be the same as the transmitting node device for collecting node device information described above. An example is shown in which a duplicated path is set between the receiving node device 93 in the ring network 63 and the receiving node device 93 in the ring network 63. The above-mentioned paper considers the case where the duplicated paths pass through the same ring network, and does not consider the case where there is a possibility that a path that passes through different ring networks may be set. For this reason, in the multi-ring communication device shown in FIG. 17, as shown in FIG. 22, the route of the path from the transmission node device 92 of the ring network 61 to the reception node device 93 of the ring network 63 is one of the ring networks. 62, the other being a ring network 64, 65, 66
Will be set via. For this reason, a large difference occurs in the delay time between the two routes. This is because the bridge node device where the packet first encounters always transfers the path to the next ring network. If there is a large difference in the delay time between routes, a large buffer is required to make the path uninterrupted.

In order to solve such a problem, there must be a restriction on which ring path each bridge node device forwards. That is, a set of ring networks through which the path from the ring network 61 to the ring network 63 passes must be uniquely determined, and each bridge node device must know the information.

Therefore, the transmission node device 91, which has transmitted the node device information collection packet, sends the node device information collection packet to the bridge node devices 71 to 84 of the ring networks 61 to 66 and, if necessary, to each node device. The collected information is distributed by a ring configuration information packet. Each of the bridge node devices 71 to 84 imposes restrictions on the path setting via the bridge node device based on the distributed information. In the example shown in FIG. 22, the bridge node device 75 is provided with a restriction on the path setting between the transmission node device 92 and the reception node device 93. Accordingly, when a class A path is set between the transmission node device 92 and the reception node device 93, a path via the ring network 12 can be set as a second route path, and the delay difference is small. A duplicated path can be obtained, and the buffer amount of the receiving node device 93 for performing instantaneous interruption switching can be reduced.

FIG. 23 shows the flow of the routing process by the bridge node device. First, the bridge node device performs routing in a restricted state. For example,
Regarding the class A path, it is assumed that all the passing routes are different but the passing ring networks are all the same, and the route connecting the ring networks is uniquely determined. This route determination is unique, and the calculation is simple because the number of nodes is the number of ring networks. Here, if it is possible to set a restricted path, the path routing is completed. In this case, instantaneous interruption is easy, and the most reliable path is obtained. If a restricted path cannot be set, the restriction is removed and routing is performed again. In this case, it is difficult to make the path uninterrupted, but a highly reliable path can be obtained. In addition, the buffer amount is reduced as a whole.

[0083]

As described above, according to the present invention,
Each node device is provided with control communication means for transmitting and receiving control signals to and from other node devices, and when a path is opened, the node device that has issued the path opening request selects a route to the end node device, Instead of using an individual operation system, by directly accessing each node device and checking the vacant state and setting sequentially, an individual operation system and a connection device for connecting the operation system and each node device that were conventionally required. Can be deleted. In addition, when setting a path route, the node device selects a route while grasping the path accommodation status of other node devices, thereby eliminating the need for a huge database management conventionally required for path accommodation design. In addition, a route selection tool based on the huge database becomes unnecessary, and an economical communication network can be realized.

Further, the route setting of the first route constituting the class A path and the second route dispersed in the different route is performed clockwise and counterclockwise in the ring using the above-described route setting algorithm. In the two node devices that are separately connected as the second route and the second route, and the connection between the rings, the route connected clockwise is a clockwise route, and the route connected counterclockwise is a counterclockwise route. To connect the first route and the second route without crossing each other. When a failure occurs on the class B path, the failure is avoided by using the above-described route setting algorithm. By restoring the path by re-establishing the path to be restored, instead of designing the path setting algorithm for path establishment and the failure recovery algorithm at the time of self-healing independently as in the past, the path By utilizing both for opening and for disaster recovery to constant algorithm, can reduce the load of the operating system, the economic can realize a communication network.

As described above, according to the present invention, a path having a plurality of transport functions can be set. According to the present invention, it is possible to set a path capable of transmitting information without loss of information even when a failure occurs in highly reliable communication information, and to perform economical information transmission for communication information that does not require reliability. it can. Also, a ring network that can ensure high reliability and survivability can be easily configured.

In the present invention, grasping and recognition of the arrangement and status of network node devices, which are necessarily required for automatic network operation, are realized under distributed control. As a result, simple and high-speed network management, node device state management, route diversity with a small delay difference, and avoidance of a plurality of control packets that may cause problems in distributed control can be achieved.

[Brief description of the drawings]

FIG. 1 is a diagram showing a first embodiment of the present invention.

FIG. 2 is a diagram for explaining a recovery method at the time of a link failure.

FIG. 3 is a diagram showing a frame configuration of a token ring.

FIG. 4 is a block diagram illustrating a configuration example of a node device.

FIG. 5 is a diagram showing an overall flow of path setting control in each node device.

FIG. 6 is a diagram showing a flow of control as a transmitting node.

FIG. 7 is a diagram showing a flow of control as a relay node.

FIG. 8 is a diagram showing a flow of control as a bridge node.

