Mesh design device, mesh design method, and mesh design program

Abstract

To provide a technology that increases the flexibility of mesh composition and enables a larger mesh capacity to be obtained.SOLUTION: A calculation unit 12 performs the following calculation. A number of links v is divided into two different numbers of links v1, v2. Multiple intermediate switches are divided into two groups by the two numbers of links v1, v2. A number of terminals n, a number of input / output switches k, a number of intermediate switches in a first group m1, a number of intermediate switches in a second group m2, a number of links v1 between one input / output switch and one intermediate switch in a first group, and a number of links v2 between one input / output switch and one intermediate switch in a second group are calculated so that the total probability of each blocking occurrence probability pertaining to the first group does not exceed a predetermined blocking rate ε1, and the total probability of each blocking occurrence probability pertaining to the second group does not exceed an allowable blocking rate ε when the number of terminals n connected to the input or output ports of multiple input / output switches is less than or equal to the sum of a first terminal number n1snb and a second terminal number n2snb.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present disclosure relates to a network design device, a network design method, and a network design program. [Background technology]

[0002] There is a network design technique that theoretically guarantees the blocking rate of paths (Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Haruto Taka and two others, “Design of Twisted and Folded Clos Network with Guaranteeing Admissible Blocking Probability”, IEEE Networking Letters, pp.1-1, 2003 Summary of the Invention [Problem to be solved by the invention]

[0004] However, in Non-Patent Document 1, if the number of links between the output switch and the middle switch is allowed to be completely free, the number of variables in the formulated ILP (Integer Linear Programming) becomes enormous, making it difficult to solve the ILP equation within a realistic time.

[0005] The present disclosure has been made in consideration of the above circumstances, and aims to provide a technique that can increase the flexibility of network configuration and obtain a larger network capacity. [Means for solving the problem]

[0006] A network design device according to one aspect of the present disclosure is a network design device that uses a plurality of intermediate switches and a plurality of input / output switches connected via one of the intermediate switches to design a network that satisfies a predetermined network capacity, the network design device comprising: an acquisition unit that acquires the number of switch ports, a plurality of input / output switches having the number of input ports and the number of output ports of the number of ports, and a plurality of intermediate switches, a terminal utilization rate of terminals connected to the plurality of input / output switches, and an allowable blocking rate that allows inability to communicate between two terminals connected to the plurality of input / output switches; and an acquisition unit that acquires the number of terminals, the number of input / output switches, and the number of intermediate switches such that in a network that satisfies the number of ports, the total number, and the terminal utilization rate, a total probability of each blocking occurrence probability when the number of terminals is increased by one from the maximum number of terminals when the network is non-blocking does not exceed the allowable blocking rate, and a network capacity obtained by multiplying the number of terminals and the number of input / output switches is maximized. and a calculation unit that calculates the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, the number of intermediate switches in the second group, the number of links between one input / output switch and one of the intermediate switches in the first group, and the number of links between one input / output switch and one of the intermediate switches in the second group, wherein the calculation unit divides the number of links into two different numbers of links and divides the multiple intermediate switches into two groups based on the two numbers of links, and calculates the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, the number of intermediate switches in the second group, the number of links between one input / output switch and one of the intermediate switches in the first group, and the number of links between one input / output switch and one of the intermediate switches in the second group, so that when the number of terminals connected to input ports or output ports of the multiple input / output switches is equal to or less than the total number of a first terminal number and a second terminal number, the calculation unit calculates the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, the number of intermediate switches in the second group, the number of links between one input / output switch and one of the intermediate switches in the first group, and the number of links between one input / output switch and one of the intermediate switches in the second group, so that the total probability of each blockage occurrence probability related to the first group does not exceed a predetermined blockage rate (not less than the allowable blockage rate and not more than 1) and the total probability of each blockage occurrence probability related to the second group does not exceed the allowable blockage rate.

