JP6637911B2 - Network design apparatus, network design method, and network design processing program - Google Patents

Description

  An embodiment of the present invention relates to a network design device, a network design method, and a network design processing program.

Wide area Ethernet (registered trademark) including a layer 2 switch and a transmission device has been widely adopted as a carrier network.
In the wide-area Ethernet, by applying ERP (Ethernet (registered trademark) Ring Protocol) to a ring topology network in which nodes are physically connected in a ring, reliability is improved due to redundant paths and equipment is economical. (Non-Patent Document 1).

  The arrangement of the network device and the configuration of the equipment in the device change depending on how the route is designed. The type and price of network equipment and equipment in the equipment are determined according to the amount of traffic that must be processed. If the traffic volume to be processed is large, the traffic capacity of the network device and the equipment must be large, and the above-mentioned price increases. Therefore, it is considered that the total cost can be reduced by designing a route that efficiently uses the capacity of the equipment (interface) in consideration of the capacity of the device or equipment to be accommodated.

For example, as a first example, in a network topology in which nodes v1 to v6 shown in FIG. 9 are connected in a ring, a line “1” having a traffic demand of 10, which is a line communicating from v1 to v4, and a traffic demand of 10 It is assumed that 30 lines “2” and 100 lines with traffic demand “3” are accommodated.
Also, it is assumed that the capacity of an interface, which is equipment of a link unit (sometimes referred to as a link) of a network device arranged in each node, is standardized, and the types of the capacities are 40 and 100. There are two routes, a route “1” passing through v1, v2, v3, and v4, and a route “2” passing through v1, v6, v5, and v4. When accommodating based on the traffic demand of each line as in the related art, the lines "1" and "2" are accommodated in the route "1", and the line "3" is accommodated in the route "2". That is, since the link unit on the route “1” processes a total of 40 traffics, the capacity of the interface is 40.
Further, since the link unit on the route “2” processes a total of 100 traffic, the capacity of the interface is 100. Here, it can be said that the total cost is kept to a minimum because the capacity of the interface can be used efficiently.

  On the other hand, in recent years, for realization of 5G (Non-Patent Document 2), various requirements such as a very large capacity, a large number of connections, and a very low delay have been mentioned, and statistics for accommodating these lines with different restrictions in a common network have been proposed. It has been advocated for the economy of multiple effects. In the network required by Non-Patent Document 2, it is expected that a plurality of lines having various restrictions are accommodated in a path designed using not only traffic demand but also QoS such as delay time as a metric. In addition, a plurality of routes are set between transmitting and receiving nodes, and a communication system in which a route to be accommodated is selected according to a policy of each constraint has been widely developed (Non-Patent Document 3).

Here, when accommodating lines having various requirements, it is necessary to design a path to be accommodated in consideration of not only traffic demand but also line restrictions.
For example, as a second example, as shown in FIG. 9, a line “4” having a delay constraint value of 3 and a traffic demand of 10, which is a line communicating from v1 to v4, and a delay constraint value of 1 It is assumed that a line “5” having a traffic demand of 100 and a line “6” having a delay constraint value of 1 and a traffic demand of 30 are accommodated. Similarly, the route “1” or the route “2” is accommodated. Here, it is assumed that the delay time of the route “1” is 3 and the delay time of the route “2” is 1. If the lines are accommodated based on the traffic demand of each line as in the related art, the lines "4" and "6" are accommodated in the route "1", and the line "5" is accommodated in the route "2".
However, considering the delay constraint value, the line “5” and the line “6” must be accommodated in the path “2” that satisfies the delay constraint value. That is, the link unit on the route “1” processes a total of 10 traffics, so the interface capacity is 40, and the link unit on the route “2” processes a total of 130 traffics, so the interface capacity is 140. As described above, in the second example, the capacity of the interface cannot be used efficiently compared to the first example.

  Therefore, in consideration of not only traffic demand but also other requirements, the route to be accommodated changes and the cost changes accordingly, so when designing a route to accommodate a line having multiple requirements, Considering the arrangement of the network devices and the configuration of the equipment in the apparatus, it is considered that the optimum network configuration can be derived so that the equipment can be used more efficiently and the total cost can be minimized.

Recommendation ITU-T G.8032 / Y.1344 “Ethernet Ring Protection Switching”, 08/2015 (7 Ring protection characteristics) DOCOMO 5G White Paper Requirements and Technology Concept for 5G Wireless Access after 2020, NTT DOCOMO, Inc., September 2014 Manayya KB, “Constraint Shortest Path First”, https://www.ietf.org/archive/id/draft-manayya-constrained-shortest-path-first-02.txt, 02/2010 Jin Y. Yen, “Finding the K Shortest Loopless Paths in a Network”, Management Science, vol. 17, No. 11, pp. 712-716, 1971

  As described above, a more optimal network design can be realized by considering the route design for accommodating the line and the configuration design of the network device and the equipment in the device. However, in the route design method for accommodating a line proposed so far, a network device and a configuration in the device are determined in advance, and the line is designed on a determined network. In the network equipment configuration design method proposed so far, the amount of traffic and power consumption are considered for each network device for specific indices, and the conditions required from the network such as the delay time due to each user's request are considered. Not considered in line accommodation.

  An object of the present invention is to provide a network design apparatus, a network design method, and a network design processing program capable of designing an optimal network configuration in an arrangement of arbitrary nodes and links, accommodation lines, and network facilities. is there.