FIG. 9 is a diagram showing a flow of control as a receiving node.

FIG. 10 is a diagram illustrating the relationship between the frequency of path connection requests and the average connection delay time.

FIG. 11 is a diagram illustrating a configuration example for instantaneous interruption switching.

FIG. 12 is a diagram showing the effect of improving the error rate.

FIG. 13 shows j including node devices A, B, C, D, and Z
The figure which shows the state in which the failure occurred in the 2nd ring network.

FIG. 14 is a diagram showing the positions of node devices and the number of paths to be restored when all the failed paths are of class B;

FIG. 15 is a diagram showing how a token is transferred to each node device.

FIG. 16 is a diagram showing a recovery rate obtained by calculation.

FIG. 17 is a diagram showing a second embodiment of the present invention.

FIG. 18 is a diagram showing a processing flow of a transmission node device that wants to collect information on node device arrangement and node device status.

FIG. 19 is a diagram showing a processing flow of the relay node device.

FIG. 20 is a diagram showing a processing flow of a bridge node device.

FIG. 21 is a diagram showing a flow of a packet process during the process shown in FIG. 20;

FIG. 22 is a diagram showing an example of path setting.

FIG. 23 is a diagram showing a flow of a routing process by the bridge node device.

FIG. 24 is a diagram showing a conventional management network architecture based on the TMN model for a single ring network.

FIG. 25 is a diagram showing a conventional distributed management network architecture in a single ring network.

[Explanation of symbols]

 11-19 Node devices 21, 22 Ring network 31 Non-stop switching device 41 Main signal processing unit 42 Path management unit 43 Control communication unit 44 Local transmission unit 45 Low-speed frame signal termination processing unit 46 Path selection processing unit 47 Route connection processing unit 48 demultiplexing processing unit 49 multiplexed frame signal termination processing unit 50 inter-station transmission unit 51 path setting unit 61-66 ring network 71-84 bridge node device 91, 92 transmission node device 93 reception node device

Claims (18)