[0007] A network design method according to one aspect of the present disclosure is a network design method for designing a network that satisfies a predetermined network capacity using a plurality of intermediate switches and a plurality of input / output switches connected via one of the intermediate switches, the method comprising the steps of: a first step in which a network design device acquires the number of switch ports, a plurality of input / output switches having the number of input ports and the number of output ports of the number of ports, a total number of a plurality of intermediate switches, a terminal utilization rate of terminals connected to the plurality of input / output switches, and an allowable blocking rate that allows inability to communicate between two terminals connected to the plurality of input / output switches; and a second step in which a network design device acquires the number of terminals, the number of input / output switches, the number of intermediate switches, and the number of intermediate switches such that, in a network that satisfies the number of ports, the total number, and the terminal utilization rate, a total probability of each blocking occurrence probability when the number of terminals is increased by one from the maximum number of terminals when the network is non-blocking does not exceed the allowable blocking rate, and a network capacity obtained by multiplying the number of terminals and the number of input / output switches is maximized. and a second step of calculating the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, the number of intermediate switches in the second group, the number of links between one input / output switch and one of the intermediate switches in the first group, and the number of links between one input / output switch and one of the intermediate switches in the second group, in such a way that when the number of terminals connected to the input ports or output ports of the plurality of input / output switches is equal to or less than the sum of the first number of terminals and the second number of terminals, the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, the number of intermediate switches in the second group, the number of links between one input / output switch and one of the intermediate switches in the first group, and the number of links between one input / output switch and one of the intermediate switches in the second group are calculated so that the total probability of each of the blockage occurrence probabilities in the first group does not exceed a predetermined blockage rate (not less than the allowable blockage rate and not more than 1) and the total probability of each of the blockage occurrence probabilities in the second group does not exceed the allowable blockage rate.

[0008] A network design program according to an aspect of the present disclosure causes a computer to function as the network design device. Effect of the Invention

[0009] According to the present disclosure, it is possible to provide a technique that can increase the flexibility of a network configuration and obtain a larger network capacity. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a diagram illustrating an example of path setting. [Diagram 2] FIG. 2 is a diagram showing a path closure situation. [Diagram 3] FIG. 3 is a diagram showing a conventional Clos network. [Figure 4] FIG. 4 is a diagram showing an example of construction of a Clos network according to this embodiment. [Diagram 5] FIG. 5 is a diagram showing examples of algorithms 1 and 2. [Figure 6] FIG. 6 is a diagram illustrating an example of a functional block configuration of a network design device. [Figure 7] FIG. 7 is a diagram showing a network design processing flow. [Figure 8] FIG. 8 is a diagram showing the results of a numerical experiment on the network capacity versus the allowable blocking rate. [Figure 9] FIG. 9 is a reference diagram for explaining the occlusion. [Figure 10] FIG. 10 illustrates an example of a hardware configuration of a network design device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0012] The present disclosure discloses a network design technique for data centers, communication buildings, and the like.

[0013] FIG. 1 is a diagram illustrating an example of path setting.

[0014] In a data center, multiple racks (racks housing switches, servers, etc.) installed within the data center are housed in a network, and a communication line (path) is set up by connecting designated rack pairs so that they can communicate with each other.

[0015] In a communications building, multiple transmission devices are housed in a network and are connected to transmission devices in other buildings, other operators, etc. Current transmission devices use packet switching technologies such as IP and Ethernet, but in the future, circuit switching technologies such as optical communications will be used to improve transmission capacity and power efficiency.

[0016] FIG. 2 is a diagram showing a path closure situation.

[0017] In circuit switching, if there is no unused path between the pair for which a connection is requested, the path cannot be set up. This is called blocking. We want to set up a path between the second port P12 of the first switch S1 and the fourth port P34 of the third switch S3. However, because there are no free paths due to the two previously set paths (dashed lines), the desired path cannot be set up.

[0018] This disclosure describes a network design technique that maximizes network capacity (the number of switches and transmission devices that can be accommodated) while satisfying a given allowable blocking rate. In particular, it targets Clos networks with a Clos structure that are widely used in data centers.

[0019] FIG. 3 is a diagram showing a conventional Clos network.

[0020] A Clos network is composed of three layers: an input layer, an intermediate layer, and an output layer. However, the input layer and the output layer are combined into one layer called the input-output stage, and the intermediate layer is folded back.

