JP4977153B2 - Node, network, node control device, link control method and program thereof - Google Patents

Description

  The present invention relates to a technique for powering off links in a network as much as possible.

  Conventionally, there are a technique for distributing the load in the network by using multipaths in the network, and a technique for improving multipath QoS (Quality of Service) and throughput (see Patent Document 1).

JP 2000-13439 A

  However, all of the above-described technologies enable communication via a link even if the link usage rate of the link connecting the nodes (the traffic volume of the link ÷ the maximum bandwidth used for the link) is reduced. Power is required. For example, when Ethernet (registered trademark) is used for a link, about 40 W of power is required for each 10 Gbps link. This is the same when the link usage rate is relatively low. Even when an optical link is used, the communication speed cannot be reduced due to the PLL (Phase-Locked Light). Therefore, even if the link usage rate is reduced, the power consumption is almost the same as when communicating at the full rate.

  Therefore, the present invention aims to solve the above-described problems and reduce the power consumption of the entire network.

  In order to solve the above-described problem, the invention according to claim 1 is a node for determining a route in a network by a shortest path routing, and shows a network topology and a link cost value of each link in the network. A storage unit that stores the topology information and link usage rate information indicating the maximum available bandwidth and link usage rate of each link connected to the node, a connection reception unit that receives connection establishment requests and disconnection requests, A path calculation unit that determines the link to be used in the shortest test path of the connection for which the establishment request is made with reference to the topology indicated in the topology information and the link cost of each link, and a connection establishment / disconnection request on the determined link Connection confirmation to disconnect the target connection A link usage rate update unit that updates the link usage rate in the link usage rate information based on the traffic volume of the established connection and the disconnected connection, and the updated link usage rate information. The link cost of the link is calculated as a lower value as the link usage rate of the link is higher with the predetermined link usage rate as an upper limit value, and the link cost of the link in the topology information is calculated using the calculated value. The cost calculation unit to be updated, the power control unit that controls power-off or power-on of the link connected to the node, and the link of the link connected to its own node with reference to the updated link usage rate information When it is determined that the usage rate is 0, a power control determination unit that instructs the power control unit to power off the link; When the connection accepting unit accepts a request for establishing a new connection, the route calculating unit refers to the link cost of each link indicated in the updated topology information, and shows the test of the connection for the establishment request. It is characterized by calculating the link used in the path.

  The invention according to claim 7 is a topology information indicating a network topology and a link cost value of each link in the network, and a link indicating a maximum available bandwidth and a link usage rate of each link connected to the node. A node having a storage unit for storing usage rate information accepts a connection establishment request and a disconnection request, refers to the topology indicated in the topology information and the link cost of each link, and A step of determining a link to be used in the shortest path, a step of establishing a connection on the determined link and disconnecting a connection for which a disconnection request is requested, and a traffic volume of the established connection or a traffic volume of the disconnected connection Based on link usage information Based on the updated link usage rate information and the link usage rate information, the link cost of the link is calculated as a lower value as the link usage rate of the link is higher with the predetermined link usage rate as the upper limit. When the link usage rate of the link connected to its own node becomes 0 as a result of the step of updating the link cost of the link in the topology information using the calculated value and the update of the link usage rate Instructing the node comprising the link to power off the link and, when receiving a request for establishing a new connection, refer to the link cost of each link indicated in the updated topology information and Determining a link to be used in the show test pass of the target connection. And the control method.

  By doing so, a node (for example, a router) sets the link cost value of a link (that is, “0.8”) as the upper limit value of the link cost of the link (that is, the link cost of the link) The higher the (traffic volume of the connection established on the link / the maximum usable bandwidth of the link), the lower the value. Here, when a node calculates a connection route, a link with the smallest possible link cost is selected by the show test path. Therefore, when the node accepts a request for establishing a new connection, a larger amount of traffic flows. The link will be selected. That is, a new connection is established in a link through which more traffic flows by repeating the above process every time the node receives a request for establishing a new connection. Here, the node disconnects the existing connection every time it receives a connection disconnection request. Therefore, the connection established in the link with a small amount of traffic is gradually disappeared. Then, the node powers off the link for the link in which the connection disappears in this way (that is, the link whose traffic volume becomes “0”). Therefore, the power consumption of the node can be reduced, and the power consumption of the entire network constituted by such nodes can also be reduced. Note that when the node calculates the link cost, the higher the link usage rate, the lower the link cost value is calculated to a predetermined link usage rate (eg, “0.8”). Therefore, the node can prevent a link having an excessive link usage rate (for example, “0.9”) from occurring.

  In the invention according to claim 2, the node according to claim 1 stores the link information indicating the identification information of the link turned off in the storage unit, and the link of the link turned off in the topology information is stored. A link management unit that sets a predetermined value smaller than the link cost value corresponding to a predetermined link usage rate as a link cost value, and is used in the shortest test path of the connection that is the target of the establishment request by the route calculation unit If the link used in the shortest test path includes a link whose link usage rate is equal to or higher than the predetermined link usage rate, the power control determination unit refers to the link information and topology information. When there is a link whose link usage rate is equal to or higher than the predetermined link usage rate and a link that is connected to the same link and that is powered off When it is disconnected, the power control unit is instructed to power on the link, and the route calculation unit again determines the link to be used in the shortest test path of the connection for which the establishment request is made, including the link that is powered on. The connection establishment unit establishes a connection on the determined link, the link usage rate update unit updates the link usage rate information based on the traffic volume of the established connection, and the cost calculation unit Based on the updated link usage rate information, the link cost of the link is calculated as a higher value as the link usage rate of the link is higher, and the link cost of the link in the topology information is updated using this calculated value. When the connection accepting unit accepts a request for establishing a new connection, the route calculating unit displays updated topology information. Referring to the link cost of each link represented, and determines the link to be used in the Shortest Path object of connection establishment request.

  The invention according to claim 8 is the link control method according to claim 7, wherein the node stores link information indicating identification information of the link that is powered off in the storage unit, and topology information, As a link cost value of a link that has been powered off, a step of setting a predetermined value smaller than a link cost value corresponding to a predetermined link usage rate, and a link used in a show test path of a connection for which an establishment request is made As a result of the determination, if there is a link whose link usage rate in the link usage rate information is equal to or higher than the predetermined link usage rate, the link usage rate is determined to It is determined that there is a link that has the same link destination as the link usage rate or higher and that is powered off. A step of instructing a node including the link to power on the link; a step of determining a link to be used again in a show test path of a connection for which an establishment request is made, including the link that is powered on; A step of establishing a connection on the determined link, a step of updating link usage rate information based on the traffic volume of the established connection, and a link cost of the link based on the updated link usage rate information Is calculated as a higher value as the link usage rate of the link is higher, and using this calculated value, a step of updating the link cost of the link in the topology information and a request for establishing a new connection are received. Request establishment by referring to the link cost of each link indicated in the updated topology information And executes a step of determining a link used in the Shortest Path of the target connection.