[Claims] A plurality of ring devices each including a plurality of node devices connected in a ring by a transmission line, wherein the plurality of ring networks are connected to each other by using any node device in each network as a bridge node. In the ring communication device, each node device includes a distributed path setting unit configured to perform a control communication with another node device using a control communication channel to set a path serving as a unit of information transmission with the other node device. This distributed path setting means classifies paths according to reliability and quality, and sets two paths that are opposite to each other in each of the ring networks through which the high-grade path passes. For paths, include means to set up a path around one direction in each ring network A multi-ring communication device, characterized in that: 2. The path setting means, when the own node device is a transmitting node, obtains a token circulating around the ring network to which the node device belongs and sends packets requesting path setting in opposite directions. Means for transmitting in two directions; means for transferring packets arriving in one direction around the same direction of the next ring network when the own node device is a bridge node; and two directions when the own node device is a receiving node. When a packet received from the GW requests a high-grade path setting, a response to the request is returned in two opposite directions, and a packet received from the two directions requests a low-grade path setting. 2. A multi-ring communication apparatus according to claim 1, further comprising means for returning a response to one of them in one direction. 3. The node device according to claim 1, wherein each node device is provided with means for selecting a high-quality path having a duplexed route having the node device as a receiving end, with a higher quality, without an instantaneous interruption. Multi-ring communication device. 4. The means for setting a path includes means for automatically relieving a relatively high-grade path among low-grade paths by rerouting when a failure occurs. 4. The multi-ring communication device according to any one of 3. 5. The method according to claim 4, wherein said rescue means includes means for looping back a token included in said control communication channel when a failure is detected in an adjacent link or node device in each node device. Multi-ring communication device. 6. Each ring network is connected to another ring network via two or more bridge nodes, and at least one node device of any one of the ring networks has a node in the ring network and another ring network. Means for transmitting a node information collection packet for collecting information on the arrangement of the device and its operation status in one direction of the ring network to which the node device belongs, and terminating the node information collection packet transmitted and returned by itself. Means for accumulating information collected by the packet, and in each node device of each ring network, write the number and status of the own node device in the received node information collection packet and transfer it to the next node device To use as a bridge node In addition to the transfer means, the bridge device temporarily stores a node information collection packet received from one of two ring networks connected to each other by the bridge node,
Means for transferring the node information collection packet stored in the temporarily storing means to the other ring network when the transmission right to the other of the two ring networks is obtained, and another bridge without obtaining the transmission right Means for erasing the node information collection packet stored in the temporarily storing means when receiving the node information collection packet from the node, and terminating the node information collection packet transferred and returned by itself. 2. The multi-ring communication device according to claim 1, further comprising means for writing back the packet and returning the packet to the original ring network. 7. A multi-ring, wherein each of a plurality of node devices includes a plurality of ring networks connected on a ring by a transmission line, and each ring network is connected to another ring network via two or more bridge nodes. In the communication device, at least one node device of the plurality of ring networks includes, in the node device, a node information collection packet for collecting information on an arrangement and an operation state of the node devices in the ring network and other ring networks. Means for transmitting in one direction to the ring network to which the node belongs, and means for terminating the node information collection packet transmitted and returned by itself and storing the information collected by the packet. The node device receives Means for writing the number and status of the own node device in the collected node information collection packet and transferring the same to the next node device. In addition to the transferring means, each of the node devices used as bridge nodes Means for temporarily storing node information collection packets received from one of the two ring networks to be connected;
Means for transferring the node information collection packet stored in the temporarily storing means to the other ring network when the transmission right to the other of the two ring networks is obtained, and another bridge without obtaining the transmission right Means for erasing the node information collection packet stored in the temporarily storing means when receiving the node information collection packet from the node, and terminating the node information collection packet transferred and returned by itself. Means for writing back the packet to the original ring network after write-protecting the packet. 8. The apparatus according to claim 7, wherein each of the node devices including the bridge node of each ring network is provided with a means for deleting the same node information collection packet when it is received within a predetermined time. Multi-ring communication device. 9. A node device for transmitting an information signal via a communication network connected to a plurality of ring networks, and performs control communication with another node device using a control communication channel to be a unit of information transmission. A distributed path setting unit for setting a logical path between the own node and the destination node; the distributed path setting unit acquires a token circling the ring network to which the node device belongs and requests path setting; Means for transmitting a packet in two directions in opposite directions to each other; and for a high-grade logical path for which reliability and quality are required, receiving a response packet to the packet requesting the path setting from the two directions, Means for setting two paths that are opposite to each other in each ring network that passes through the Means for setting a path in one direction in each ring network after receiving from one direction a response packet to a packet requesting the path setting for a low-grade logical path whose quality is not so required. Characteristic node device. 10. The distributed path setting means further classifies a low-grade logical path into two grades, and when a failure occurs in the relatively high-grade logical path, automatically re-routes the logical path by rerouting. The node device according to claim 9, further comprising: 11. A node device for transmitting an information signal via a communication network connecting a plurality of ring networks, wherein a logical path serving as a unit of information transmission between a start node and an end node is defined by a start node and an end node. And a distributed path setting means for performing control communication with another node device using a control communication channel in order to set a path between the two. A node device for determining whether or not a logical path can be relayed, adding information of the determination result to the packet, and relaying the packet to the next node device. 12. The node device according to claim 11, wherein the distributed path setting unit includes a unit that loops back a token included in the control communication channel when a failure is detected in an adjacent ring or a node device. . 13. A node device used as a bridge node connecting two ring networks for transmitting information signals to each other, wherein a logical path serving as a unit of information transmission between a start node and an end node is defined by a start node and an end point. A distributed path setting unit for performing control communication with another node device using a control communication channel in order to establish a setting between the node and the node, the distributed path setting unit arriving in one ring network in one direction; It is determined whether a logical path can be relayed around the same direction of the other ring network with respect to the packet for which the path setting is requested, and information of the determination result is added to the packet, and the packet is transmitted around the same direction. Means for transmitting to the one ring network a response packet to which information of the determination result is added. And a node device. 14. A node device for transmitting an information signal through a communication network connected to a plurality of ring networks, and performs control communication with another node device using a control communication channel to form a unit of information transmission. A distributed path setting means for setting a logical path between the start node and the own node; the distributed path setting means, when receiving a packet requesting the setting of a logical path ending at the own node from two directions, A response packet indicating whether or not the logical path can be set is transmitted in the opposite directions in the two directions for a high-grade logical path whose reliability and quality are required. For a low-grade logical path that is not so required, a means for transmitting in the opposite direction in one of the two directions is included. Device. 15. The node apparatus according to claim 14, further comprising means for selecting a path of higher quality from a high-grade logical path set between the node and the start node without any instantaneous interruption. 16. A node device for transmitting an information signal via a communication network in which a plurality of ring networks are connected to another ring network via two or more bridge nodes, respectively, comprising: a ring network to which the own node device belongs; Means for transmitting a node information collection packet for collecting information on the arrangement and operation status of node devices in the ring network in one direction of the ring network to which the own node device belongs, and node information transmitted and returned by itself Means for terminating the collected packet and storing information collected by the packet. 17. A node device for transmitting an information signal via a communication network in which a plurality of ring networks are connected to another ring network via two or more bridge nodes, respectively, when a node information collection packet is received. A node device comprising means for writing the number and status of the own node device in a packet and transferring the number and status to a next node device. 18. A node device used as a bridge node connecting two ring networks for transmitting an information signal to each other, upon receiving a node information collection packet, writes the number and status of the own node device in the packet, Means for transferring to the other node device, means for temporarily storing a node information collection packet received from one of the two ring networks connected to the own node device, and transmission right to the other of the two ring networks. Means for transferring the node information collection packet stored in the temporarily storing means to the other ring network, and when a node information collection packet from another bridge node is received without obtaining a transmission right. The node information collection packet stored in the temporarily storing means. And means for erasing the door, the node device according to claim that it has a means for returning to the original ring network by the packet write protected with terminating node information collecting packet which has returned to the transfer.