[0021] Non-Patent Document 1 discloses a network design technique that theoretically guarantees a blocking rate for such a Clos network having a Clos structure called TF-Clos (Twisted and Folded Clos Network).

[0022] Terminals such as racks and transmission devices are connected to the input / output switches in the input / output layer. Terminal t is a transmitting terminal, and terminal r is a receiving terminal. The number of input / output switches in the input / output layer is k, and the number of intermediate switches in the intermediate layer is m. For all switches, the input / output switches are numbered S1, ..., S k , S k+1 …S k+m The input / output switches and the intermediate switches are connected by v links at the input and output, respectively. Each input / output switch has n terminals connected at the input and output, respectively.

[0023] Now, consider the following network design problem.

[0024] Design a network with the largest possible capacity using a N×N switches with N ports each for input and output. The number of terminals nk (=n×k; input and output are separate) in the entire network is taken as the network capacity. However, the blocking rate must be less than or equal to the specified allowable blocking rate ε. Also, the terminal utilization rate p of each terminal is assumed to be known, as it can be estimated from past service performance, etc.

[0025] When this problem is formulated as ILP (Integer Linear Programming), it becomes equations (1a) to (1i).

[0026]

number

[0027] N, a, p, ε are given as problems. snb , k, m, and v are the decision variables of the optimization problem. θ is a parameter required only for the solution. snb is the number of terminals that are non-blocking. snb is an abbreviation for "strict-sense non-blocking". θ is a threshold for distinguishing the number of terminals that are blocking and non-blocking.

[0028] Equation (1a) is an objective function for maximizing network capacity. Equation (1b) is a constraint on the number of ports of input and output switches. Equation (1c) is a constraint on the number of ports of middle switches. Equation (1g) is a constraint on the number of switches.

[0029] Equation (1d) indicates that the total probability of blockage occurrence is equal to or less than the allowable blockage rate ε. Equation (1f) indicates that there is no blockage (blockage rate is zero). Equation (1e) indicates that the conditions of equations (1d) and (1f) do not coexist.

[0030] In detail, equation (1f) is expressed as n snb This represents a situation where no more than n terminals are in use and communicating. snb +1 (=θ+1)” or more terminals are in use.

[0031] Equation (1f) is the number of non-blocking terminals n snb Since the maximum value of "n snb Blockages can occur when more than one terminal is connected. The left side of equation (1d) calculates the probability of such a blockage occurring.

[0032] In formula (1d), the sum symbol (Σ) on the left side means "n snb +1” pieces, “n snb +2”, …, n, terminals are used, the blockage occurrence probabilities of each are added together. Finally, equation (1d) shows the condition under which the total occurrence rate of the blockage occurrence probabilities that can cause blockage does not exceed the allowable blockage rate ε.

[0033] In this way, Non-Patent Document 1 divides the equation into equation (1d) for when blockage occurs and equation (1f) for when no blockage occurs, and separates equations (1d) and (1f) by equation (1e), so that the blockage rate can be theoretically guaranteed.

[0034] Non-Patent Document 1 is groundbreaking in that it can theoretically guarantee the blocking rate. However, when the terminal utilization rate p is large (close to 1) or when the allowable blocking rate ε is 10 -5 In the case where the capacity is small, it is not possible to design a network with a large capacity. Therefore, the method of Non-Patent Document 1 is improved.

[0035] In Non-Patent Document 1, the number of links v between switches is set to a uniform value. On the other hand, in this embodiment, a different number of links is allowed for each switch, which increases the flexibility of the network configuration and makes it possible to obtain a larger network capacity. However, if the number of links v is allowed to be completely arbitrary, the number of variables in the ILP formulation becomes enormous, making it impossible to solve the ILP equation within a realistic time frame.

[0036] Therefore, in this embodiment, the number of links v is set to only two different numbers of links v1 and v2, and the constraint condition of the allowable blocking rate ε is set to two levels accordingly, thereby obtaining a formulation that can be solved realistically. In other words, in a network design that theoretically guarantees the blocking rate, the number of links between switches is not uniformly set to v, but is set to a different value of v1 or v2 for each intermediate switch, thereby increasing the flexibility of the network configuration and obtaining a larger network capacity.