  According to such a node, when there is a link (overload link) having a predetermined link usage rate (for example, “0.9”) or more, this overload link is selected from among the links that have been powered off. Select the link with the same connection destination as, and power it on. Note that the node calculates the link cost as a higher value as the link usage rate is higher based on the updated link usage rate information after powering on the link that was powered off. For a link that is powered on, when the link is powered off, a link cost value (for example, “99”) corresponding to the predetermined link usage rate (for example, “0.9”). Since a value smaller than () is set, the power-on link is selected as a route for a new connection. In other words, when there is a link where traffic is excessively concentrated, a new connection is established on the link that was previously powered off. Therefore, it is possible to eliminate excessive traffic concentration on the links in the network and secure a stable communication environment.

  A third aspect of the present invention is a network comprising a plurality of nodes according to the first or second aspect.

  According to such a network, the power consumption of the whole network can be reduced and a stable communication environment can be ensured.

  According to a fourth aspect of the present invention, there is provided a node control apparatus for controlling a node in a network, the topology information indicating a network topology and a link cost value of each link in the network, and a link connected to the node. A storage unit that stores link usage rate information indicating the maximum available bandwidth and link usage rate, a connection reception unit that receives connection establishment requests and disconnection requests, and the topology and link of each link indicated in the topology information Referring to the cost, the path calculation unit that determines the link to be used in the show test path of the connection that is the target of the establishment request, and the instruction for the establishment of the connection on the determined link and the disconnection of the target connection of the disconnection request Connection establishment control unit to send to the The link usage rate update unit that updates the link usage rate in the link usage rate information based on the traffic volume of the link and the traffic volume of the disconnected connection, and the link cost of the link based on the updated link usage rate information A cost calculation unit that updates the link cost of the link in the topology information using the calculated value as a lower value as the link usage rate of the link is higher. As a result of calculation of the link usage rate by the power control unit that transmits to the node a power off or power on instruction of the link connected to the node and the link usage rate update unit, the link usage rate becomes zero. A power control determining unit that instructs the power control unit to turn off the link.

  By doing so, the node control device controls each node in the network and establishes a new connection by biasing it to a predetermined link. As a result, the traffic volume is likely to be “0”, so the link whose traffic volume is “0” can be powered off. Therefore, the power consumption of the node can be reduced, and the power consumption of the entire network constituted by such nodes can also be reduced. This node control device is realized by a PCE (Path Computation Element).

  In the fifth aspect of the invention, the node control device according to the fourth aspect of the present invention provides the node control device according to the fourth aspect of the present invention, by using the OSPF path information, the identification information of the link that has been powered off and the information indicating the link cost value of the link. The received OSPF path information receiving unit and the link information indicating the identification information of the link turned off shown in the received OSPF path information are stored in the storage unit, and the link of the link turned off in the topology information A link management unit that sets a value smaller than the link cost value corresponding to a predetermined link usage rate as a cost value is determined, and the link used in the shortest test path of the connection of the establishment request is determined by the route calculation unit As a result, the link usage rate in the link usage rate information is higher than the predetermined link usage rate for the link used in the show test pass. If there is a link, the power control determination unit refers to the link information and the topology information, and the link whose link usage rate is equal to or higher than the predetermined link usage rate is the same link as the connection destination, and the link is powered off. When it is determined that there is a link, the power control unit is instructed to power on the link, and the path calculation unit includes the link that has been turned on again and uses it again in the show test path of the connection for which the establishment request is made. The link is determined, the connection establishment control unit transmits an instruction to establish the connection on the determined link to the node, and the link usage rate update unit determines the link based on the traffic volume of the established connection. The usage rate information is updated, and the cost calculation unit calculates the link cost of the link based on the updated link usage rate information. The higher the value is, the higher the value is calculated, and using this calculated value, the link cost of the link in the topology information is updated, and when the connection reception unit receives a request for establishing a new connection, the route calculation unit The link cost of each link indicated in the updated topology information is referred to determine a link to be used in the short test path of the connection that is the target of the establishment request.

  In this way, when there is a link whose link usage rate exceeds a predetermined value (overload link), the node controller can connect the destination of this overload link from the links that have been powered off until now. Select the same destination link and power on. Then, a new connection is established using the power-on link. Therefore, excessive traffic concentration in the network can be eliminated and a stable communication environment can be secured.

  The invention according to claim 6 is a node that performs power-off and power-off of a link based on an instruction from the node control device according to claim 4 or 5, and when the power of the link is performed. The OSPF path information in which the link identification information and the link cost value of the link are set is advertised to the node control device.

  By doing in this way, the node notifies the node control apparatus of the identification information of the link powered off by the OSPF path information and the link cost set for the link. Therefore, the node control apparatus can know that the link is surely powered off at this node and the link cost of the powered off link.

  The invention according to claim 9 is a program for causing a computer to execute the link control method according to claim 7 or claim 8.

  According to such a program, it is possible to cause a general computer (for example, a router) to execute the link control method according to claim 7 or claim 8.

  According to the present invention, the power consumption of the entire network can be reduced.

It is the figure explaining the outline | summary of the control of the power-off or power-on of the link by the node of 1st Embodiment. It is the figure explaining the outline | summary of the control of the power-off or power-on of the link by the node of 1st Embodiment. It is the block diagram which showed the structure of the node of 1st Embodiment. It is the figure which illustrated the topology information of FIG. It is the flowchart which showed the process sequence of the node of FIG. It is a block diagram of the system containing the node control apparatus of 2nd Embodiment. It is the block diagram which showed the structure of the node control apparatus of 2nd Embodiment. It is the block diagram which showed the structure of the node of 2nd Embodiment. It is the flowchart which showed the process sequence of the node control apparatus of FIG. It is the figure quoted in order to demonstrate the effect of each embodiment. It is the figure which illustrated the calculation formula and the graph of the link cost which the cost calculation part of each embodiment calculates. (A) And (b) is the figure which showed the structural example of the node of FIG. (C) is the figure quoted in order to demonstrate other embodiment.

  Hereinafter, modes for carrying out the present invention will be described by dividing them into a first embodiment and a second embodiment.