  In order to achieve the above object, a first aspect of a network design device according to an embodiment of the present invention includes a plurality of nodes and a plurality of link units connecting between network devices arranged in the nodes. In the network, (1) a route in which all the lines are accommodated, (2) the presence or absence of a network device to be arranged in each of the nodes, and (3) all of the link units in all of the network devices. A network design apparatus for determining a type of an interface to be installed, (1) information on a connection state between the nodes, (2) information on a plurality of lines accommodated in the network, and (3) An input unit for inputting information on equipment of a network device and a link unit arranged in the node; and A first calculating unit that calculates candidates for disjoint routes accommodating the line and a set of link units used by the candidates based on input information that is input to the link, and (1) configuring the link unit Candidate combinations of disjoint interfaces, (2) the number of interfaces included in the combination, (3) the total capacity of the interfaces included in the combination, and (4) the total equipment of the interfaces included in the combination. A first processing unit having a second calculation unit for calculating a set of cost values; (1) the input information; and (2) a route obtained by the first calculation unit and a second calculation unit obtained by the second calculation unit. Collection of candidates for the combination of the route and the interface such that the line satisfies the constraint conditions required from the network, based on the candidate for the combination with the interface. The calculated, in the set of the obtained candidate, such as the total installed cost value in the entire network is minimized, to provide a device and a second processing unit that derives a combination of candidates of the optimum path and interface.

  In a second aspect of the network design apparatus having the above configuration, in the first aspect, based on the candidates determined by the second processing unit, (1) a route for each of the lines is determined; A node for arranging the network device is determined under a condition that a network device in which none of the interfaces is installed in all link units in the device is not arranged in the node. (3) An interface to be arranged in a link unit in the network device Is provided for each of the link units, and an output unit for outputting the obtained result is further provided.

  In order to achieve the above object, a first aspect of the network design method according to the embodiment of the present invention is as follows. (1) All the lines are accommodated in the network of the first aspect of the network design apparatus having the above configuration. A network design method for determining a route, (2) presence / absence of a network device arranged in each of the nodes, and types of interfaces respectively installed in all the link units in all the network devices. There, based on the input information, by using the k-shortest path algorithm, to accommodate the line, calculate a set of disjoint routes candidate, by listing all combinations of the interface, A set of candidates for a combination of disjoint interfaces constituting the link unit is calculated, and the input information Based on the calculated set of route candidates, and the calculated set of interface combination candidates, a candidate set of route and interface combinations, such that the line satisfies the constraints required from the network, In this set of candidates, solve as an optimization problem and find the optimal combination of route and interface that is the optimal solution, so that the combination of route and interface that minimizes the equipment cost value in the entire network is obtained. A method is provided for deriving candidates.

  The second aspect of the network design method described above provides the method according to the first aspect, wherein the optimization problem is a problem formulated as a 0-1 linear programming problem.

  In order to achieve the above object, an aspect of a network design processing program according to an embodiment of the present invention is a program used in a computer that operates as a part of the first aspect or the second aspect of a network design device, A program for causing a computer to function as the input unit, the first processing unit, and the second processing unit is provided.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to design an optimal network configuration in arrangement of arbitrary nodes and links, accommodation lines, and network facilities.

FIG. 1 is a diagram illustrating an example of a network design apparatus according to an embodiment of the present invention. FIG. 4 is a diagram showing a flow of network design by the network design device according to the embodiment of the present invention. The figure which showed the example of the topology. The figure which showed the legend of the figure of the example of a topology. The figure which showed the legend of the example of the equipment which accommodates a network. The figure which showed an example of the network structure. FIG. 3 is a diagram illustrating an example of a network configuration when a link has no physical cable. FIG. 1 is a diagram illustrating an optimal network configuration according to an embodiment of the present invention. The figure which showed the example of the network topology connected in a ring.

  According to the present invention, a network configuration pattern representing a combination of a path in which all lines are accommodated and interfaces arranged in all link units in the apparatus is used as a decision variable, and a requirement for a line to be required for a network is restricted. The optimal network configuration is designed by solving the combination optimization problem that minimizes the objective function value with the total facility cost as the objective function value.

  Specifically, a network design device according to the present invention is configured such that, in a network including a plurality of nodes and a plurality of link units connecting the network devices arranged in the nodes, an input unit, a first processing unit, A second processing unit.

  The input unit includes: (1) network topology information (network) including a connection matrix indicating a connection state between nodes in the network, a delay time in the link unit, and a link cost calculated using the delay time as a metric. (2) information on the starting and ending points of communication among the nodes, that is, a pair of nodes serving as end points of the line, traffic demand of the line, and delay time of the line Information on a plurality of lines, including the upper limit value of (i), (3) the type of interface constituting a link unit in the network device arranged in the node, the traffic capacity of the interface, and the price of the interface as metrics. The calculated equipment cost value and the Including the upper limit of the number of volumes of interfaces, inputs the information on a network device and a link portion of the equipment placed in the node, the input information including.

The first processing unit has a first calculation unit and a second calculation unit.
The first calculation unit includes: (1) a plurality of nodes accommodating the line and a link unit based on the connection matrix as the input information, the pair of nodes as the input information, and the link cost as the input information. Then, a set of disjoint route candidates in which the shortest route according to the link cost is determined in ascending order by an arbitrary number, and (2) a set of used links for each route are calculated.

  The second calculation unit is configured to: (1) a set of candidates for disjoint interface combinations obtained by determining an arbitrary number of interface combinations including a plurality of interfaces constituting the link unit from the information on the equipment as the input data; (2) the number of interfaces included in the interface combination, (3) the total traffic capacity of the interfaces included in the interface combination, and (4) the total equipment cost value of the interfaces included in the interface combination. Is calculated for each of the above combinations.

