JP4720897B2 - Node device, optical network, and route calculation method - Google Patents

Description

  The present invention relates to a node device, an optical network, and a route calculation method, and more particularly, to a method for determining interface attributes of a node including an optical switch and an electrical switch.

  The network is composed of nodes having switching functions and links serving as data transmission paths connecting the nodes. FIG. 16 is a diagram illustrating a configuration example of an optical network. In the optical network of FIG. 16, there are nodes 110 to 115, between node 110 and node 111, between node 111 and node 112, between node 110 and node 113, between node 111 and node 114, and between node 112 and node 115. Bidirectional links are set between the node 113 and the node 114 and between the node 114 and the node 115, respectively. In addition, the node connection point at both ends of the link is an interface, which is indicated by a black circle in FIG. An identifier (ID) is assigned to each interface, and the interface can be uniquely specified by the ID.

  When an attempt is made to set 1 Gb / s and 10 Gb / s paths from the node 110 to the node 112 in the optical network as shown in FIG. 16, each interface can process in order to search for a route that satisfies the requested bandwidth. It is necessary to know the bandwidth. In the example of FIG. 16, the processable bandwidth of the nodes 110 and 112 to 115 is 10 Gb / s, and the node 111 is 1 Gb / s. For this reason, the processable bandwidth of the interface in the node 111 and the interface of the link connected to the node 111 is 1 Gb / s.

  That is, the bandwidth of the interfaces I1 and O1 of the node 110, the interfaces I1 to I3 and O1 to O3 of the node 111, the interfaces I1 and O1 of the node 112, the interfaces I1 and O1 of the node 114 is 1 Gb / s, and the interface of the node 110 I2, O2, interfaces I2, O2 of node 112, interfaces I1, I2, O1, O2 of node 113, interfaces I2, I3, O2, O3 of node 114, and interfaces I1, I2, O1, O2 of node 115 are 10 Gb / s.

  By grasping the bandwidth of each node, the shortest route to reach the node 112 from the node 110 via the node 111 is selected as the 1 Gb / s path, and the node 111 cannot pass through the 10 Gb / s path due to insufficient bandwidth. Therefore, it is possible to select a route from the detour node 110 to the node 112 via the node 113, the node 114, and the node 115. As described above, in routing, it is necessary to grasp an attribute of an interface such as a bandwidth.

  The attributes of the interface include a protocol, a switching capability, a metric, and the like in addition to the bandwidth.

  As shown in FIG. 17, in the conventional optical network in which the node 1 is composed of the electrical switch 10, the attribute of the interface has the attributes of the electrical switch 10 as they are, the interfaces I1 to I3, It was used as an attribute of O1 to O3.

  The reason is that in an electrical switch, the band and protocol that can be processed to convert an optical signal into an electrical signal are limited, and the interface attribute is uniquely determined by assigning the switch attribute as it is. In FIG. 17, since the node 1 is composed of the electrical switch 10, the switching capability, which is one of the attributes of each interface, is uniquely determined by PSC (Packet Switching Capability). PSC indicates that switching is performed in units of packets. Also, the control device 30 shown in FIG. 17 is connected to the electrical switch 10 and the control device of the adjacent node, and exchanges interface attributes with the adjacent node and controls the electrical switch 10.

  However, when the signal speed in the optical network becomes high, an electrical switch for dealing with a high-speed signal becomes expensive, and an optical switch that can be switched independently of the signal speed may be cheaper. Therefore, as shown in FIG. 18, the node 2 is configured by combining the electrical switch 10 and the optical switch 20, and an optical signal that does not require electrical processing is processed only by the optical switch 20. Node configurations that reduce the scale have been proposed.

  A node configured by the electrical switch 10 and the optical switch 20 as in the node 2 in FIG. 18 has a problem that the interface attribute of the node 2 is not uniquely determined depending on the connection state of the optical switch 20 in the node 2. For example, in a state where the interface I5 of the optical switch 20 and the port O4 are connected, the switching capability that is one of the attributes of the interface I5 is a PSC for processing a packet. However, when the interface I5 and the interface O5 are connected, the switching capability of the interface I5 is LSC (Lambda Switching Capability). LSC indicates that switching is performed in units of wavelengths.

  Conventionally, for the problem that the interface attributes differ depending on the switch connection status in the node as described above, for example, as disclosed in Patent Document 1, the switching capability of one interface is set to a plurality of values such as “PSC + LSC”. Shown in combination. Further, according to the technique described in Patent Document 1, the interface attribute is determined from the usage state of whether the device in the node is unused or in use.

  The technique described in this publication will be described with reference to FIGS. As an attribute of the interface, consider switching capability as an example. In FIG. 18, the interfaces I 1, I 2, O 1, and O 2 of the node 2 are uniquely determined as PSC from the attribute of the electrical switch 10. On the other hand, the interfaces I5, I6, O5, and O6 of the optical switch 20 can take either a PSC or an LSC depending on the connection state of the optical switch 20. Therefore, both PSC and LSC (denoted as “PSC + LSC”) are given to the interfaces I5, I6, O5, and O6.

  FIG. 19 shows a state in which the optical path P3 is set in the node 2 in FIG. The optical path P3 passes through the optical switch 20 and the electrical switch 10 in the node 2. Since all the ports I3 and O3 of the electrical switch 10 connected to the optical switch 20 are in use in the node 2 by this path P3, the attributes of the interfaces I6 and O6 cannot be PSC, only the LSC. To be judged.