<< First Embodiment >>
First, an outline of link power-off or power-on control by a node according to the first embodiment will be described with reference to FIGS. 1 and 2. Here, the node is a node that establishes a connection and transmits / receives data through the established connection. This node selects a path (link) used in this connection by show test path routing such as OSPF (Open Shortest Path First). That is, it is assumed that a node that has received a connection establishment request establishes a connection to a link with a lower link cost among links connected to this node. The link usage rate in the following description is (traffic amount of connection established for the link / maximum usable bandwidth of the link). Further, as shown in the lower left of FIG. 1, this node 10 is connected to another node 10 by a plurality of links, for example, a first link and a second link.

Here, the node calculates the link cost of the link connected to its own node based on, for example, Expression (1).
Max (10 × 1 / (1-link usage rate), 10 × 1 / (link usage rate)) (1)
That is, the node selects, as the link cost, the larger one of the value obtained by multiplying the reciprocal of the link unused ratio by 10 and the value obtained by multiplying the reciprocal of the link utilization ratio by 10. The graph on the right side of FIG. 1 is a graph showing the link cost calculated by equation (1). That is, if the link usage rate is less than “0.5”, the higher the link usage rate, the smaller the value of the link cost. Therefore, if the link usage rate is less than “0.5”, a link with a high link usage rate is likely to be used in the show test path routing. On the other hand, when the link usage rate becomes “0.5” or more, the higher the link usage rate, the larger the link cost value. In other words, it is difficult to use in show test pass routing.

  Therefore, for example, the link usage rates of the first link and the second link connected to the node are both “0.5” or less, and as shown in the upper left of FIG. If the link usage rate of the link is high, the cost of the first link is calculated low. Therefore, since the node selects the first link as the establishment destination of the new connection, the traffic amount of the first link increases. That is, the link usage rate increases. This further reduces the link cost of the first link and concentrates new connections on the first link. On the other hand, no new connection is established on the second link, and if the existing connection disappears, the traffic amount is reduced as it is, and finally the traffic disappears. And since there is no need to power on the link whose traffic has disappeared, the node powers off this link.

  This will be described in detail with reference to FIG. For example, when the link usage rate is “0.2” for both the first link and the second link, the link cost is both “50”. In this case, the node selects an arbitrary link as the establishment destination of the new connection. Here, when the node selects the first link as the establishment destination of the new connection, the traffic amount of the first link increases. For example, the link usage rate becomes “0.3” and the link cost becomes “33”. Further, the link usage rate of the second link is “0.1”, and the link cost is “70”. Therefore, the node selects the low-cost link, that is, the first link as the establishment destination of the new connection, and no new connection is established on the second link. Therefore, the connection of the second link gradually decreases, and the link usage rate of the second link decreases (this increases the link cost). Then, when all the connections of the second link disappear, the node powers off the second link. Note that the node sets the link cost of the power-off link to, for example, “99”. On the other hand, since new connections are established for the first link, the link usage rate is “0.4” and the link cost is “25”.

  Thereafter, a processing procedure when the traffic amount of the first link further increases will be described with reference to FIG. When the node powers off the second link, the node continues to establish new connections on the first link. As a result, for example, when the link usage rate of the first link becomes “0.7”, the node calculates the link cost as about “33” based on the above-described equation (1). Since the link usage rate is “0.5” or more, the node calculates the link cost by using the value of 10 × 1 (1-link usage rate) in the equation (1) (in the graph of FIG. 1). (Refer to the link usage rate of “0.5” or higher).

  Thereafter, if new connections continue to be established on the first link, the first link becomes heavily loaded, and the link cost further increases. For example, the link cost of the first link is “120”, and the link usage rate is approximately “0.83”. If new connections continue to be established on the first link, the first link becomes overloaded, and the node powers on the second link. That is, since the link usage rate in the graph of FIG. 1 is “0.9” or more, the second link that is currently powered off is powered on. Here, the link cost of the second link that is powered on is “99” as described above. That is, since this link cost value is lower than the link cost (“120”) of the first link, the node selects the second link as the establishment destination of the new connection. As a result, traffic of a new connection starts to flow through the second link. For example, the link usage rate of the first link is “0.85”, the link cost is about “66”, and the link usage rate of the second link is “ 0.15 "and the link cost is also about" 66 ". That is, it is possible to prevent traffic from being concentrated on one link (for example, the first link).

  As described above, when the link usage rate exceeds a predetermined value (for example, “0.9”), the node powers on the link that has been powered off and establishes a new connection to this link. To. Here, when the link usage rate exceeds a predetermined value (for example, “0.5”), the node calculates the link cost as a higher value as the link usage rate increases (see Expression (1) and the graph). ). That is, it is difficult to establish a new connection for a high-load link. Further, a value (for example, “99”) lower than the link cost of the high-load link is set as the link cost of the link that is powered off. By doing this, when this link is powered on, the link that is powered on has a lower link cost than a heavily loaded link, and a new connection is established on this powered on link. The load on high-load links is reduced.

  By constructing a network using the nodes described above, traffic flowing between the nodes can be brought to a predetermined link, and a link where traffic does not flow can be created. And since the link where this traffic stops flowing is powered off, the power consumption of the entire network can be reduced. In addition, when there is a high load link, the node powers on and restores the link that has been powered off, and allows traffic to flow to this link. Therefore, it is possible to secure a stable communication environment while reducing the power consumption of the entire network as much as possible.

<Configuration>
The configuration of such a node will be described with reference to FIG. As illustrated in FIG. 3, the node 10 includes an interface unit 11 (11A, 11B,..., 11N), a processing unit 12, a storage unit 13, and a power control unit 14. This node 10 is realized by a router that establishes a connection with another node 10. The processing unit 12 is realized by a program execution process by a CPU (Central Processing Unit) included in the node 10 or a dedicated circuit. Further, the storage unit 13 includes a storage medium such as a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), and a flash memory. When the node 10 is realized by program execution processing, the storage unit 13 stores a program for realizing the function of the node 10.

  The interface unit 11 accommodates a link that is a data transmission path with another node 10 (not shown). The interface unit 11 outputs the data received via the link to the processing unit 12, and outputs the data output from the processing unit 12 to another node 10 via the link. The interface unit 11 may be an optical link interface or an interface such as Ethernet. The interface unit 11 is powered on or powered off by a power control unit 14 (details will be described later). When the power is turned on, data communication via the link accommodated in the interface unit 11 becomes possible, and when the power is turned off, data communication via the link accommodated in the interface unit 11 cannot be performed.

  The processing unit 12 controls the entire node 10, and here determines a path (link) mainly used for connection, and controls link power-off and power-off. Such a processing unit 12 includes a connection reception unit 121, a route calculation unit 122, a connection establishment unit 123, a link usage rate update unit 124, a cost calculation unit 125, a cost setting unit 126, and a power control determination unit. 127 and a link management unit 128.