[0037] FIG. 4 is a diagram showing an example of construction of a Clos network according to this embodiment.

[0038] The middle switches are divided into two groups based on the number of links v1 and v2 with the input / output switches. In other words, the number of middle switches with v1 links to each input / output switch is m1, which constitutes the first group (First-step group). The number of middle switches with v2 links to each input / output switch is m2, which constitutes the second group (Second-step group). These two groups are then used as follows:

[0039] In the first step, the number of terminals n connected to either the input or output side of the connection request is equal to the number of terminals n1 snbWhen the number of terminals is n1 or less, a connection is made using one of the m1 intermediate switches in the first group. snb is obtained as part of the solution. The probability of being able to connect in the first step is less than or equal to the allowable blocking rate ε. The probability of not being able to connect in the first step is less than or equal to a predetermined blocking rate ε1 (ε≦ε1≦1).

[0040] In the second step, the number of terminals n connected to either the input or output side of the connection request is increased to the number of terminals n1 snb Beyond the individual, "n1 snb +n2 snb When the number of nodes is less than or equal to m, one of the m2 intermediate switches included in the second group is further used to make a connection. The probability that a connection cannot be made even in the second step is equal to or less than the allowable blocking rate ε.

[0041] If none of the above is true, the connection request is rejected with a rejection probability equal to or less than the allowable blocking rate ε.

[0042] The ILP of this embodiment is expressed by formulas (2a) to (2m).

[0043]

number

[0044] The objective function of formula (2a) is the same as formula (1a). The constraints of formulas (2b) to (2d) and (2i) are basically the same as those of formulas (1b), (1c) and (g), respectively. Formulas (2j) to (2m) are snb , n2 snb , m1, m2, v1, and v2. Therefore, hereinafter, formulas (2e) to (2h) will be explained.

[0045] The constraints in (2e) and (2f) are the same as (1d) and (1f), respectively, without the subscript "1". In (1d), a parameter θ is used, but θ is a threshold value for distinguishing between blocking and non-blocking terminal numbers, so the number of non-blocking terminals n snbThat is, the first step is basically the same as that in Non-Patent Document 1, which uses only m1 middle switches.

[0046] Next, we consider how many of the connection requests that could not be connected in the first step can be saved in the second step. In other words, we consider whether the blocking rate can be reduced to the allowable blocking rate ε. Of the terminal utilization rate p, the proportion of connections that could not be connected in the first step and are sent to the second step is ε1, so the terminal utilization rate p seen from the second step is ε1p. Also, in the first step, the maximum number of unused terminals was n, but in the second step, it is n1 snb Since n is already in use in the first step, the number of available terminals is "n-n1 snb " pieces.

[0047] Therefore, in the second step, for the connection requests that could not be connected in the first step, in equations (2e) and (2f) used in the first step, p is replaced with ε1p and n is replaced with "n-n1 snb " Use equation (2g) and equation (2h) which have been replaced with ".

[0048] An algorithm for calculating equations (2a) to (2m) using a computer will now be described.

[0049] In Non-Patent Document 1, all decision variables are integers, but in this embodiment, a real decision variable ε1 is included. Therefore, the optimal solution (optimum network configuration) cannot be obtained by the same method as Non-Patent Document 1. Therefore, the optimal solution of Equation (2) is calculated by the following method.

[0050] In Algorithm 1, we use integer decision variables n, k, and n1. snb , n2 snb Search for all possible combinations of values ​​for m1, m2, v1, and v2. Specifically, check whether each combination satisfies the constraints (2b) to (2d), (2f), (2h), and (2i) that do not include ε1, and select a feasible combination of values ​​that satisfies all of these conditions.

[0051] In Algorithm 2, it is checked whether there exists an ε1 that satisfies the remaining constraint conditions for the combination of possible values ​​selected in Algorithm 1. Note that the value of ε1 is not necessary for operation, so it is sufficient to check only the existence of a feasible ε1.