  Note that the attribute of the interface of the node 2 is advertised to other nodes in the optical network via the control unit 30, and each of the other nodes also advertises the attribute of the interface of the own node.

  Further, in the conventional route calculation method, as shown in Non-Patent Document 1, the unreserved bandwidth of each link is managed. When setting a path, a route is calculated using only links whose unreserved bandwidth is equal to or greater than the bandwidth of the path to be set. Thereby, it is possible to set a path of the requested bandwidth.

Japanese Patent Laying-Open No. 2003-234761 (page 3-11, FIG. 1-13)

RFC 3630, Traffic Engineering (TE) Extensions to OSPF Version 2, Chapter 8, 2.5, September 2003

  FIG. 20 is a diagram illustrating another configuration example of the nodes in the optical network. In FIG. 20, the node 3 includes an electrical switch 40, optical switches 60 and 70, and a control unit 30 that controls these switches. The optical switch 60 and the optical switch 70 are connected via the electrical switch 40.

  When a plurality of optical switches are connected via electrical switches as in the node 3 shown in FIG. 20, even the interfaces described as “PSC + LSC” may not be connected due to the switching capability of the LSC. However, there is a problem that network resources cannot be used effectively.

  In FIG. 20, since the interfaces I1, I2, I9, I10, O1, O2, O9, and O10 of the optical switches 60 and 70 can be PSC or LSC depending on the connection state of the optical switch, the technology described in Patent Document 1 is used. Then, the switching capability is “PSC + LSC”, and the attributes of each interface stored in a database (not shown) of the control unit 30 are, for example, as shown in FIG. The combination is described in the interface that can take a plurality of values as in the switching capability column of FIG.

  A path that passes only through the same optical switch as in path P4 in FIG. 20 can be connected by “LSC”. However, when the interfaces I2 and O9 of optical switch 60 and optical switch 70 are connected as in path P5. Since it is necessary to pass through the electrical switch 40 in the node 3, the actual switching capability is “PSC”.

  Problems in using the node configuration shown in FIG. 20 for an optical network will be described with reference to FIG. The optical network in FIG. 21 includes nodes 501 to 505. The node configurations of the nodes 501, 503 to 505 are the same as those of the node 2 shown in FIG. 19, and the configuration of the node 502 is the same as that of the node 3 shown in FIG. .

  In an optical network, when the transmission distance of an optical signal becomes long, the power is lowered or the waveform is broken, so that the signal quality is deteriorated. Therefore, when data is transmitted over a long distance in an optical network, an optical signal is converted into an electric signal on the way to reproduce the signal quality, and then electricity is converted into an optical signal again for transmission. Since the regenerative repeater used at this time is an electrical device, and the one corresponding to the high-speed signal is expensive, route setting or the like is performed so as to reduce the number of uses of the regenerative repeater. The optical signal regenerative relay is also performed when an optical switch converts an optical signal into electricity and converts it again into an optical signal. In the optical network of FIG. 21, it is assumed that the number of hops that can be transmitted as an optical signal is two hops.

  A path P6 is set from the node 501 to the node 505. As the path P6, paths 501, 502, 504, and 505 are selected. Since both the interface I1 and the interface O9 of the node 502 are advertised to the entire network with the switching capability “PSC + LSC”, the interface I1 and the interface O9 seem to be connected by “LSC” as a whole network. Since the path P6 passes two hops between the node 501 and the node 502 and between the node 502 and the node 504 with an LSC, that is, an optical signal, it is determined that it is necessary to perform regenerative relay at the node 504. The interfaces I5 and O5 are used as “PSC”.

  However, the node 502 passes through the electrical switch 512, and regenerative relay is performed. Actually, since the regenerative relay is performed by the electric switch 512 of the node 502, the electric switch 514 is used wastefully at the node 504.

  As described above with reference to FIG. 20 and FIG. 21, in the technique described in Patent Document 1, when a plurality of optical switches are connected via an electrical switch, the interfaces described as “PSC + LSC” are connected to each other. Even in such a case, connection may not be possible due to the switching capability of the LSC, so that network resources cannot be used effectively.

  In the technique described in Patent Document 1, a combination of a plurality of values such as “PSC + LSC” is defined as a new attribute value. When there are many combinations of a plurality of values, it is necessary to determine attribute values for all the combinations, and the attribute values increase. For example, if there are 10 types of attribute values representing bandwidths, such as 10 Mb / s, 1 Gb / s, 10 Gb / s, etc., depending on the connection status of the switch, all of these may be taken. At this time, it is necessary to prepare 1024 attribute values for all combinations of 10 types of values, and there are many attribute values, which makes the process troublesome.

  Further, in the conventional route calculation method, as shown in Non-Patent Document 1, the unreserved bandwidth of each link is managed. When setting a path, a route is calculated using only links whose unreserved bandwidth is equal to or greater than the bandwidth of the path to be set. However, in the case of a node including an optical switch, it may not be possible to correctly determine whether each link can accommodate the requested path by this conventional method.