  A second processing unit configured to request the network from the network based on the input information and the path obtained by the first processing unit and a candidate set of interface combinations; (1) In the above-mentioned link section, a constraint condition regarding traffic capacity that the total amount of traffic flowing on each link section does not exceed the sum of the traffic capacity of the interfaces installed in each link section, and (2) each The constraint condition concerning the delay time that the delay time of the route accommodating the line does not exceed the upper limit value of the delay time of each line, and (3) the number of interfaces accommodated in the network device arranged in each node in the node is the above Restriction on equipment that does not exceed the upper limit of the number of interfaces , To set the.

  The second processing unit restricts the candidate set of the combination of the route and the interface obtained by the first processing unit to the candidate set of the combination of the route and the interface that satisfies all of the above-mentioned constraints. In the candidate set, a candidate of a combination of a route and an interface that derives the minimum total equipment cost value in the entire network is derived.

  The output unit outputs, from the optimal candidates obtained by the second processing unit, a route for each line, and a network device obtained under the condition that no network device in which no interface is installed in all the link units in the network device is arranged in a node. And outputs the combination of the arrangement node and the interface for each link unit in each network device.

  In addition, the network design method according to the present invention provides a first procedure, a second procedure, and a second procedure in a network including a plurality of nodes and a plurality of link units connecting between network devices arranged in the nodes. There are three procedures.

  In the first procedure, a link comprising a plurality of nodes and a link unit accommodating the line is obtained from the connection matrix as the input information, the node pair as the input information, and the link cost as the input information. The k-shortest path is derived as a means for calculating a set of disjoint path candidates and use links for each candidate, which are obtained in the order of the shortest path according to the cost in the order of the shortest number.

  In the second procedure, from the information on the equipment, which is the input information, a plurality of interfaces constituting the link unit, a combination of interfaces, and any number of disjoint interface combination candidates obtained by an arbitrary number; Means for calculating the number of included interfaces for each combination of the interfaces, the total traffic capacity of the included interfaces for each combination of the interfaces, and the total equipment cost value of the included interfaces for each combination of the interfaces Enumerate all interface combinations.

  In the third procedure, the network is requested from the network based on the input information, the route obtained in the first and second procedures, and the candidate set of the interface combination. As conditions relating to the equipment, (1) in the above-mentioned link unit, a constraint condition regarding traffic capacity that the total amount of traffic flowing on each link unit does not exceed the sum of the traffic capacity of the interfaces installed in each link unit, and (2) In the above-mentioned line, the delay time of the route accommodating each line does not exceed the upper limit of the above-mentioned delay time of each line. Said that the number of accommodated interfaces does not exceed the upper limit of the number of accommodated interfaces. Constraints related to equipment, to set the.

  In the third procedure, from the candidate set of the combination of the route and the interface obtained by the first processing unit, the combination is limited to the combination of the route and the interface that satisfies all the above-mentioned constraints. As a means for deriving a candidate of a combination of a route and an interface that minimizes the total equipment cost value in the entire network, an optimal combination of a route and an interface that is an optimal solution is obtained as an optimization problem.

  In the third procedure in the network design method according to the present invention, in each of the lines, a candidate for each path is selected from a set of candidate combinations of the path and the interface obtained in the first and second procedures. A variable indicating whether or not a candidate for a combination of each interface is selected in a link portion in all network devices as a decision variable to which an integer 0 or 1 is set, By formulating as a 0-1 linear programming problem, it is possible to apply the LP relaxation and the branch and bound method, efficiently solve the subproblems by dividing them into subproblems, and efficiently derive the optimal combinations of paths and interfaces I do.

  That is, in the third procedure of the network design method according to the present invention, the LP mitigation and the branch-and-bound method can be applied by expressing the decision variables as binary variables, and even a problem in a large-scale network is divided into partial problems. By solving and pruning unnecessary partial problems, a solution can be efficiently derived.

  In the third procedure in the network design method according to the present invention, the route selected based on the variables related to the route candidates and the set of use links for each route obtained in the second procedure is determined. A set of link parts to be used is obtained, and based on the set of link parts and the traffic demand of each line as the input information, a traffic amount of the corresponding link part is obtained, and a total amount of traffic flowing on an arbitrary link part is calculated. Is expressed as the sum of the traffic volume on the line. Further, in the third procedure, the traffic capacity of an arbitrary link part is determined from the variables related to the candidates for the interface combination and the total traffic capacity of the interfaces included in the combination obtained in the second procedure. , It is possible to linearly express the above-mentioned constraint condition regarding the traffic volume.

  In the third procedure of the network design method according to the present invention, the selected route is used based on the variables relating to the route candidates and the set of used links for each route obtained in the second procedure. A set of link parts is obtained, and a delay time of each link part of the corresponding line is obtained based on the set of link parts and the delay time of each link part as the input information, and a path for accommodating an arbitrary line is determined. Since the delay time is expressed as the sum of the delay times in all the link units in the line, it is possible to linearly express the constraint condition regarding the delay time.

  Further, in the third procedure in the network design method according to the present invention, the number of interfaces of the selected combination is determined based on the variables related to the interface combination and the number of interfaces included in the combination obtained in the second procedure. Based on the number of interfaces and the row of each node of the connection matrix as the input information, the number of interfaces of each link unit connected to the corresponding node is determined, and the accommodation interface in the network device to be arranged at an arbitrary node Since the number is represented as the sum of the number of accommodated interfaces in all the link units connected to the corresponding node, it is possible to linearly represent the condition regarding the above facilities.