[0052] FIG. 5 is a diagram showing examples of algorithms 1 and 2.

[0053] We will explain Algorithm 1.

[0054] Lines 3 and 4 initialize the variables for saving the optimal solution. Line 5 loops through each possible combination of values ​​for the integer decision variables. It is desirable to loop in descending order of nk. This allows the loop to be terminated if an "optimal config" has been found after processing a certain nk.

[0055] In line 6, we check the feasibility of the integer decision variables. In line 7, we check whether there is a feasible ε1. Details will be explained in Algorithm 2 below. Note that ε1 min is defined as in equation (3). The meaning of equation (3) is also explained in Algorithm 2.

[0056]

number

[0057] If lines 6 and 7 both determine that the solution is feasible, line 8 is called. If the solution is more optimal than the previous solution (i.e., the network capacity nk is larger), it is saved as the new optimal solution. Line 15 returns the final optimal solution.

[0058] We will explain Algorithm 2.

[0059] In algorithm 2 (ε1-existence judgmen algorithm; ε1-JA), ε1 lower and ε1 upper If it finds an ε1 that satisfies formula (2g) within the range, it returns “true”; otherwise it returns “false”.

[0060] The first time it is called is in line 7 of Algorithm 1. At this time, the range that ε1 can take is ε1 lower and ε1 upper is set. ε1 lower is the minimum value that satisfies both equations (2e) and (2j), and is expressed as equation (3). upper is "1". ε1 upper =1 clearly satisfies both equations (2e) and (2j).

[0061] Algorithm 2 performs recursive calls while gradually narrowing the range of possible values ​​for ε1. Lines 4 and 5 return "false" if a feasible ε1 is not found after searching a sufficiently small range. Lines 7 and 8 return "true" if a feasible ε1 is found.

[0062] In addition, f(ε1 lower ) is defined by equation (4). Equation (4) is just equation (2g) rewritten in the form f(ε1)≧0. Note that the notation "lower" is omitted on the left side of equation (4).

[0063]

number

[0064] In line 11, the range of ε1 is split in half and "ε1-JA" is recursively called, and if either is "true", it returns "true" by connecting with "or". Here, the if statement in line 10 is a condition to eliminate unnecessary searches. + is defined as in equation (5).

[0065]

number

[0066] The if statement in line 10 eliminates unnecessary searches because f(ε1)<0 in line 7 and f + If (ε1+δ)<0, f + This is because there is no δ' (0<δ'≦δ) that satisfies (ε1+δ')≧0. In other words, since ε1p<(ε1+δ)p, f + (ε1+δ)≧f(ε1+δ), and f + Since (ε1+δ) is a non-decreasing function of δ, f + (ε1+δ')<0 is f + This is because it folds when (ε1+δ).

[0067] In addition, since f(ε1) is basically a monotonically decreasing function, the same results can be obtained even if the recursive search in lines 9 to 13 of Algorithm 2 is omitted. min If x is negative, then larger values ​​will also be negative, so it is sufficient to only look at the minimum. Therefore, this recursive search is optional.

[0068] Furthermore, equation (2) may be solved using methods other than algorithms 1 and 2. Algorithms 1 and 2 are examples of methods for solving equation (2).

[0069] FIG. 6 is a diagram showing an example of a functional block configuration of the network design device 1 according to the present embodiment.

[0070] The network design device 1 is a device that designs a network that satisfies a predetermined network capacity by using a plurality of middle switches and a plurality of input / output switches connected via one middle switch. The network design device 1 includes an acquisition unit 11, a calculation unit 12, and a storage unit 13.

[0071] The acquisition unit 11 has a function of acquiring N, a, p, and ε given for calculating a network optimization problem. That is, the acquisition unit 11 acquires the number of switch ports N, the total number a of a plurality of input / output switches and a plurality of intermediate switches each having an input port with the number of ports N and an output port with the number of ports N, the terminal utilization rate p of terminals connected to the plurality of input / output switches, and the allowable blocking rate ε that allows inability to communicate between two terminals connected to the plurality of input / output switches.