  An example in which path accommodation cannot be correctly determined will be described with reference to FIG. FIG. 23 is a diagram showing a configuration example of an optical network for explaining the problems of the conventional route calculation method. As shown in FIG. 23, nodes 1001 to 1004 are connected by links 901 to 903 to form an optical network. The node 1001 includes an optical switch 501, a client device 801, a control device 701, and a wavelength multiplexer / demultiplexer 600. The other nodes 1002 to 1004 are similarly composed of an optical switch, a client device, a control device, and a wavelength multiplexer / demultiplexer. Note that p0 to p15 indicate interfaces of the optical switches 501 to 504 of the nodes 1001 to 1004, respectively. The control device 701 manages the initial state of the unreserved bandwidth of the link 901 as 40 Gbps as an example.

  For example, when the control device 701 is requested to set a wavelength path whose bandwidth from the client 801 to the client 803 is 10 Gbps, the control device 701 searches for a route using only a link whose unreserved bandwidth is 10 Gbps or more. The path P21 is selected. The unreserved bandwidth of the link 901 after setting the path P21 is 30 Gbps. Similarly, when the wavelength paths P22 and P23 having a bandwidth of 10 Gbps are set, the unreserved bandwidth of the link 901 is 10 Gbps.

  Here, when a wavelength path with a bandwidth of 2.5 Gbps is requested from the client 801 to the client 802, the path P24 is set. In the conventional method of managing the unreserved bandwidth after setting the path P24, the unreserved bandwidth of 7.5 Gbps is advertised as the remaining allocation of 10 Gbps to 2.5 Gbps. However, in practice, even if an attempt is made to set a wavelength path of 7.5 Gbps or less for this link 901, the interfaces p8 to p11 connected to the link 901 of the optical switch 501 are already used, and there is no usable interface. Therefore, the path cannot be set.

  An object of the present invention is to provide a node device, an optical network, and a route calculation method that can correctly determine whether each link can accommodate a path for which a setting is requested.

  The node device according to the present invention is an interface related to an optical switch capable of exchanging input signals in units of wavelengths and a link that receives a path setting request and connects between the node device and the node device adjacent to the node device. Advertising means for calculating the unreserved bandwidth of the link using the number of unreserved interfaces and the maximum bandwidth that each of the interfaces can process and advertising to other node devices. .

  An optical network according to the present invention includes the above node device.

  The path calculation method according to the present invention relates to a step of receiving a path setting request, and a link connecting a node device including an optical switch capable of exchanging input signals in units of wavelengths and a node device adjacent to the node device. And calculating the unreserved bandwidth of the link by using the number of unreserved interfaces among the interfaces to be used and the maximum bandwidth that can be processed by each of the interfaces, and advertising to other node devices. To do.

  According to the present invention, the node device searches for a route in the own node connecting the input interface and the output interface of the own node, and for each searched route, an attribute satisfying an attribute of a device existing on the route is obtained. By setting the attributes of the input interface and the output interface, it is possible to effectively use network resources.

  Furthermore, according to the present invention, in the optical network, it is possible to correctly determine whether or not a path can be accommodated in each link by performing route calculation based on the product of the number of unreserved interfaces and the maximum bandwidth that can be processed by the interface. The effect is obtained.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram showing a configuration of a node according to an embodiment of the present invention. In FIG. 1, the node 4 according to the embodiment of the present invention includes optical switches 60 and 70, an electric switch 40, and a control device 400.

  The optical switches 60 and 70 switch the optical signal input from the port as light. Therefore, switching is possible without depending on the protocol. The electrical switch 40 once converts the optical signal input from the port into an electrical signal and performs switching. In order to convert an optical signal into an electrical signal and process it as an electrical signal, the protocol that can be processed by the electrical switch 40 is limited. In addition, since optical-electrical-optical conversion is performed, optical signals can be reproduced simultaneously.

  The control device 400 controls the switches inside the node 4. Although not shown in FIG. 1, when a transmitter, a receiver, and a shutter for preventing erroneous light output are present in the node 4, these controls are also performed.

  The control device 400 includes a switch characteristic database (DB) 401 in which characteristics of each switch constituting the node 4 are recorded in advance, an internal link DB 402 in which connection relationships between devices including the switches in the node 4 are recorded in advance, An interface (IF) attribute determining unit 403 that determines an interface attribute in the own node from information recorded in the characteristic DB 401 and the internal link DB 402, and an IF attribute that records the interface attribute determined by the IF attribute determining unit 403 DB 404, a flooding unit 405 that advertises information in IF attribute DB 404 to other nodes in the optical network, a route calculation unit 406 that performs route calculation for the requested path using the information in IF attribute DB 404, and a route Based on the route calculated by the calculation unit 406 And a switch controller 407 which controls the node 4 inside of each switch.

  FIG. 2 is a diagram illustrating an example of the switch characteristic DB 401 for the optical switch 60 of FIG. This example can be used when the ports in the switch have the same attributes except for the input and output directions. The switch ID is used to identify each switch, and represents that a switch connection is possible between ports of the same switch ID. The number of input ports and the number of output ports each represent the number of ports that can be switched. Here, the ports and the interfaces are collectively referred to as ports without distinguishing between the ports and the interfaces.