  Further, in the third procedure in the network design method according to the present invention, the interface combination included in each link unit and the interface included in the combination obtained in the second procedure are indicated by variables related to the interface combination. From the total cost value of the above, the total cost value of the selected combination is obtained. From the total cost value and the row of each node of the connection matrix that is the input information, the total cost value of each link portion connected to each node is calculated. The sum of the total cost value of the link unit and the sum of all the link units connected to each node is obtained, and the total equipment cost value is expressed as the sum of the sum of all the link units and the sum of all the nodes. It is possible to linearly express the objective function value of the programming problem.

  That is, in the network design method according to the present invention, since all equations representing the constraint conditions and the objective function are linearly represented, it is possible to derive an exact solution using the interior point method.

Therefore, the present invention provides a network design method for realizing a combination of an optimum route, an arrangement of network devices, and an interface in a network device, which realizes accommodation of a plurality of lines having different network requirements at a minimum equipment cost. Can be.
That is, according to the present invention, it is possible to design an optimal network configuration in the arrangement of arbitrary nodes and link units, accommodation lines, and network facilities.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
The embodiment described below is an example of the embodiment of the present invention, and the present invention is not limited to the following embodiment.
These embodiments are merely examples, and the present invention can be implemented in various modified and improved forms based on the knowledge of those skilled in the art. In the specification and the drawings, components having the same reference numerals indicate the same components.

Next, a network design apparatus for realizing the present invention will be described. FIG. 1 is a diagram illustrating an example of a network design apparatus according to an embodiment of the present invention.
As shown in FIG. 1, the network design device 10 includes an input unit 11, a first processing unit 12, a second processing unit 13, and an output unit 14.
The user of the network design apparatus 10 has (1) information on a connection state between node apparatuses, (2) information on a line accommodated in a network, (3) information on equipment (interface) of a network apparatus and a link unit. Is input to the input unit 11 of the network design device 10.
The network design device 10 provides input information from the input unit 11 to the first processing unit 12. The first processing unit 12 calculates, for each line, a first calculation unit 12a that calculates a candidate set of disjoint routes accommodating the line, and for each link unit, a candidate for a combination of disjoint interfaces arranged in the link unit. A second calculating unit 12b for calculating a set.
The first calculation unit 12a calculates a set of route candidates for each line and link information used by the candidates based on the input information.
Based on the input information, the second calculator 12b calculates (1) an interface combination candidate set for each link unit, (2) the number of interfaces included in the candidate, and (3) the total number of interfaces included in the candidate. The capacity and (4) the total equipment cost value of the interface included in the candidate are calculated.
The network design device 10 gives a candidate set and information on candidates from the first processing unit 12 to the second processing unit 13.

  The second processing unit 13 includes a calculation unit 13a that calculates a combination of an optimum route and an interface of a link unit in the optimum network device. The calculating unit 13a calculates an optimal combination of an optimal route for each line and an optimal interface for each link unit based on the information provided from the first processing unit 12. The network design device 10 gives the information of the optimum candidate from the second processing unit 13 to the output unit 14.

  The output unit 14 is based on the information obtained from the second processing unit 13, (1) route information for each line, (2) node information for arranging network devices, and (3) link unit of each network device. And output the interface information included in. The network design device 10 provides output information to the user from the output unit 14.

4 shows an example of a flow of a network design method using the network design device for realizing the present disclosure.
FIG. 2 shows a flow example of the network design method.
The network design method includes steps S1, S2, and S3. Step S1 uses the k-shortest path algorithm based on the input information to list, for each line, candidate routes that are disjoint with respect to the link unit and the node to be used. It is implemented in the first calculation unit 12a of the processing unit 12.
In step S2, based on the input information, for each link unit of the network device, a list of interface combination candidates that are relatively prime to the traffic capacity is listed, and the second calculation unit of the first processing unit 12 12b.
In step S3, based on the input information, the candidate set obtained in steps S1 and S2, and information on each candidate, each candidate is determined as a decision variable, a line request is used as a constraint, and the total equipment cost value is used as an objective function. The method is characterized in that an optimization problem to be minimized is solved as a value, and is implemented in the second processing unit 13.

An example for realizing the present disclosure will be described for each processing unit of the network design device 10.
First, an embodiment of the input unit 11 will be described. Examples of information provided by the user of the network design device 10 to the input unit in the present embodiment include (1) an example of information on a connection state between network devices arranged in a node, and (2) an example of information on lines accommodated in a network. , (3) shows an example of information on network devices and link unit equipment arranged in a node.

  FIGS. 3 and 4 show examples of information relating to the connection state between network devices arranged in the nodes. The information includes (1) a connection matrix representing the connection state between nodes in the network, (2) a delay time in the link, (3) Includes link cost calculated using delay time as a metric.

  FIG. 4 shows a legend used for the example topology (FIG. 3). The node “1” in FIG. 4 indicates that the node whose node number is “1”. The link “1” in FIG. 4 indicates that the link has the link number 1 and is connected to the node “1”.

  FIG. 3, which is a topology for realizing the present disclosure using the above legend, shows that nodes “1” to “4” are connected via links “1” to “5”. Represents

  The connection matrix M of the topology shown in FIG. 3 is shown in the following equation (A).