  The connection receiving unit 121 receives a connection establishment request from an external device via the interface unit 11. This connection establishment request includes information such as a connection start point node, an end point node, and a traffic amount allocated to the connection.

  The route calculation unit 122 refers to the topology of the network indicated in the topology information 131 (described later) and the link cost of each link, and determines a route (link) to be used for the connection that is the target of the establishment request. The route calculation here is performed by the route calculation unit 122 using a shortest path routing such as OSPF to select a link that minimizes the link cost. The protocol used in the route calculation unit 122 is not limited to OSPF as long as it is a protocol that determines a route from a shortest path (shortest route) based on a link cost value. For example, an IS-IS (Intermediate System-to-Intermediate System) protocol may be used.

  The connection establishment unit 123 uses the link selected by the route calculation unit 122 to establish a connection from the start point node to the end point node.

  The link usage rate updating unit 124 updates the link usage rate of each link in the link usage rate information 132 (described later) based on the traffic volume of the connection established on the link and the traffic volume of the disconnected connection.

  The cost calculation unit 125 calculates the link cost of the link in the topology information 131 based on the updated link usage rate information 132 and updates the link cost value in the topology information 131 using the calculated value. The value of the link cost here is calculated based on the above-described formula (1), for example. That is, as illustrated in the graph of FIG. 1, the cost calculation unit 125 increases the link usage rate if the link usage rate is less than a predetermined value (“0.5” in the case of Expression (1)). The link cost value is calculated as a small value. That is, the link cost is calculated as a smaller value for a link with a larger traffic volume. On the other hand, if the link usage rate is equal to or greater than a predetermined value (“0.5” in the case of equation (1)), the higher the link usage rate, the larger the link cost value is calculated. That is, the link cost is calculated as a larger value for a link with a larger traffic volume. Here, since the route calculation unit 122 selects a link having a lower link cost by the above-described shortest path routing, the link usage rate becomes a predetermined value (“0.5” in the case of Expression (1)). Until then, a new connection is established on a link with a larger traffic volume. When the link usage rate exceeds a predetermined value (in the case of equation (1), “0.5”), the higher the link usage rate, the higher the link cost. Thus, for example, if a link is newly powered on when the link usage rate of this link exceeds “0.9”, the link cost may be lower for the newly powered on link. Get higher. Therefore, a new connection is easily established on the newly powered-on link, and the load distribution of the link that has been heavily loaded can be achieved.

  The cost setting unit 126 sets the link cost of each link calculated by the cost calculation unit 125 for the interface unit 11 that accommodates the link.

  The power control determination unit 127 determines whether to instruct the power control unit 14 (described later) to power off the link or to instruct power on. Specifically, when the link usage rate of the link connected to its own node 10 becomes “0” in the link usage rate information 132, the power control determination unit 127 causes the power control unit 14 to Instruct power off. That is, a link that does not need to carry traffic is powered off. When the link usage rate of the link connected to the node 10 in the link usage rate information 132 exceeds a predetermined value (for example, “0.9”), the power control determination unit 127 displays the power control unit 127. 14 is instructed to turn on the power-off link connected to its own node 10. The link instructing power-on here is a link connected to the same connection destination (node 10) as the link (high load link) whose link usage rate exceeds a predetermined value. When there is no link (high load link) whose link usage rate exceeds a predetermined value and no link connected to the same connection destination (node 10), the power control determination unit 127 blocks this high load link. To do. That is, this high load link is not used for communication. By doing in this way, it can avoid that the load of a high load link becomes still higher.

  The link management unit 128 writes the power-on / power-on information for each link in the link information 133 (described later). For example, the identification information of the link powered off by the power control unit 14 is written in the link information 133. When the link is powered on, the link management unit 128 deletes the identification information of the powered-on link from the link information 133. Also, the link management unit 128 sets a predetermined value as the link cost of the link that has been powered off in the topology information 131. Here, the value of the link cost set for the link that is powered off is the threshold value of the link usage rate used when the power control determination unit 127 determines that other links are powered on (for example, “0.9”). )) Is set to a value smaller than the link cost value. For example, when the threshold of the link usage rate is “0.9”, the link cost value corresponding to the link usage rate “0.9” in the formula (1) is “100”. A value “99” smaller than “100” is set as the cost. In this way, a new connection is easily established on the link that is powered on.

  Based on an instruction from the power control determination unit 127, the power control unit 14 performs power on or power off of the link to the interface unit 11 that accommodates the link to be powered on or powered off.

  The storage unit 13 stores topology information 131, link usage rate information 132, and link information 133.

  The topology information 131 is information indicating which node 10 and which node 10 are linked in the network, and includes the value of the link cost of each link in the network. For example, as shown in FIG. 4, the topology information 131 is information indicating how the nodes 10 (10A, 10B, 10C, 10D) in the network are linked to each other. , Including the link cost value of each link. For example, in the topology information 131 of FIG. 4, the node 10A is connected to the node 10B by two links, and the link costs of the links are “30” and “10”, respectively. The node 10A is connected to the node 10C by one link, and the link cost of the link is “10”. The same applies to other cases. The topology information 131 is referred to when the route calculation unit 122 calculates the connection route. The initial value of the link cost value in the topology information 131 is initially set, and then updated to the link cost value calculated by the cost calculation unit 125.

  The link usage rate information 132 is information indicating the link usage rate of each link connected to its own node. As illustrated in Table 1 below, this link usage rate information 132 is, for each link ID of a link connected to its own node 10, the maximum usable bandwidth of the link, the traffic amount in the link, The amount of traffic divided by the link usage rate which is a value of the maximum available bandwidth is indicated. The link usage rate information 132 is updated by the link usage rate update unit 124. For example, when a new connection is established, the link usage rate update unit 124 increases the traffic volume of the link that is the route of the new connection, and increases the value of the link usage rate accordingly. Further, when the connection is disconnected, the link usage rate update unit 124 decreases the traffic amount of the link that was the path of the connection, and also decreases the value of the link usage rate accordingly. The value of the link usage rate in the link usage rate information 132 is referred to when the path calculation unit 122 calculates the link cost or when the power control determination unit 127 determines whether the link power is on or off. .

  The link information 133 (power-off / on information) is information indicating identification information of a link that is powered off in its own node 10. The link information 133 may include identification information of the link that is powered on. For example, as illustrated in Table 2 below, for each link identification information, whether the link is powered on or powered off. May be described. By referring to such link information, the power control determination unit 127 can know whether each link is powered on or powered off.

<Processing procedure>
Next, the processing procedure of the node 10 will be described with reference to FIG. First, the connection reception unit 121 of the node 10 in FIG. 3 receives a request for establishing a new connection (S101). Next, the route calculation unit 122 calculates a show test path (S102). That is, the route calculation unit 122 refers to the network topology indicated in the topology information 131 and the link cost of each link, and among the links connected to its own node 10, the link used in the connection that is the target of this establishment request. Select one with a low link cost.