  The switching capability represents the switching granularity in this switch, “LSC” represents switching processing in units of wavelengths, and “PSC” represents switching processing in units of packets. The bandwidth represents the maximum bandwidth that can be processed by this switch. The protocol represents a protocol that can be processed by this switch, and “ALL” in FIG. 2 indicates that the optical switch can be switched at the wavelength granularity regardless of the protocol. If the protocol that can be processed is, for example, only IPv4 (Internet Protocol Version 4), “IPv4” is entered in the protocol column. Loss, chromatic dispersion, and polarization mode dispersion are attributes unique to an optical switch, and these values can be used to determine degradation of optical signal quality. Each is filled with a value added by passing the optical signal through this switch.

  FIG. 8 is a diagram illustrating an example of the switch characteristic DB 401 for the electrical switch 40 of FIG. Since it is an electrical switch, the switching capability is “PSC” and the bandwidth is 1 Gb / s, which is lower than that of the optical switch.

  FIG. 3 is a diagram showing another example of the switch characteristic DB 401 for the optical switch 60 of FIG. In this example, attributes are entered for each switch port. Each port is identified by the port ID, and it is possible to cope with a case where each port has a different attribute. The contents of the attributes shown in FIG. 3 are the same as the contents of the attributes described in FIG. The characteristics of the optical switch 70 recorded in the switch characteristics DB 401 are the same as those in FIGS.

  FIG. 4 is a diagram illustrating an example of the internal link DB 402 of FIG. The output side indicates that the port is an output port of the switch, the input side indicates that the port is an input port of the switch, and indicates that the connection is connected as a link in the node. 4, in the node 4 of FIG. 1, between the port O3 of the switch 60 and the I4 of the switch 40, between the port O4 of the switch 40 and the port I3 of the switch 60, between the port O7 of the switch 40 and the I8 of the switch 70, It is shown that the port O8 of 70 and the port I7 of the switch 40 are connected by a link.

  The contents of the database shown in FIGS. 3 and 4 can be input to the control unit by an administrator, or can be input using a file or a network. In addition, each switch has a storage device, and the characteristics of each switch are stored in this storage device. When a node is configured, the storage contents of the storage device are input to the control unit through the control line. May be.

  FIG. 5 is a diagram showing an example of the IF attribute DB 404 in which the interface attribute of the node 4 determined by the IF attribute determination unit 403 in FIG. 1 is recorded. The difference from the conventional example of FIG. 22 is that a sub-switch ID is added and the attribute given to the combination of the interface ID and the sub-switch ID is limited to one type. In the conventional example, it is determined that switching is possible between interfaces of the same node. However, in this embodiment, it is determined that switching is possible only between input / output interfaces having the same node ID and sub switch ID shown in FIG. The attribute including the sub switch ID is determined according to the operation flow shown in FIGS.

  6 and 7 are flowcharts showing the operation of the IF attribute determination unit 403 in FIG. 1, FIG. 6 is a flowchart for determining the interface attribute from information in the node, and FIG. 7 shows the sub switch ID in the node. It is a flowchart to determine.

  Hereinafter, a processing procedure for determining an interface attribute will be described with reference to FIGS. 6 and 7.

  First, when a start command for determining an interface attribute in a node is input, the interface attribute determination unit 403 starts a procedure for determining an attribute of the interface in the node, and then assigns a sub switch ID. As a method of issuing the start command, there is a method in which the administrator inputs a command, the administrator issues a command via the control network, and automatically starts when the node is activated.

  First, the interface attribute determination unit 403 stores the characteristic data of each switch in the node as shown in FIG. 3 and FIG. 8 stored in the switch characteristic DB 401, and as shown in FIG. 4 stored in the internal link DB 402. Internal link data is acquired (step S140). The interface attribute determination unit 403 uses ports stored in the switch characteristic DB 401 and also stored in the internal link DB 402 as ports used for internal connection, and those not stored in the internal link DB 402. Are distinguished as interfaces connectable to adjacent nodes (step S141).

  Next, the interface attribute determination unit 403 searches for attributes that each interface can take from the acquired switch characteristic data and internal link data (steps S142 to S149).

  The interface attribute determination unit 403 searches for an input interface that has not received the IF attribute determination procedure in the node (step S142). If there is an interface whose attribute is not yet determined, one interface is selected from the interfaces. For example, in the switch characteristic DB 401 shown in FIG. 3, the following description will be given assuming that the interface I1 is selected from the ports (that is, interfaces) not stored in the internal link DB 402 shown in FIG. This selection method may be any method, for example, ascending order or descending order of ID.

  The interface attribute determination unit 403 searches the selected input interface I1 for an output interface that has not been searched for a route in the interface attribute determination procedure flow (step S144). For the interface I1, the output interfaces O1, O2, O5, O6, O9, and O10 correspond to output interfaces that have not been searched for routes.

  The interface attribute determination unit 403 selects one of the output interfaces O1, O2, O5, O6, O9, and O10. The method for determining the selection order is not limited, but there is, for example, an ascending order or descending order of IDs. Here, first, the interface O1 is selected. Then, the interface attribute determination unit 403 searches for a path connecting the input interface I1 and the selected output interface O1 (step S145). As a path that connects the interfaces I1 and O1, first, a path that directly connects the interfaces I1 and O1 in the optical switch 60 is conceivable.

  Since the interface attribute determination unit 403 has not yet determined the attributes of the interfaces I1 and O1 for this route (step S146), the interface attribute determination unit 403 determines this based on the information in the switch characteristic DB 401 (step S147). Since only the optical switch 60 exists on this path, the characteristics of the optical switch 60 in the switch characteristics DB 401 shown in FIG. 10 are determined as the attributes of the interfaces I1 and O1 as they are (step S147), and the determined attributes are determined. Is recorded in the IF attribute DB 404 (steps S148 and S149).