  In the connection matrix M, each row corresponds to a node, each column corresponds to a link, and “1” is stored in a portion where the link is connected to the node, and “0” is stored in a portion where the link is not connected. . Table 1 below shows an example of the delay time in each link.

  Next, an example of information on lines accommodated in the network is shown in Table 2 below.

  In Table 2, as information on the lines accommodated in the network, for each line, (1) a pair of nodes serving as end points of the line, (2) traffic demand of the line, and (3) an upper limit value of the delay time of the line. Set. For example, the line “1” having the line number “1” performs communication of the traffic amount “1” between the node “1” and the node “2”, and the delay time between the node “1” and the node “2” is Indicates that it must be less than or equal to "3".

  Next, FIG. 5 and Table 3 below show an example of information relating to the network device and the equipment of the link unit arranged in the node.

  FIG. 5 shows a legend of the network device. One network device is arranged at an arbitrary node. The network device in FIG. 5 includes a switch housing unit “1”, a slot “1-1”, a slot “1-2”, and a slot “1-3”. The switch housing unit “1” has a function of determining a slot (corresponding to a link) for outputting data based on the destination node indicated in the data received by the network device in FIG. A slot corresponds to a connection between a node and a link. The slot accommodates an interface constituting a link unit in the network device.

  FIG. 6 shows an example in which network devices are arranged in the network shown in FIG. 3 in order to show a connection method between network devices. In this example, network devices are installed in nodes "1" to "4", and in each node, the slots of the switches are connected by cables via links.

  When there is no cable directly connecting the node “1” and the node “3”, as shown in FIG. 7, the actual cable is the link “1”, the node “2”, and the link “2”. , But this cable does not pass through the switch installed at this node “2”.

Table 3 above shows legends of examples of facilities of network devices and link units arranged in the nodes as examples of facilities accommodated in the network.
Here, for the various interfaces “1” to “3”, (1) the traffic capacity of the interface, (2) the equipment cost value calculated by using the price of the interface as a metric, and (3) the number of interfaces accommodated in the network device Set the upper limit of. As for the switch housing, (1) the traffic capacity of the switch housing, (2) the equipment cost value obtained by calculating the price of the switch housing as a metric, and (3) the housing interface of the switch housing. Set the upper limit of the number.

  Here, the switch housing has a capacity of 100 Gbit / s per slot of the interface, and can accommodate three slots per switch housing. In this case, the traffic volume that can be processed by the entire switch casing is 100 Gbit / s * 3 = 300 Gbit / s (see Table 3), and the cost value of the switch casing is “5” (see Table 3). ).

  Further, as shown in Table 3, for example, the interface “1” whose interface number is “1” can process 10 Gbit / s of traffic, can install one interface in one slot, and has a cost value of Is 3.62.

Next, an example of the first processing unit 12 will be described.
In step S1 implemented in the first calculation unit 12a of the first processing unit 12 in the present embodiment, a route candidate (a set of route candidates) for each line based on the information regarding the connection state and the information regarding the line. And a set of use links of this candidate.

  Here, the first processing unit 12 uses the k-shortest path algorithm (Non-Patent Document 4) to set k = 3 from the node pairs in Table 2, the delay times in Table 1, and the connection matrix of Expression A. A description will be given of obtaining route candidates up to the third route. Table 4 below shows an example of a link used for each route candidate of each line as an example of a route complement set of lines accommodated in the network.

  The k-shortest path algorithm is an algorithm that, when given a weighted graph G, a start point s, and an end point t, searches for k paths in a path that does not include a loop from s to t in ascending order of cost. As shown in Table 2, the line “1” is a line for performing communication between the node “1” and the node “2”.

Based on the delay time of each link shown in Table 1, the first processing unit 12 determines the shortest path candidate “1-1” as the path between the node “1” and the node “2”. , "1-2" and "1-3". As a result of this search, the route candidate “1-1” is a route that passes through the node “1”, the link “1”, and the node “2”. The route candidate “1-2” is a route passing through the node “1”, the link “4”, the node “4”, the link “3”, the node “3”, the link “2”, and the node “2”. is there. The route candidate “1-3” is a route that passes through the node “1”, the link “5”, the node “3”, the link “2”, and the node “2”.
Therefore, a set of links used by the route candidate “1-1” is the link “1”. A set of links used by the route candidate “1-2” is the link “4”, the link “3”, and the link “2”. A set of links used by the route candidate “1-3” is the link “5” and the link “2”.

  From the delay time of each link shown in Table 1, the delay time of the route candidate “1-1” is “1”, the delay time of the route candidate “1-2” is “5” (= 2 + 1 + 2), and the route candidate is The delay time of “1-3” is “5” (= 3 + 2), and is set as a path candidate with a smaller number in ascending order of delay time. The route candidates for other lines are calculated in the same manner.

  Next, in step S2 implemented in the second calculation unit 12b of the first processing unit 12 in the present embodiment, based on the input information about the facilities, the combination of disjoint interfaces arranged in the link unit is determined. A set of candidates is calculated for each link.

  (1) Interface combinations for each interface combination candidate number, and (2) interfaces included in the combinations, obtained from the interface types, traffic capacities, and equipment cost values in Table 3, which are information on network devices. And an example of (3) total traffic capacity (total capacity of interfaces included in the combination) and (4) total cost value (total equipment cost value (total of equipment costs of interfaces included in the combination)) Table 5 below shows an example of a configuration candidate set of parts to be performed.