  Here, the power control determination unit 127 determines whether or not there is a link (high load link) having a link usage rate equal to or higher than a predetermined value (for example, “0.9”) among the selected links ( S103). Here, if there is no link whose link usage rate is a predetermined value (for example, “0.9”) or more (No in S103), the connection establishment unit 123 performs the new connection establishment process received by the connection reception unit 121. Do. Thereby, the packet is transferred to the link (new connection) of the shortest path calculated in S102 (S107). Therefore, the traffic volume of the link of this show test path increases.

  Thereafter, the node 10 updates the link cost (S108). That is, first, the link usage rate updating unit 124 of the node 10 calculates the link usage rate of this link based on the traffic amount of the link obtained by adding the traffic amount of the new connection, and updates the link usage rate information 132. To do. Then, the cost calculation unit 125 calculates the link cost of the link that has established the new connection using the above-described equation (1) based on the link usage rate indicated in the updated link usage rate information 132. calculate. Next, the cost calculation unit 125 updates the link cost value of the link that has established a new connection in the topology information 131 to the calculated link cost value. Further, the cost setting unit 126 sets the link cost value calculated by the cost calculation unit 125 in the interface unit 11.

  After the link cost is updated, the power control determination unit 127 refers to the link usage rate information 132, and the link connected to its own node 10 has a link usage rate of “0”. It is determined whether or not there is (S109). That is, it is determined whether or not there is a link having a traffic volume of “0”. Here, if there is a link whose link usage rate is “0” (Yes in S109), the power control determination unit 127 instructs the power control unit 14 to power off the link (S110). In response to this, the power control unit 14 executes power-off of the link. Then, the link management unit 128 writes in the link information 133 that the link is powered off. Further, the link management unit 128 sets a predetermined value (for example, “99”) in the topology information 131 as the value of the link cost of the power-off link. On the other hand, if there is no link whose link usage rate is “0” in S109 (No in S109), the process is terminated as it is.

  In S103, when the power control determination unit 127 determines that there is a link (high load link) having a link usage rate of a predetermined value (for example, “0.9”) or more among the selected links ( In step S103, the power control determination unit 127 refers to the topology information 131 and the link information 133, and a link that is connected to the same connection destination (node 10) as the high-load link is a link that is powered off. It is determined whether or not there is (S104). Here, if there is a link that is connected to the same connection destination (node 10) as this high-load link and that is powered off (Yes in S104), the power control determination unit 127 causes the power control unit 14 to The power-on to the link that is turned off is instructed. Then, the power control unit 14 powers on the link that is powered off (S105). Then, the link management unit 128 writes in the link information 133 that this link is powered on, and returns to S102. That is, the show test path is calculated again including the link that is powered on. On the other hand, in S104, the power control determination unit 127 refers to the topology information 131 and the link information 133, and the link that is connected to the same connection destination (node 10) as the high-load link is a power-off link. If not (No in S104), the high load link is blocked (S106). Then, the process ends. By doing in this way, it can avoid that the load of a high load link becomes still higher.

  Although not described here, every time the node 10 receives a connection disconnection request from the connection reception unit 121, the connection establishment unit 123 performs the connection disconnection process, and the link usage rate update unit 124 The traffic volume of the link whose link is disconnected is reduced, and the link usage rate information 132 is updated. Then, as in S109, the power control determination unit 127 refers to the updated link usage rate information 132, and if there is a link whose link usage rate is “0”, the power control determination unit 127 turns off the power of the link. The power control unit 14 is instructed. Therefore, the node 10 executes the processing described above every time it receives a request for establishing a new connection, so that traffic flowing through a plurality of links can be concentrated on a predetermined link with a predetermined link usage rate as an upper limit. . As a result, the connection disappears gradually, and a link with a traffic amount of “0” is easily created. By powering off this link, the power consumption of the entire network can be reduced. In addition, after a link is powered off in this way, if the power-on link becomes heavily loaded, the power-off link is restored and a new connection is established for this link. It can prevent becoming a load.

<< Second Embodiment >>
Next, a second embodiment of the present invention will be described with reference to FIG. The second embodiment is a system including a node 30 and a node control device 20 that establishes a connection between the nodes 30 and calculates a route in the connection. This node control device 20 controls the power off or power on of the link of each node 30. That is, the node control device 20 acquires the traffic amount of each link of the network, and updates the link cost of each link based on the acquired track amount. Therefore, the node control device 20 can determine which link among the links of the entire network should be powered off and which link should be powered off. The node control device 20 is realized by, for example, a PCE that remotely controls the node 30 (30A, 30B, 30C, 30D).

<Configuration>
The configuration of such a node control device 20 will be described with reference to FIG. As illustrated in FIG. 7, the node control device 20 includes an interface unit 21, a processing unit 22, a storage unit 23, and a power control unit 24. The processing unit 22 is realized by a program execution process by a CPU provided in the node control device 20, a dedicated circuit, or the like. Furthermore, the storage unit 23 includes a storage medium such as a RAM, a ROM, an HDD, and a flash memory. When the node control device 20 is realized by program execution processing, the storage unit 23 stores a program for realizing the function of the node control device 20.

  The interface unit 21 receives a request for establishing a new connection from an external device such as the node 30 or establishes a connection between the nodes 30. In addition, a link power-off instruction and a power-on instruction are transmitted to the node 30. Further, the OSPF route information advertised from the node 30 is received.

  The processing unit 22 controls the entire node control device 20, and here, mainly determines a path used for connection, and controls link power-off and power-on. Such a processing unit 22 includes a connection reception unit 221, a route calculation unit 222, a connection establishment unit 223, a link usage rate update unit 224, a cost calculation unit 225, a power control determination unit 227, and a link management unit. 228 and an OSPF route information receiving unit 229.

  The functions of the connection receiving unit 221, the route calculating unit 222, the cost calculating unit 225, and the power control determining unit 227 are the functions of the connection receiving unit 121, the route calculating unit 122, the cost calculating unit 125, and the power control determining unit 127 of the node 10, respectively. The explanation is omitted because it is the same as.

  The connection establishment unit (connection establishment control unit) 223 uses the link selected by the route calculation unit 222 as the path of this connection to provide route control information instructing to establish a connection from the start point node to the end point node. It transmits to the node 10 which has a link.

  The link usage rate updating unit 224 updates the link usage rate in the link usage rate information 232 (described later) based on the traffic volume of the connection established on each link of the network and the traffic volume of the disconnected connection. .