  As a path connecting the interfaces I1 and O1, there is a path that reaches the interface O1 from the interface I1 via the ports O3, I4, O4, and I3, and the interface attribute determination unit 403 still has the interfaces I1 and O1 for this path. Since the attribute has not been determined (step S146), it is determined based on the information in the switch characteristic DB 401 (step S147). Since the optical switch 60 and the electrical switch 40 exist on this path, the characteristics that can pass through these two switches are selected. That is, the interface attribute determination unit 403 acquires the attributes of the interfaces I1 and O1 and ports O3, I4, O4, and I3 existing on this route from FIGS. 3 and 8, and newly sets values that satisfy these attributes to the interface. It determines as an attribute of I1 and O1 (step S147). Since this route once passes through the electrical switch 40, the switching capability is “PSC”, and the band needs to be adjusted to a small value, and is “1 Gb / s”. The protocol is only “IPv4” that can be processed by the electrical switch 40. The attributes of the interfaces I1 and O1 determined in this way are also recorded in the IF attribute DB 404 (steps S148 and S149).

  If the attributes have been determined for all routes connecting the interfaces I1 and O1, the process returns to step S144. If a route search has already been performed between the interface I1 and each of all output interfaces in the node 4 (step S144), the process returns to step S142.

  For the path connecting the interfaces I1 and O10, a value that can pass through the optical switches 60 and 70 and the electrical switch 60 is taken as an attribute of the interfaces I1 and O10. The attributes of the interfaces I1 and O10 for this route are also limited by the characteristics of the electrical switch 40, the switching capability is “PSC”, the bandwidth is “1 Gb / s”, and the protocol is “IPv4”. In the example shown here, only the optical switch and the electrical switch are devices on the path, but there are, for example, a multiplexer, a duplexer, a shutter, a transmitter, and a receiver in addition to the switch. In this case, the attributes are determined so as to satisfy the attributes of those ports.

  As described above, by performing the processing of steps S142 to S149 in the flowchart shown in FIG. 6 for all interfaces, an IF attribute DB in which no sub switch ID is assigned can be created as shown in FIG. .

  As described above, there is also a wavelength in which a value that can be commonly used by devices on a path is used as an attribute value. A method for determining an attribute when the attribute is a wavelength will be described with reference to FIG. In FIG. 12, the node 300 includes optical switches 301 to 303. Although not shown in FIG. 12, the node 300 has a control unit similar to the control unit 400 of FIG. The selected paths are the interfaces I1 and O1 of the optical switch 301, the interfaces I1 and O1 of the optical switch 302, and the interfaces I1 and O1 of the optical switch 303. The wavelengths that can be processed by each interface are the interface I1 of the optical switch 301, the wavelength λ1, the wavelength λ2, and the wavelength λ3 of the optical switch 301, the interface I1 and O1 of the optical switch 302, the wavelength λ2, the wavelength λ3, the wavelength λ4, and the interface of the optical switch 303. I1 and O1 have a wavelength λ3, a wavelength λ4, and a wavelength λ5. In this case, the IF attribute determination unit of the control unit of the node 300 determines the wavelength λ3, which is a wavelength that can be processed in common for all interfaces, as the interface attribute value.

  Loss, chromatic dispersion, and polarization mode dispersion (PMD) use values obtained by adding the values of each interface as interface values. A method for determining an attribute when the attribute is loss, chromatic dispersion, or polarization mode dispersion will be described with reference to FIG. 13 taking loss as an example. In FIG. 13, the selected paths in the node 300 composed of the optical switches 301 to 303 are the interfaces I1 and O1 of the optical switch 301, the interfaces I1 and O1 of the optical switch 302, and the interfaces I1 and O1 of the optical switch 303. is there. The loss of each interface is 2 dB for the interfaces I1 and O1 of the optical switch 301, 3 dB for the interfaces I1 and O1 of the optical switch 302, and 4 dB for the interfaces I1 and O1 of the optical switch 303. The IF attribute determination unit of the control unit of the node 300 determines 9 dB obtained by adding all these loss values along the route as the interface attribute value. For the chromatic dispersion and polarization mode dispersion, the interface attribute value is determined by the same calculation method.

  Since loss occurs when passing from the input port to the output port of the optical switch, a loss value may be given to only one of the input or output of the switch, and the other value may be set to 0. A half value of the switch loss may be given to each of the outputs.

  If the electrical switch is in the middle of the path, regenerative relaying is performed by the electrical switch, so the calculation is performed with the loss reset to 0 at the port of the electrical switch.

  When using the maximum bandwidth or unreserved bandwidth that can be processed by the device as an attribute (the path is not set on the optical network when the attribute is determined, this unreserved bandwidth corresponds to the maximum bandwidth). The minimum value of the maximum bandwidth that can be processed by each device existing on the searched route is determined as an interface attribute. In addition, as an attribute, the minimum bandwidth that can be processed by the device (for example, even if the maximum bandwidth that can be processed is 10 Gbps, a value such as 0.5 Gbps cannot be supported, and there is a case where 1 Gbps can be supported. 1 Gbps is the minimum bandwidth), the maximum value among the minimum bandwidths that can be processed by each device existing on the searched path in the own node is determined as the interface attribute.