From Table 3, the number of slots in the switch housing is “3”, and each slot accommodates one of the interface “1”, the interface “2”, and the interface “3”. Further, as shown in Table 5, in the interface combination candidate “1”, the slot “1” indicating the first slot accommodates the interface “1”, and the slot “2” indicating the second slot. Does not accommodate any interface (“empty”), and slot “3” indicating the third slot does not accommodate any interface (“empty”).
Similarly, other interface combination candidates shown in Table 5 also indicate which interface is accommodated in each slot. In this example of the interface combination candidates, for the sake of simplicity of the embodiment, an example in which the same interface is not accommodated in the same switch except for the interface of the maximum capacity, and a plurality of interfaces are not accommodated in one slot is used. However, the present invention can be implemented in any combination of interface candidates.

Next, an example in the second processing unit 13 will be described.
The second processing unit 13 restricts the set of candidates obtained by the first processing unit 12 of the present embodiment to a plurality of candidates satisfying all the constraint conditions, and the equipment cost value is minimized from the plurality of restricted candidates. Derive the best candidate.

  In step S3 implemented in the second processing unit 13, in order to formulate the problem as a 0-1 integer linear programming problem and efficiently derive an exact solution, a decision variable, a constraint condition, , An objective function value is determined.

  The determination variables set in step S3 in this embodiment will be described. Table 6 below shows a variable x indicating whether or not each route candidate is selected for all the lines.

Each variable has a value of either 0 or 1. When the variable x _1,1 is 1, the route candidate “1” is selected as the route on the line “1”, and the route of the line “1” passes through the node “1”, the link “1”, and the node “2”. Represent. When the variable x _1,1 is 0, it indicates that the route candidate “1” is not selected as a route on the line “1”. Other lines and route candidates are set similarly.
Table 7 below shows variables that are set for each of the candidates for the interface combination of each link.

Each variable has a value of either 0 or 1. When the variable y _1,1 is 1, the interface combination candidate “1” is selected at the link “1”, and the interface “1” is inserted into the slots of the nodes “1” and “2” connected to the link “1”. "Means that it is accommodated. When the variable y _1,1 is 0, it indicates that the interface combination candidate “1” has not been selected for the link “1”. Other link and interface combination candidates are similarly set.

  Next, a description will be given of the constraint condition regarding the decision variable, the constraint condition regarding the traffic capacity, the constraint condition regarding the delay time, and the constraint condition regarding the facilities, which are set in step S3 in the present embodiment.

  First, the constraints on the decision variables will be described. The following conditional expression indicating that one of the route candidates “1”, “2”, or “3” is always selected for each line regarding the constraint condition regarding the variable x. (1) and (2) are provided.

  Here, i indicates a line number, and j indicates a route candidate number.

Equation (1) indicates that the sum of the values of the variable x in each line is 1. Equation (2) indicates that the value of the variable x in each line is either 1 or 0. For example, when i is 1, become x _1,1 + x _1,2 + x _1,3 is 1, x _{1, 1,} x _{1, 2} and x _{1, 3} since is 0 or 1, x _1, The value of only one of ₁ , x _1,2 or x _1,3 becomes 1, and the other values become 0. That is, it indicates that one of the route candidates is selected for each line.
Similarly to the variable x, the following conditional expressions (3) and (4) are provided for the variable y, and these indicate that a candidate for any one interface combination is selected for each link.

  Here, h indicates a link number, and l indicates a candidate number of an interface combination.

  Next, the following conditional expression (5) shows an expression of a constraint condition regarding the traffic volume. This equation indicates that the total traffic volume of any link does not exceed the traffic capacity.

  Here, i represents a line number, j represents a route candidate number, h represents a link number, and l represents a combination number of an interface.

The parameter capacity _l indicates the total traffic capacity of the interface combination 1 obtained by the first processing unit 12 shown in the row of the total traffic capacity in Table 5. The parameter demand _i represents the traffic demand of the line i, which is input data, shown in the traffic demand row of Table 2. The parameter route _{i, j, h} is one variable of the set of used links obtained by the first processing unit 12 shown in the row of used link in Table 4, and indicates whether the route candidate j of the line j uses the link h. Represents

  Here, for the sake of explanation, the route candidate “1” is selected on the line “1”, the route candidate “1” is selected on the line “2”, the route candidate “1” is selected on the line “3”, It is assumed that the route candidate “2” is selected on the line “4”, the route candidate “2” is selected on the line “5”, and the route candidate “2” is selected on the line “6”. Further, it is assumed that the interface combination candidate “1” is selected for all the links. This selection pattern is also used in the following description of the constraint conditions and the objective function value. Here, it is assumed that the variable x is set as shown in Table 8 below.

  It is assumed that the variable y is set as shown in Tables 9 and 10 below.

  First, a method for obtaining the total traffic volume of an arbitrary link will be described. As an example, an equation for calculating the total traffic volume of link "1" is shown below.

In this equation, the values of a variable x (Table 8), a parameter demand (see Table 2 as described above), and a parameter route (see Table 4 as described above) are substituted. For example, route _1,1,1 is a parameter indicating whether the route candidate “1” of the line “1” uses the link “1”. From Table 4, since the link “1” is used, route _1,1,1 = 1.
As described above, by setting the value of the variable x, the total traffic amount per link can be obtained.

  Next, a method for obtaining the traffic capacity will be described. As an example, an equation for determining the traffic capacity of link "1" is shown below.

  In this equation, the variable y (see Tables 9 and 10) and the value of the parameter capacity (see Table 5 as described above) are substituted. In this way, by setting the value of the variable y, the traffic capacity per link can be obtained.