  The link management unit 228 writes the link power-off and power-on states indicated by the OSPF path information received by the OSPF path information receiving unit 229 in the link information 233. Further, the link management unit 228 sets the value of the link cost indicated in the OSPF route information in the topology information 231 (details will be described later). That is, the value of the link cost indicated in the OSPF route information transmitted from the node 30 is reflected in the topology information 231.

  The OSPF route information receiving unit 229 receives the OSPF route information transmitted from the node 30. This OSPF path information includes identification information of the link that is powered off or powered on at the node 30 and the link cost of the link. That is, the OSPF path information receiving unit 229 can receive the OSPF path information from the node 30 to know whether the link is surely powered off or powered on at the node 30. The OSPF route information received by the OSPF route information receiving unit 229 is output to the link management unit 228.

  The power control unit 24 instructs each node 30 to power off or power on the link based on an instruction from the power control determination unit 227. This instruction is instruction information indicating which link of the node is to be powered off or powered on.

  The storage unit 23 stores topology information 231, link usage rate information 232, and link information 233.

  The topology information 231 is the same as the topology information 131 of the node 10 described above, but the link cost value here is the value of the link cost of each link of the entire network.

  The link usage rate information 232 is the same as the link usage rate information 132 of the node 10 described above, but the link usage rate of each link of the entire network is described.

  The link information 233 (power off / on information) is the same as the link information 133 of the node 10 described above, but information on all the links that are powered off in the network is described.

  Next, the configuration of the node 30 will be described. The node 30 is a router that establishes a connection with another node 10 and performs link power-off and power-on based on control from the node control device 20. As shown in FIG. 8, the node 30 includes an interface unit 31 (31A, 31B,..., 31N), a processing unit 32, a storage unit 33, and a power control unit 34. The processing unit 32 is also realized by a program execution process by a CPU included in the node 30, a dedicated circuit, or the like. Furthermore, the storage unit 33 includes a storage medium such as a RAM, a ROM, an HDD, and a flash memory. When the node 30 is realized by program execution processing, the storage unit 33 stores a program for realizing the function of the node 30.

  Since the interface unit 31 is the same as the interface unit 11 of the node 10, the description thereof is omitted.

  The processing unit 32 establishes or disconnects a connection based on the path control information from the node control device 20. The processing unit 32 includes a connection establishment unit 323, an OSPF route information advertisement unit 325, a cost setting unit 326, and a link management unit 328.

  The connection establishment unit 323 establishes a connection with another node 30 based on the path control information from the node control device 20.

  When the link is powered off, the OSPF path information advertising unit 325 displays the OSPF path information indicating the identification information of the link that has been powered off and the link cost set for the link by the link management unit 328. 20 is transmitted. When the link is powered on, OSPF path information indicating the identification information of the link that has been powered on and the link cost of the link is transmitted to the node control device 20.

  The cost setting unit 326 and the link management unit 328 are the same as the cost setting unit 126 and the link management unit 128 of the node 10 and thus will not be described.

  The power control unit 34 powers off or powers off the link based on the control information from the node control device 20.

  The storage unit 33 stores topology information 331 and link information 333. The topology information 331 is the same information as the topology information 231 of the node control device 20. The link information 333 is information indicating identification information of a link that is powered off in its own node 30.

<Processing procedure>
Next, the processing procedure of the node control apparatus 20 will be described with reference to FIGS.

  First, the connection accepting unit 221 of the node control device 20 in FIG. 7 accepts a request for establishing a new connection from an external device or the like (S201). Next, the route calculation unit 222 calculates a show test path for this new connection (S202). That is, the route calculation unit 222 refers to the topology of the network indicated in the topology information 231 and the link cost of each link, and determines the route (link connection) from the start node to the end node of the connection that is the target of the establishment request. Then, the route that minimizes the total value of link costs is calculated.

  Next, similarly to S103 of FIG. 5 described above, the power control determination unit 227 has a link usage rate that is equal to or higher than a predetermined value (for example, “0.9”) among the links of the route calculated in S202 ( It is determined whether there is a high load link) (S203). Here, if there is no link whose link usage rate is equal to or higher than a predetermined value (for example, “0.9”) (No in S203), the connection establishment unit 223 makes this to the node 30 serving as the shortest path. The short test path route control information is transmitted (S209). That is, the connection establishment unit 223 instructs the node 30 (see FIG. 8) to establish a new connection on the shortest path. As a result, each node 30 establishes a connection on the shortest path and traffic flows.

  Thereafter, similarly to S108 of FIG. 5, the node control device 20 updates the link cost (S210). That is, the link usage rate updating unit 224 of the node control device 20 calculates the link usage rate of this link based on the traffic amount of the link obtained by adding the traffic amount of the new connection, and updates the link usage rate information 232. To do. Then, the cost calculation unit 225 calculates the link cost of the link that has established the new connection using the above-described equation (1) based on the link usage rate indicated in the updated link usage rate information 232. calculate. Next, the cost calculation unit 225 writes the calculated link cost value in the topology information 231.

  After the link cost is updated in S210, the power control determination unit 227 refers to the link usage rate information 232, and the link (0) of the link usage rate is included in the route (link) of this show test path. It is determined whether or not there is (S211). Here, if there is a link whose link usage rate is “0” (Yes in S211), the power control determination unit 227 instructs the node 30 having the link to power off the link via the power control unit 24. (S212). After that, when the node 30 powers off the link and receives OSPF path information indicating the power off of the link from the node 30 (S213), the link management unit 228 updates the link information 233, The cost is updated (S214). That is, the link management unit 228 writes in the link information 233 that the link is powered off. Further, the link management unit 228 updates the topology information 231 using the value of the link-off link cost indicated in the received OSPF path information. On the other hand, when it is determined that there is no link having a link usage rate of “0” in the shortest path (link) (No in S211), the process is terminated.

  In S203, when the power control determination unit 227 determines that there is a link (high load link) having a link usage rate of a predetermined value (for example, “0.9”) or more among the selected links ( In step S203, the power control determination unit 227 refers to the topology information 231 and the link information 233 in the same manner as in step S104 in FIG. 5, and is a link connected to the same connection destination (node 10) as the high-load link. It is determined whether there is a link that is powered off (S204). Here, if there is a link that is connected to the same connection destination (node 10) as this high-load link and is powered off (Yes in S204), the node 30 having that link via the power control unit 24. Is instructed to power on the link (S205). Thereafter, when the node 30 powers on the link and receives OSPF path information indicating the power-on of the link from the node 30 (S206), the link management unit 228 updates the link information 233 and the link cost. (S207). That is, the link management unit 228 writes in the link information 233 that the link is powered on. Further, the link management unit 228 updates the topology information 231 by using the value of the link-on link cost indicated in the received OSPF path information.