  After determining the interface attribute for each route in the node 4 of FIG. 1 according to the processing procedure shown in FIG. 6, the sub-switch ID assignment flow shown in FIG. 7 is performed. First, the IF attribute determination unit 403 assigns a sub switch ID for each attribute to an interface in the same switch (step S151). Each of the interfaces I1, I2, O1, and O2 of the optical switch 60 has an attribute including the switching capability “LSC” and an attribute including “PSC” as shown in FIG. “60” and “65” are assigned. Similarly, for the optical switch 70, sub-switch IDs “70” and “75” are assigned to each interface. For the electrical switch 40, since all the interfaces are PSC, “40” is assigned as the sub switch ID to all the interfaces. As a result, the temporary sub switch ID is entered in the IF attribute DB 404 as shown in FIG.

  Next, the IF attribute determination unit 403 refers to the internal link DB 402 and checks whether an interface having the same attribute exists between adjacent switches (step S152). Referring to the internal link DB 402 shown in FIG. 4, it can be seen that the optical switch 60 and the electrical switch 40 are connected. In FIG. 10, since the PSC interface exists in the optical switch 60 and the electrical switch 40, the IF attribute determination unit 403 integrates these sub-switch IDs (step S153). Here, a small value is adopted, and the sub-switch ID for the PSC interface in the optical switch 60 and the electrical switch 40 is set to “40”.

  Since the electrical switch 40 and the optical switch 70 are adjacent to each other (step S154), the process returns to step S152, and the electrical switch 40 and the optical switch 70 also have interfaces having the same PSC attribute (step S152). The value of the switch ID is integrated with “40” (step S153).

  When sub-switch ID integration (steps S152 and S153) is completed for all switches, the sub-switch ID assignment flow is also completed, and an IF attribute DB as shown in FIG. 5 is completed.

  The functions of the control apparatus 400 described above can be realized by different devices or can be integrated into the same device.

  As described above, the attribute of each interface of the node 4 can be determined based on the data obtained from the switch characteristic DB 401 and the internal link DB 402 of the node 4. However, in order to actually control the network, it is necessary to give the interface not only the interface information of the own node but also an attribute that can be used in common with the attribute of the adjacent interface to which the link is connected as a link attribute.

  FIG. 14 is a diagram illustrating an example of an optical network including node devices according to an embodiment of the present invention. The optical network shown in FIG. 14 includes nodes 201, 202, 203, 204, and 205. The node 201 includes an electrical switch 211, an optical switch 221, and a control device 231, and the node 202 includes an electrical switch 212, optical switches 222 and 223, and a control device 232. The node 203 includes an electrical switch 213, an optical switch 224, and a control device 233. The node 204 includes an electrical switch 214, an optical switch 225, and a control device 234. The node 205 includes an electrical switch 215. , An optical switch 226, and a control device 235. The control devices 231 to 235 have the same configuration as the control device 400 in FIG. 1, and the node 202 corresponds to the node 4 in FIG.

  In the optical network of FIG. 14, the interface O1 of the node 202 is connected to the interface I5 of the node 203, and both interfaces can correspond to “PSC” and “LSC”. . In general, the nodes are separated from each other by several tens of kilometers, and a connection error between the link and the interface may occur, and the connection destination interface cannot be easily specified.

  Therefore, in order to identify an interface connected between adjacent nodes and exchange interface attributes at both ends of the link, an LMP (Link Management Protocol) proposed by the IETF (Internet Engineering Task Force) is used. For LMP, see J.H. Lang, “Link Management Protocol (LMP)”, draft-ietf-ccamp-lmp-06.

  FIG. 11 shows a sequence of exchanging attributes at both ends of the link using LMP between the node 202 and the node 203 in FIG. 14 as an example. FIG. 11 shows only the interfaces O1, O2, I5, and I6 related to the link between the node 202 and the node 203, and the other interfaces of the nodes 202 and 203 are omitted.

  For exchanging interface attributes, a Link Summary message, a Link Summary Ack message, and a Link Summary Nack message proposed in the IETF draft of LMP can be used.

  First, the attribute of the interface O1 is entered in the Link Summary message from the node 202 and transmitted. The node 203 that has received the Link Summary message checks whether the interface attribute described in the received Link Summary message matches the attribute of the interface I5 of the node 203 to which the interface O1 of the node 202 is connected. If only the input / output directions are different and other attribute values match, the interfaces O1 and I5 at both ends of the link have the same attribute, so the Node 203 sends a Link SummaryAck message to the node 202, and the node 203 The attribute of the interface I5 of the node 203 is determined as the link attribute between the interfaces O1 and I5. When the node 202 also receives the LinkSummaryAck message, the node 202 determines the attribute of the interface O1 of the node 202 as the link attribute between the interfaces O1 and I5 because it can confirm that the attributes match at both ends of the link.

  When the node 203 receives the Link Summary message, if the interface attribute described in the received Link Summary message does not match the attribute of the interface I5 of the node 203, the node 203 sends a Link Summary Nack message to the node 202. To do. The contents of the LinkSummaryNack message are attribute values that satisfy both the attribute described in the received LinkSummary message and the attribute of the interface I5 of the node 203. The method for determining the attribute satisfied by both the interfaces is determined by the same method as the method of step S147 in FIG. When an attribute satisfying the attributes of both interfaces is obtained in this way, it is transmitted to the node 202 as a LinkSummaryNack message. If there is no attribute that simultaneously satisfies the attributes of the interfaces at both ends of the link, this is notified by an error code in the LinkSummaryNack message, and both the interfaces are set to “unusable”. When there is an attribute that satisfies both interfaces, the node 203 that has transmitted the Link Summary Nack message and the node 202 that has received the link adopt this attribute as the link attribute between the interfaces O1 and I5.