  Next, the following conditional expression (6) shows an expression of a constraint condition regarding the delay time. Indicates that the delay time does not exceed the upper limit value for an arbitrary line.

Here, i represents a line number, j represents a route candidate number, h represents a link number, and l represents a combination number of an interface. The parameter "latency _h" represents the delay time of the link h, which is input data, shown in the delay time row of Table 1. The parameter limit _i represents the upper limit of the delay time of the line i, which is input data, shown in the upper limit row of Table 2. Here, the variable x uses the values shown in Table 8, and the variable y uses the values shown in Tables 9 and 10.

  This section shows how to calculate the delay time of an arbitrary line. As an example, an equation for calculating the delay time of the line “1” is shown below.

  In this equation, the variable x (see Table 8 as described above), the value of the parameter route (see Table 4), and the value of the parameter latency (see Table 1 as described above) are substituted.

Next, the following conditional expression (7) shows the expression of the constraint condition regarding the equipment.
This expression indicates that, in an arbitrary node, the number of interfaces in the network device does not exceed the upper limit of the number of accommodated interfaces. In this embodiment, the upper limit of the number of accommodated interfaces is set to 3. That is, three slots are provided per network device.

Here, i is a line number, j is a route candidate number, h is a link number, l is a combination number of an interface, and m is a node number. The parameter part _l indicates the number of accommodated interfaces in the interface combination l obtained by the first processing unit 12 shown in the row of the number of interfaces in Table 5. The parameter ^t M _{h, m} is the (h, m) component of the transpose of the connection matrix M in equation (A). Here, the variable x uses Table 8, and the variable y uses Tables 9 and 10.

  A method for obtaining the number of interfaces accommodated in a network device of an arbitrary node will be described. As an example, an expression for obtaining the number of accommodated interfaces of the node “1” is shown below.

For this equation, and from the connection matrix M of equation (A), the variable y (see Tables 9 and 10), the parameter M (that is, ^t M _{h, m} ), and the value of the parameter part (as described above) Into Table 5).

  Next, the following equation (8) shows an equation for calculating the total equipment cost value.

Here, h indicates a link number, 1 indicates a combination number of an interface, and m indicates a node number. Parameters cost _l is shown in the line of the total cost values of Table 5, representing the total value of the equipment cost in all interfaces included in the combination of the interface obtained in the first processing unit 12.

Here, as described above, the variable x uses Table 8, and the variable y uses Table 9 and Table 10.
The formula for calculating the total equipment cost value in the entire network is shown below.

For this equation, a variable y (see Tables 9 and 10), a parameter M (that is, ^t _{Mh, m} ) and a connection matrix M of Tables 9, 10, and 5 and Equation (A) are used. , And the value of the parameter cost (see Table 5).

  As described above, the 0-1 integer linear programming problem is solved using the above decision variables, constraints, and objective functions. Here, an exact solution is obtained by applying the LP relaxation method and the branch and bound method, which are known algorithms and are effective for the 0-1 integer programming problem.

  Finally, an example of the output unit 14 will be described. Here, an exact solution (optimum network configuration) obtained by solving the 0-1 integer programming problem represented by the above variables and the above parameters is shown.

  Table 11 below shows the optimal solution of the variable x.

  Tables 12 and 13 below show optimal solutions of the variable y.

FIG. 8 shows an example of a network in which a line is accommodated in a path and an interface is accommodated as in this optimal solution.
In this example, from Table 11, line “1”, line “2”, line “4”, and line “5” use link “1”. Further, the line “2”, the line “3”, the line “5”, and the line “6” use the link “2”.

  Further, from Tables 12 and 13, the link “1” uses the interface “1”. The link “2” uses the interface “1”. Interface “1” at the connection between node “1” and link “1”, interface “1” at the connection between node “2” and link “1”, and connection between node “2” and link “2” The interface “1” is accommodated in the connection part of the node “3” and the link “2”, respectively. That is, from Table 3, since there are four interfaces “1” in total, the total equipment cost value is 3.62 × 4 = 14.48.

Next, it is confirmed whether various constraints are satisfied.
First, the constraints on the traffic capacity will be described. From Table 2, since the traffic demands of the line “1”, the line “2”, the line “4”, and the line “5” are each “1”, the total traffic volume of the link “1” is “4”. The traffic capacity of the link “1” is “10”. Similarly, since the traffic demand of the line “2” and the line “5” is “1” and the traffic demand of the line “3” and the line “6” is “2”, the total traffic volume of the link “2” Is "6". The traffic capacity of the link “2” is “10”. Therefore, in all the links, the total traffic volume is equal to or less than the traffic capacity, and satisfies the traffic capacity constraint.

  Next, a constraint condition regarding the delay time will be described. From Table 2, the upper limit of the line “1” to the line “3” is “3”, and the upper limit of the line “4” to the line “6” is “5”. From Table 1, the delay time of the link “1” is “1”, and the delay time of the link “2” is “2”. The sum of the delay time of link “1” and the delay time of link “2” is “3”. Therefore, the delay time is 3 or less in all the lines, and satisfies the constraint on the delay time.

  Lastly, a description will be given of constraints on equipment. Node “1” uses slot “1-1” because it has one interface “1”, and node “2” uses slot “2-1” and slot “1” because it has two interfaces “1”. 2-2 ", and the node" 3 "uses the slot" 3-1 "since there is one interface" 1 ". Therefore, in all nodes, the number of used slots is 3 or less, which satisfies the constraint on equipment.