[0072] The calculation unit 12 has a function of calculating decision variables of an optimization problem.

[0073] In other words, similar to Non-Patent Document 1, the calculation unit 12 calculates the number of terminals n, the number of input / output switches k, the number of intermediate switches m, and the number of links v between one input / output switch and one intermediate switch so that, in a network that satisfies the number of switch ports N, the total number of switches a, and the terminal utilization rate p, the total probability of each blocking occurrence probability when the number of terminals n is increased by 1 from the maximum value of the number of terminals n when the network is non-blocking does not exceed the allowable blocking rate ε, and the network capacity nk obtained by multiplying the number of terminals n and the number of input / output switches k is maximized.

[0074] In this embodiment, the calculation unit 12 divides the number of links v into two different numbers of links v1 and v2, and divides the multiple middle switches into two groups according to the two numbers of links v1 and v2.

[0075] In this embodiment, the calculation unit 12 calculates whether the number of terminals n connected to the input ports or output ports of the multiple input / output switches is equal to or greater than the first number of terminals n1 snb In the following case, the number of terminals n, the number of input / output switches k, the number of intermediate switches m1 in the first group, and the number of links v1 between one input / output switch and one intermediate switch in the first group are calculated so that the total probability of each blocking occurrence probability related to the first group does not exceed the allowable blocking rate ε.

[0076] In this embodiment, the calculation unit 12 calculates whether the number of terminals n connected to the input ports or output ports of the multiple input / output switches is equal to or greater than the first number of terminals n1snb and the second terminal number n2 snb If the total number of the blockage occurrence probabilities in the first group is equal to or less than the total number of the blockage occurrence probabilities in the second group, the number of terminals n, the number of input / output switches k, the number of intermediate switches m1 in the first group, the number of intermediate switches m2 in the second group, the number of links v1 between one input / output switch and one intermediate switch in the first group, and the number of links v2 between one input / output switch and one intermediate switch in the second group are calculated so that the total probability of the blockage occurrence probabilities in the first group does not exceed a predetermined blockage rate ε1 (ε≦ε1≦1) and the total probability of the blockage occurrence probabilities in the second group does not exceed the allowable blockage rate ε.

[0077] The storage unit 13 has a function of storing data to be processed by the network design device 1.

[0078] FIG. 7 is a diagram showing a flow of a network design process performed by the network design device 1. As shown in FIG.

[0079] Step S101; The acquisition unit 11 acquires the number of ports N of a switch, the total number a of switches, a terminal utilization rate p, and an allowable blocking rate ε given to calculate the network optimization problem.

[0080] Step S102; The calculation unit 12 calculates decision variables of the optimization problem using the parameters acquired in step S101 and equations (2a) to (2m).

[0081] (First step) The number of terminals connected to either the input or output switch is n1. snb If the number of switches is less than m, only the m1 middle switches in the first group are used, and the decision variables are optimized using equations (2e) and (2f) with ε on the right hand side of “if m2=0”.

[0082] In this case, the number of terminals n, the number of input / output switches k, the number of intermediate switches m1 in the first group, and the number of links v1 between one input / output switch and one intermediate switch in the first group are calculated.

[0083] (Second step) The number of terminals connected to either the input or output switch is n1. snb Beyond the individual, "n1 snb +n2 snb If the number of intermediate switches is less than or equal to m1, the number of intermediate switches in the first group and the number of intermediate switches in the second group are used, and for the first group, the decision variables are optimized using equations (2e) and (2f) with the right-hand side of ε1 for “if m2>0”, and for the second group, the decision variables are optimized using equations (2g) and (2h).

[0084] In this case, the number of terminals n, the number of input / output switches k, the number of intermediate switches m1 in the first group, the number of intermediate switches m2 in the second group, the number of links v1 between one input / output switch and one intermediate switch in the first group, and the number of links v2 between one input / output switch and one intermediate switch in the second group are calculated.

[0085] Step S103; The calculation unit 12 outputs the decision variables optimized in step S102 as network design data, and causes the storage unit 13 to store it.