  The node control device 20 executes the process described above every time a request for establishing a new connection is received. Also in this case, every time the node control device 20 receives a connection disconnection request from the connection reception unit 221, the connection establishment unit 223 performs the connection disconnection process, and the link usage rate update unit 224 disconnects the connection. The traffic amount of the link is reduced, and the link usage rate information 232 is updated. Then, as in S211, the power control determination unit 227 refers to the updated link usage rate information 232, and if there is a link whose link usage rate is “0”, the power control determination unit 227 turns off the power of the link. The node 30 having this link is instructed via the power control unit 24.

  By controlling the node 30 in the network using such a node control device 20, it is possible to secure a stable communication environment while reducing the power consumption of the entire network as much as possible. In the second embodiment, the value of the link cost of the power-off link (power-on link) may be a value determined by the node 30 or a value determined by the node control device 20 side.

  According to the node 10 and the node control device 20 described above, for example, as shown in FIG. 10, when a plurality of links are used, the total load of the first link and the second link is relatively low (low load). Time), a new connection is established on one link (for example, the first link), and a new connection is not established on the second link. Therefore, the connection of the second link is naturally extinguished, and the power can be turned off when the traffic amount becomes “0”. Further, when the total load of the first link and the second link is relatively high (at the time of high load), the power-off link (for example, the second link) is powered on, and a new connection is also made to the second link. To establish. That is, as shown in the graph on the right side of FIG. 10, when the traffic volume is large, traffic is flowed using a plurality of links. When the traffic volume is small, the link is powered off as much as possible. Therefore, the power consumption of the entire network can be reduced compared to the case where a plurality of links are always used in the network.

In the node 10 and the node control device 20 described above, the formula used by the cost calculation units 125 and 225 for link cost calculation is not limited to the formula (1). In the case of formula (1), when the link usage rate is “0.5”, the link cost value is minimized. For example, the link usage rate value at which the link cost value is minimized is reduced to about In order to set “0.7”, the following equation (2) may be used.
Max (5 × 1 / (1-link usage rate), 10 × 1 / (link usage rate)) (2)

  That is, the value of the link usage rate that minimizes the link cost value can be changed by arranging the coefficients 1 / (1-link usage rate) and 1 / (link usage rate).

The calculation formula used by the cost calculation unit 225 is as follows: (1) When the link usage rate is equal to or lower than a predetermined link usage rate (for example, “0.5” or “0.7”), the link usage rate is The higher the link usage rate, the smaller the link cost value is calculated. (2) When the link usage rate exceeds the predetermined link usage rate, the higher the link usage rate, the higher the link cost value is calculated. If it is a calculation formula, it is not limited to Formula (1) or Formula (2). For example, the following equation (3) may be used.
y = a | x-A | + B (3)
y: Link cost value x: Link usage rate a: Predetermined coefficient A: The higher the link usage rate, the higher the link usage rate, the lower the link cost value, the lower the link cost value B: The lowest link cost Value If this equation (3) is shown in a graph, the graph on the right side of FIG. 11 is obtained.

  Note that the node 30 may be a node connected by a multi-link as illustrated in FIG. 12A, or may be a node having a multi-route as illustrated in FIG. That is, a node that establishes connections on a plurality of routes (routes) from a certain start point node to an end point node and flows traffic to both of these routes may be used. The link cost calculation described above and the short test path routing based on the calculated link cost may be applied to the routing of a node that performs connectionless communication such as IP (Internet Protocol). For example, as shown in FIG. 12C, consider a case where the network is configured with nodes 40 (40A, 40B, 40C, 40D, and 40E) that perform connectionless communication. In this case, it is assumed that routes A and B having a low cost (total value of link costs) are selected from a plurality of routes in order to transfer a packet from the node 40A to the node 40E. When the cost of the route A is “100” and the cost of the route B is “50”, the node 40A allocates the packet to be transferred by the reciprocal of the cost of each route. That is, in the example shown in FIG. 12A, “Route A traffic volume: Route B traffic volume” is approximately “1: 2”. In the node 40 (40A) that performs such packet allocation, the link cost value decreases as the link usage rate increases (that is, as the traffic volume increases), similar to the cost calculation units 125 and 225 described above. As the link usage rate decreases (that is, the traffic volume decreases), the value of the link cost is increased. By doing so, the traffic to the route B continues to increase, and the traffic amount of the route A continues to decrease. That is, the traffic is biased toward route B. Finally, when the traffic volume of the route A falls below a predetermined threshold (for example, when the traffic volume becomes “0”), the link of the route A is powered off. In this way, power consumption can be reduced even in a network that performs connectionless communication.

  The node 10 and the node control device 20 according to the present embodiment can be realized by a program for executing the processing as described above, and the program is stored in a computer-readable recording medium (CD-ROM or the like). Can be provided.

10, 30 (30A, 30B, ..., 30N), 40 (40A to 40E) Nodes 11, 21, 31 Interface unit 12, 22, 32 Processing unit 13, 23, 33 Storage unit 14, 24, 34 Power control unit 20 Node control device 121, 221 Connection accepting unit 122, 222 Route calculating unit 123, 223, 323 Connection establishing unit 124, 224 Link usage rate updating unit 125, 225, 325 Cost calculating unit 126, 326 Cost setting unit 127, 227 Power control Judgment part 128,228,328 Link management part 131,231,331 Topology information 132,232 Link utilization information 133,233,333 Link information 229 OSPF path information receiving part 325 OSPF path information advertising part

Claims (9)