  The output information output from the output unit 14 will be described above. This output information is obtained by (1) obtaining a route for each line, and (2) arranging a network device under the condition that a network device having no interface installed in all link units in the network device is not arranged in a node. As a result of obtaining the node, (3) the result of obtaining the interface to be arranged in the link unit in the network device for each link unit.

  As a result of obtaining the route for each line, the link used for the line “1” and the line “4” is the link “1”, and the link used for the line “2” and the line “5” is the link “1” And the link "2", and the link used by the line "3" and the line "6" is the link "2".

  In addition, the result of obtaining the node in which the network device is arranged indicates that the network device is arranged in the node “1”, the node “2”, and the node “3”.

  In addition, the result of obtaining the interface to be arranged in the link unit in the network device for each link unit (the interface information included in the link unit of each network device) indicates that the network device of the node “1” has the interface “ 1 is arranged, the network device of the node "2" is arranged with one interface "1" at each of two link units, and the network device of the node "3" is arranged with the interface "1" at one link unit. Is arranged.

  As described above, according to the present invention, it is possible to design an optimal network configuration in the arrangement of arbitrary nodes and links, accommodation lines, and network facilities.

  The present invention is not limited to the above-described embodiment, and can be variously modified in an implementation stage without departing from the gist of the invention. In addition, the embodiments may be appropriately combined and implemented, and in that case, the combined effects can be obtained. Further, the above-described embodiment includes various inventions, and various inventions can be extracted by combinations selected from a plurality of disclosed constituent features. For example, even if some components are deleted from all the components shown in the embodiment, if the problem can be solved and an effect can be obtained, a configuration from which the components are deleted can be extracted as an invention.

  In addition, the method described in each embodiment can be executed by a computer (computer) as a program (software means) such as a magnetic disk (floppy (registered trademark) disk, hard disk, or the like), an optical disk (CD-ROM, It can be stored in a recording medium such as a DVD, MO, or a semiconductor memory (ROM, RAM, flash memory, etc.), and can also be transmitted and distributed via a communication medium. The programs stored in the medium include a setting program for causing the computer to execute software means (including not only the execution program but also tables and data structures) to be executed by the computer. A computer that implements the present apparatus reads a program recorded on a recording medium, and in some cases, constructs software means using a setting program, and executes the above-described processing by controlling the operation of the software means. The recording medium referred to in this specification is not limited to a medium for distribution, but includes a storage medium such as a magnetic disk or a semiconductor memory provided in a computer or a device connected via a network.

  The present invention can be applied to the information and communication industry.

  DESCRIPTION OF SYMBOLS 10 ... Network design apparatus, 11 ... Input part, 12 ... First processing part, 12a ... First calculation part, 12b ... Second calculation part, 13 ... Second processing part, 13a ... Calculation part, 14 ... Output part.

Claims (5)

In a network composed of a plurality of nodes and a plurality of link units connecting between network devices disposed in the nodes, (1) a path in which all lines are accommodated, and (2) a path in which all nodes are accommodated. A network design device for determining the presence / absence of network devices to be arranged, and (3) types of interfaces respectively installed in all the link units in all the network devices,
(1) information about a connection state between the nodes, (2) information about a plurality of lines accommodated in the network, and (3) information about network devices and link unit facilities arranged in the nodes. An input unit for inputting,
A first calculation unit that calculates candidates for disjoint routes accommodating the line and a set of link units used by the candidates based on input information from the input unit; and (1) the link unit And (2) the number of interfaces included in the combination, (3) the total capacity of the interfaces included in the combination, and (4) the interfaces included in the combination. A first processing unit having a second calculating unit that calculates a set of total equipment cost values of
The line is requested from the network based on (1) the input information and (2) a combination candidate of the route obtained by the first calculation unit and the interface obtained by the second calculation unit. A set of candidates for the combination of the route and the interface that satisfies the constraint conditions is obtained, and in the obtained set of candidates, the optimal route and interface are set such that the total equipment cost value in the entire network is minimized. And a second processing unit that derives a combination candidate. On the basis of the candidates obtained by the second processing unit, (1) a route for each line is obtained, and (2) a network device in which none of the interfaces is installed in all link units in the network device is assigned to the node. A node for arranging the network device under the condition that the network device is not arranged; and (3) an output unit for obtaining an interface to be arranged in a link unit in the network device for each of the link units, and outputting the obtained result. Item 2. The network design apparatus according to item 1. 2. The network according to claim 1, wherein (1) a route in which all lines are accommodated, (2) presence / absence of a network device arranged in each of the nodes, and all routes in all of the network devices. 3. A network design method for determining a type of an interface installed in each of the link units,
Based on the input information, by using the k-shortest path algorithm, to accommodate the line, to calculate a set of disjoint path candidates,
By enumerating the combinations of all the interfaces, a set of candidates for combinations of disjoint interfaces constituting the link portion is calculated,
Based on the input information, the calculated set of route candidates, and the calculated set of interface combination candidates, such that the line satisfies the constraints required by the network, A candidate set of combinations is determined, and the set of candidates is solved as an optimization problem, and an optimal solution, that is, a combination of an optimal path and an interface is determined, so that the equipment cost value in the entire network is minimized. A network design method for deriving candidates for combinations of interfaces and interfaces. The network design method according to claim 3, wherein the optimization problem is a problem formulated as a 0-1 linear programming problem. A program used in a computer that operates as a part of the network design device according to claim 1 or 2,
Said computer,
A network design processing program for functioning as the input unit, the first processing unit, and the second processing unit.