  The nodes 202 and 203 record the determined link attribute in a link attribute DB (not shown) in the own node.

  The IF attribute determination unit of each node in the optical network shown in FIG. 14 creates an IF attribute DB as shown in FIG. 5 in accordance with the flow shown in FIGS. 6 and 7, so that the path P8 from the node 201 to the node 205 in the optical network. In the node 202, the interfaces I1 and O9 cannot be connected because the sub switch ID is different in the LSC, and it is understood that the connection is made in the PSC (see FIG. 5). For this reason, since it is understood that the regenerative relay is performed in the node 202, the path P8 does not pass through the electrical switch 214 in the node 204, and network resources can be used effectively. In the optical network of FIG. 14 as well, the number of hops that can be transmitted as an optical signal is two hops as in the optical network of FIG.

  As described above, in the embodiment of the present invention, since the attributes of the input / output interface are determined for each route connecting the input / output interfaces in the node, the types of switches having different input / output interfaces from the attribute information for each route. It is possible to determine whether or not the connection is made through the network, and thus network resources can be used effectively.

  In addition, since interfaces that have the same attribute within a node and can be connected to each other are grouped and an identifier (sub switch ID) for identifying each group is assigned, the same group is referred to by referring to the sub switch ID. It can be determined that the input / output interface to which it belongs can be connected.

  The processing operation of the control device 400 according to the flowcharts shown in FIGS. 6 and 7 is executed by causing a computer, which is a CPU (control unit), to read a program stored in a storage medium such as a ROM in advance. Of course, this can be realized.

  FIG. 15 is a diagram showing an example of an IF attribute DB according to another embodiment of the present invention. In FIG. 15, unlike FIG. 5, when the same interface has a plurality of attributes, a different interface ID is given for each attribute. A physical interface ID is given to a physical interface so that the same interface can be identified.

  By providing an interface ID for each attribute in this way, each interface ID has only one attribute, and there is no need to prepare an attribute value for a combination of multiple values such as “PSC + LSC”. This leads to a reduction in attribute values, and the network control processing load can be reduced compared to the case where there are many types of attribute values.

  When the IF attribute determination unit of each node in the optical network shown in FIG. 14 creates the IF attribute DB shown in FIG. 5 or FIG. 15, when setting the path P8 from the node 201 to the node 205 in the optical network, the path P8 The control device of each of the above nodes uses the flooding unit 405 to advertise to other nodes that the interface on the path P8 in its own node is in use. For example, the interface I1 of the node 202 corresponding to the node 4 in FIG. 1 has two attributes as shown in FIG. 5 and FIG. 15, and among these, the path P8 includes the switching capability “PSC”. The attribute is used, and the attribute including the switching capability “LSC” is not used.

  Therefore, when setting the path P8, the control device 232 of the node 202 advertises that in addition to the selected attribute of the interface I1 on the path, the attribute not selected of the interface I1 is in use. When deleting the path P8, the control device 232 of the node 202 advertises that both of the two attributes of the interface I1 are unused. For example, there is a method of advertising the band of each attribute of the interface as “band = 0” in order to indicate that it is in use and “band = the maximum band that can be processed by the interface” in order to indicate that it is not used. This prevents the remaining attributes of the physically same interface from being used for a new path.

  Next, another embodiment of the present invention will be described with reference to the drawings. The configuration of the optical network according to another embodiment of the present invention is the same as the configuration shown in FIG. 23, and will be described with reference to FIG.

  In this embodiment, the control device 701 determines the product of “the number of unreserved interfaces and the maximum bit rate (maximum bandwidth) that can be processed by each interface” of the optical switch 501 connected to the link 901 with the adjacent node 1002. "Is managed as a" pseudo unreserved bandwidth ". Here, this maximum bandwidth is assumed to be 10 Gbps. The control device 701 advertises the pseudo unreserved bandwidth to all other control devices via a control channel connecting the control devices using a routing protocol such as OSPF (Open Shortest Path First). Similarly, the other control devices 702 to 704 also calculate the pseudo unreserved bandwidth from the number of unreserved interfaces of the links managed by each of the control devices 702 to 704 and advertise the calculated value to all other control devices. As a result, all the control devices 701 to 704 have pseudo unreserved bandwidths for all the links 901 to 903. In the initial state, since the number of unreserved interfaces of all links 901 to 903 is 4, the pseudo unreserved bandwidth is 40 Gbps.

In setting the wavelength path, in the conventional GMPLS (Generalized Multiprotocol Label Switch), the path calculation is performed using only the link in which the unreserved bandwidth satisfies Expression 1 below.
(Unreserved bandwidth) ≧ (bandwidth of requested wavelength path) (Equation 1)
The pseudo unreserved bandwidth used in this embodiment also represents the maximum value of the bandwidth that can be reserved by calculating using the “maximum bandwidth that can be processed by the interface”. Therefore, by replacing the unreserved bandwidth in Equation 1 with the pseudo unreserved bandwidth of the present embodiment, the route can be calculated without changing the current GMPLS route calculation algorithm.