A node that determines a route in a network by show test path routing,
Stores topology information indicating the topology of the network and the link cost of each link in the network, and link usage rate information indicating the maximum available bandwidth and link usage rate of each link connected to the node. A storage unit to
A connection accepting unit that accepts the connection establishment request and the disconnection request;
A path calculation unit that refers to the topology indicated in the topology information and the link cost of each link, and determines a link to be used in a show test path of the connection of the target of the establishment request;
A connection establishment processing unit that establishes a connection on the determined link and disconnects the target connection of the disconnection request;
A link usage rate updating unit that updates a link usage rate in the link usage rate information based on the traffic volume of the established connection and the traffic volume of the disconnected connection;
Based on the updated link usage rate information, the link cost of the link is calculated as a lower value as the link usage rate of the link is higher, with the predetermined link usage rate as the upper limit value, and using this calculated value A cost calculation unit for updating the link cost of the link in the topology information;
A power control unit for controlling power off or power on of a link connected to the node;
Power control for instructing the power control unit to power off the link when it is determined that the link usage rate of the link connected to its own node is 0 with reference to the updated link usage rate information A determination unit,
When the connection accepting unit accepts a request for establishing a new connection, the route calculating unit refers to the link cost of each link indicated in the updated topology information, and A node characterized by calculating a link used in a show test pass. Link information indicating identification information of the power-off link is stored in the storage unit, and the link cost value of the power-off link in the topology information corresponds to a predetermined link usage rate. A link management unit for setting a predetermined value smaller than the link cost value;
As a result of calculating the link used in the show test path of the connection of the establishment request target by the route calculation unit, the link usage rate in the link usage rate information is equal to or higher than the predetermined link usage rate for the link used in the show test path. If you have a link,
The power control determination unit refers to the link information and the topology information, and a link whose link usage rate is equal to or higher than the predetermined link usage rate and a connection destination is the same link and is a power-off link When it is determined that there is, the power control unit is instructed to power on the link,
The path calculation unit determines the link to be used in the shortest test path of the connection of the establishment request again, including the link that has been powered on,
The connection establishment processing unit establishes a connection on the determined link;
The link usage rate update unit updates the link usage rate information based on the traffic volume of the established connection,
The cost calculation unit calculates the link cost of the link as a higher value as the link usage rate of the link is higher based on the updated link usage rate information, and uses the calculated value to calculate the topology information. Update the link cost of the link at
When the connection accepting unit accepts a request for establishing a new connection, the route calculating unit refers to the link cost of each link indicated in the updated topology information, and The node according to claim 1, wherein a link used in the show test path is determined.   A network comprising a plurality of nodes according to claim 1 or 2. A node control device for controlling a node in a network,
Stores topology information indicating the topology of the network and the link cost of each link in the network, and link usage rate information indicating the maximum available bandwidth and link usage rate of each link connected to the node. A storage unit to
A connection accepting unit that accepts the connection establishment request and the disconnection request;
A path calculation unit that refers to the topology indicated in the topology information and the link cost of each link, and determines a link to be used in a show test path of the connection of the target of the establishment request;
A connection establishment control unit for transmitting an instruction to establish a connection on the determined link and to disconnect the target connection of the disconnection request to the node;
A link usage rate updating unit that updates a link usage rate in the link usage rate information based on the traffic volume of the established connection and the traffic volume of the disconnected connection;
Based on the updated link usage rate information, the link cost of the link is calculated as a lower value as the link usage rate of the link is higher, with the predetermined link usage rate as the upper limit value, and using this calculated value A cost calculation unit for updating the link cost of the link in the topology information;
A power control unit for transmitting to the node a power-off or power-on instruction of a link connected to the node;
As a result of calculation of the link usage rate by the link usage rate update unit, when the link usage rate becomes 0, the power control unit is provided with a power control determination unit that instructs to power off the link. Node control device. An OSPF path information receiving unit that receives, from the node, the identification information of the link that has been powered off and information indicating the value of the link cost of the link by OSPF path information;
The link information indicating identification information of the link turned off shown in the received OSPF path information is stored in the storage unit, and the value of the link cost of the link turned off in the topology information is as follows: A link management unit for setting a value smaller than the value of the link cost corresponding to the predetermined link usage rate;
As a result of determining the link used in the shortest test path of the connection of the establishment request target by the path calculation unit, the link usage rate in the link usage rate information is equal to or higher than a predetermined link usage rate for the link used in the shortest test path If there is a link,
The power control determination unit refers to the link information and the topology information, and a link whose link usage rate is equal to or higher than the predetermined link usage rate and a connection destination is the same link and is a power off link When it is determined that there is, the power control unit is instructed to power on the link,
The path calculation unit determines the link to be used in the shortest test path of the connection of the establishment request again, including the link that has been powered on,
The connection establishment control unit transmits an instruction to establish a connection on the determined link to the node;
The link usage rate update unit updates the link usage rate information based on the traffic volume of the established connection,
The cost calculation unit calculates the link cost of the link as a higher value as the link usage rate of the link is higher based on the updated link usage rate information, and uses the calculated value to calculate the topology information. Update the link cost of the link at
When the connection accepting unit accepts a request for establishing a new connection, the route calculating unit refers to the link cost of each link indicated in the updated topology information, and 5. The node control device according to claim 4, wherein a link used in the show test path is determined. Based on an instruction from the node control device according to claim 4 or 5, the node performs link power off and power off,
A node characterized in that when the link is powered off, OSPF path information in which identification information of the link and a link cost value of the link are set is advertised to the node control device. The topology information indicating the topology of the network and the link cost value of each link in the network, and the link usage rate information indicating the maximum available bandwidth and the link usage rate of each link connected to the node are stored. A node comprising a storage unit
Receiving a connection establishment request and a disconnection request;
Determining a link to be used in a shortest test path of the connection of the establishment request with reference to the topology indicated in the topology information and the link cost of each link;
Establishing a connection on the determined link and disconnecting the target connection of the disconnection request;
Updating the link usage rate in the link usage rate information based on the traffic volume of the established connection or the traffic volume of the disconnected connection;
Based on the updated link usage rate information, the link cost of the link is calculated as a lower value as the link usage rate of the link is higher, with the predetermined link usage rate as the upper limit value, and using this calculated value Updating the link cost of the link in the topology information;
As a result of updating the link usage rate, when the link usage rate of a link connected to its own node becomes 0, instructing a node including the link to power off the link;
When the connection accepting unit accepts a request for establishing a new connection, the link used in the shortest test path of the connection of the establishment request is referred to by referring to the link cost of each link indicated in the updated topology information And a step of determining the link control method. The node is
Storing link information indicating identification information of the link that has been powered off in the storage unit;
Setting a predetermined value smaller than the link cost value corresponding to the predetermined link usage rate as the link cost value of the link that has been powered off in the topology information;
As a result of determining a link to be used in the shortest test path of the connection of the establishment request target, when there is a link whose link usage rate in the link usage rate information is equal to or higher than a predetermined link usage rate,
With reference to the link information and the topology information, when it is determined that there is a link in which the link usage rate is equal to or higher than the predetermined link usage rate and the connection destination is the same link and is powered off, Instructing a node comprising the link to power on the link;
Determining the link to be used in the shortest test path of the connection for which the establishment request is made, including the power-on link;
Establishing a connection on the determined link;
Updating the link usage rate information based on the traffic volume of the established connection;
Based on the updated link usage rate information, the link cost of the link is calculated as a higher value as the link usage rate of the link is higher, and using this calculated value, the link cost of the link in the topology information is calculated. A step of updating
Determining a link to be used in a shortest test path of the connection of the target of the establishment request with reference to a link cost of each link indicated in the updated topology information when a new connection establishment request is received; The link control method according to claim 7, wherein the link control method is executed.   A program for causing a computer to execute the link control method according to claim 7 or 8.

Images