[0086] FIG. 8 is a diagram showing the results of a numerical experiment on the network capacity versus the allowable blocking rate ε.

[0087] Numerical experiments are performed to show how much the network capacity can be increased by this embodiment. The objects for comparison are a non-blocking network (ε=0) and the method of Non-Patent Document 1. "One-step" is the method of Patent Document 1, and "Two-step" is the method of this embodiment. N=200 and a=160 are set.

[0088] In the method of Patent Document 1, when the terminal utilization rate p is large (close to 1) or when the allowable blocking rate ε is 10 -5 On the other hand, in this embodiment, even under such conditions, the network capacity is larger than that of the method of Patent Document 1.

[0089] FIG. 9 is a reference diagram for explaining the occlusion.

[0090] A further explanation on blockage.

[0091] First, we will supplement the usage of the network that is the subject of this embodiment. In the subject network, the terminals and the input ports or output ports are connected from the beginning to the end. For example, the terminal t1 1 is connected to switch S1, but this wiring does not change from beginning to end. The same is true for the other terminals. All terminals t and r are connected to either the inlet or outlet switch from beginning to end.

[0092] What changes within the network are the connection configuration within the switches and the presence or absence of signals between terminals. Communication takes place between transmitting terminal t and receiving terminal r. A terminal cannot communicate with multiple terminals simultaneously, but it can communicate with different terminals at different times of the day. The percentage of time that each terminal is attempting to communicate is called the "terminal utilization rate."

[0093] For example, two terminals t1 1 , r1 k Consider a situation where two terminals t1 and t2 want to communicate using only idle links (links with no signal). 1 , r1 k Prepare a path in between.

[0094] In the example of FIG. 9, the inlet / outlet switch S1 and the intermediate switch S k+1 There is an empty link between the k+1 and in / out switch S k Suppose there is also an empty link between and . Therefore, we introduce an intermediate switch S k+1 Change the internal configuration of the two in / out switches S1 and S k The internal structure of the same is two terminals t1 1 , r1 k As a result, terminal t1 1 From terminal r1 kA path consisting of only free links is created until terminal t1. 1 When terminal r1 sends a signal k It will be possible to reach.

[0095] When communication begins in this way, the links on the path become in use and are no longer free links, so they cannot be used for other communications. After that, when the communication ends, they return to being free links. Links between input / output switches and terminals are bound to specific terminals, but links between input / output switches and intermediate switches can be used by any terminal. If the number of free links between switches decreases, communication may not be possible because there is no path made up of free links even if communication is attempted. This is called "blocking." Equations (1f), (2f), and (2h) express the conditions under which blocking will never occur. It has been mathematically proven that blocking will never occur, regardless of how the terminals to be communicated with are selected or the order in which they are selected.

[0096] As described above, according to this embodiment, the number of links between switches is not uniformly set to v, but is made to take a different value of v1 or v2 for each intermediate switch, thereby increasing the flexibility of the network configuration and achieving a larger network capacity.

[0097] The present disclosure is not limited to the above-described embodiment, and various modifications are possible within the scope of the present disclosure.

[0098] The network design device 1 of this embodiment described above can be realized, for example, using a general-purpose computer system including a CPU 901, a memory 902, a storage 903, a communication device 904, an input device 905, and an output device 906, as shown in FIG. 10.

[0099] The memory 902 and the storage 903 are storage devices. In the computer system, the CPU 901 executes a predetermined program loaded onto the memory 902, whereby each function of the network design device 1 is realized.

[0100] The network design device 1 may be implemented by one computer. The network design device 1 may be implemented by multiple computers. The network design device 1 may be a virtual machine implemented in a computer.