  For example, when the control device 701 is requested to set a wavelength path whose bandwidth from the client 801 to the client 803 is 10 Gbps, the control device 701 calculates the shortest path using only a link whose pseudo unreserved bandwidth is 10 Gbps or more. Then, a signaling message for setting a path along the obtained route is transmitted. In the process of this signaling, the control device 701 assigns the interfaces p4 and p8 of the optical switch 501 to the wavelength path P21. As a result, the number of unreserved interfaces in the link 901 is 3, and the pseudo unreserved bandwidth is 30 Gbps. The control device 701 immediately advertises the pseudo unreserved bandwidth of the link 901 after the change to another control device. Similarly, when the wavelength paths P22 and P23 having bandwidths of 10 Gbps and 2.5 Gbps are set, the number of unreserved interfaces of the link 901 is 1, and the pseudo unreserved bandwidth is 10 Gbps.

  Here, when a wavelength path with a bandwidth of 2.5 Gbps is requested from the client 801 to the client 802, the path P24 is set. In the conventional method of managing the unreserved bandwidth after setting the path P24, the unreserved bandwidth of 7.5 Gbps is advertised as the remaining allocation of 10 Gbps to 2.5 Gbps. However, in practice, even if an attempt is made to set a wavelength path of 7.5 Gbps or less for this link, there is no usable interface, and therefore the path cannot be set. In the method according to the present embodiment, since the number of unreserved interfaces becomes 0 after setting the path P24, the pseudo unreserved bandwidth to be advertised is also 0. Can be advertised. This is because the number of unreserved interfaces used in this embodiment does not depend on the bit rate.

It is a figure which shows the structure of the node apparatus by the Example of this invention. It is a figure which shows the example of switch characteristic DB of FIG. 1 about an optical switch. It is a figure which shows another example of switch characteristic DB of FIG. 1 about an optical switch. It is a figure which shows the example of internal link DB of FIG. It is a figure which shows the example of IF attribute DB of FIG. It is a flowchart which shows operation | movement of the IF attribute determination part of FIG. It is a flowchart which shows operation | movement of the IF attribute determination part of FIG. It is a figure which shows the example of switch characteristic DB of FIG. 1 about an electrical switch. FIG. 7 is a diagram illustrating an example of the IF attribute DB of FIG. 1 in a state where IF attributes are determined according to the flow of FIG. 6. It is a figure which shows the example of IF attribute DB of FIG. 1 in the state where the temporary subswitch ID attribute was filled in. FIG. 5 is a diagram illustrating a sequence for exchanging attributes between adjacent nodes of an optical network according to an embodiment of the present invention. It is a figure for demonstrating the determination method of the attribute value in case the attribute is a wavelength in the Example of this invention. It is a figure for demonstrating the determination method of the attribute value in case the attribute is loss in the Example of this invention. It is a figure which shows the example of the optical network comprised from the node apparatus by the Example of this invention. It is a figure which shows the example of IF attribute DB by the other Example of this invention. It is a figure which shows the structural example of an optical network. It is a figure which shows the structural example of the node of an optical network. It is a figure which shows another structural example of the node of an optical network. It is a figure for demonstrating the determination method of the conventional interface attribute. It is a figure which shows another structural example of the node of an optical network. It is a figure which shows the structural example of the optical network using the node structure of FIG. It is a figure which shows the attribute of each interface determined by the determination method of the conventional interface attribute. It is a figure which shows the structural example of the optical network for demonstrating the problem of the conventional route calculation method.

Explanation of symbols

4 node equipment
40 electrical switch 60, 70 optical switch 400 control device 401 switch characteristic DB
402 Internal link DB
403 IF attribute determination unit 404 IF attribute DB
405 Flooding unit 406 Route calculation unit 407 Switch control unit

Claims (7)

An optical switch that can exchange input signals in units of wavelengths;
In response to a path setting request, the number of unreserved interfaces among the interfaces related to the link connecting the node device and the node device adjacent to the node device, and the maximum bandwidth that each of the interfaces can process are used. Advertising means for calculating the unreserved bandwidth of the link and advertising to other node devices;
A node device comprising:   The node device according to claim 1, wherein the unreserved bandwidth of the link is calculated as a product of the number of unreserved interfaces and a maximum bandwidth that can be processed by each of the interfaces. 3. The node device according to claim 1, wherein path calculation of a path is performed using only a link whose unreserved bandwidth is equal to or greater than a requested path bandwidth . An optical network comprising the node device according to any one of claims 1 to 3. Receiving a path setting request;
The number of unreserved interfaces among the interfaces related to the link connecting the node device including the optical switch that can exchange the input signal in units of wavelengths and the node device adjacent to the node device, and each of the interfaces processes Calculating the unreserved bandwidth of the link using the maximum possible bandwidth and advertising to other node devices;
A route calculation method comprising: 6. The route calculation method according to claim 5, wherein the unreserved bandwidth of the link is calculated as a product of the number of unreserved interfaces and the maximum bandwidth that can be processed by each of the interfaces . The route calculation method according to claim 5 or 6 , wherein the path route calculation is performed using only a link whose unreserved bandwidth is equal to or greater than the bandwidth of the requested path .

Images