[0101] The program for the network design device 1 can be stored in a computer-readable recording medium such as a HDD, SSD, USB memory, CD, or DVD. The computer-readable recording medium is, for example, a non-transitory recording medium. The program for the network design device 1 can also be distributed via a communication network. [Explanation of symbols]

[0102] 1 Network design device 11 Acquisition Department 12 Calculation section 13 Storage section 901 CPU 902 Memory 903 Storage 904 Communication equipment 905 Input Device 906 Output Device

Claims

1. 1. A network design device for designing a network having a predetermined network capacity using a plurality of intermediate switches and a plurality of input / output switches connected via one of the intermediate switches, comprising: an acquisition unit that acquires the number of ports of a switch, the total number of a plurality of input / output switches and a plurality of intermediate switches each having the number of input ports and the number of output ports, a terminal utilization rate of a terminal connected to the plurality of input / output switches, and an allowable blocking rate that allows communication to be disabled between two terminals connected to the plurality of input / output switches; a calculation unit that calculates the number of terminals, the number of input / output switches, the number of intermediate switches, and the number of links between one of the input / output switches and one of the intermediate switches, so that in a network that satisfies the number of ports, the total number, and the terminal utilization rate, a total probability of each blocking occurrence probability when the number of terminals is increased by one from a maximum value of the number of terminals when the network is non-blocking does not exceed the allowable blocking rate, and a network capacity obtained by multiplying the number of terminals by the number of input / output switches is maximized; The calculation unit is dividing the number of links into two different numbers of links, and dividing the plurality of intermediate switches into two groups according to the two numbers of links; when the number of terminals connected to the input ports or output ports of the plurality of input / output switches is equal to or less than the total number of the first terminals and the second terminals, the total probability of each of the blockage occurrence probabilities related to the first group does not exceed a predetermined blockage rate (not less than the allowable blockage rate and not more than 1), and the total probability of each of the blockage occurrence probabilities related to the second group does not exceed the allowable blockage rate, A network design device that calculates the number of terminals, the number of input / output switches, the number of intermediate switches in a first group, the number of intermediate switches in a second group, the number of links between one input / output switch and one intermediate switch in the first group, and the number of links between one input / output switch and one intermediate switch in the second group.

2. The calculation unit is When the number of terminals connected to the input ports or the output ports of the plurality of input / output switches is equal to or less than a first number of terminals, the total probability of each of the blockage occurrence probabilities related to the first group does not exceed the allowable blockage rate, 2. A network design device according to claim 1, further comprising: a calculation unit for calculating the number of terminals, the number of input / output switches, the number of intermediate switches in the first group, and the number of links between one input / output switch and one intermediate switch in the first group.

3. 1. A network design method for designing a network having a predetermined network capacity using a plurality of intermediate switches and a plurality of input / output switches connected via one of the intermediate switches, comprising: A network design device, a first step of acquiring a number of ports of a switch, a total number of a plurality of input / output switches and a plurality of intermediate switches each having the number of input ports and the number of output ports, a terminal utilization rate of terminals connected to the plurality of input / output switches, and an allowable blocking rate that allows communication to be disabled between two terminals connected to the plurality of input / output switches; a second step of calculating the number of terminals, the number of input / output switches, the number of intermediate switches, and the number of links between one of the input / output switches and one of the intermediate switches, so that in a network that satisfies the number of ports, the total number, and the terminal utilization rate, the total probability of each blocking occurrence probability when the number of terminals is increased by one from the maximum number of terminals when the network is non-blocking does not exceed the allowable blocking rate, and the network capacity obtained by multiplying the number of terminals by the number of input / output switches is maximized; In the second step, dividing the number of links into two different numbers of links, and dividing the plurality of intermediate switches into two groups according to the two numbers of links; when the number of terminals connected to the input ports or output ports of the plurality of input / output switches is equal to or less than the total number of the first terminals and the second terminals, the total probability of each of the blockage occurrence probabilities related to the first group does not exceed a predetermined blockage rate (not less than the allowable blockage rate and not more than 1), and the total probability of each of the blockage occurrence probabilities related to the second group does not exceed the allowable blockage rate, A network design method that calculates the number of terminals, the number of input / output switches, the number of intermediate switches in a first group, the number of intermediate switches in a second group, the number of links between one input / output switch and one intermediate switch in the first group, and the number of links between one input / output switch and one intermediate switch in the second group.

4. A network design program for causing a computer to function as the network design device according to claim 1.

Images