JPH03270528A - Line capacity deciding method for communication network - Google Patents

Abstract

PURPOSE:To decide a line capacity form in a complicated communication network within a practical processing time by synthesizing line capacity forms obtained from split partial networks so as to form the line capacity form of a communication network being a processing object. CONSTITUTION:A communication network development supporting device 10 consists of an input/output control section 11, a processing data reception section 12, a split processing section 13, an optimizing processing section 14 and a synthesis processing section 15. The input/output control section 11 executes the interface processing between the device 10 and a man-machine interface 20. The processing data reception section 12 acquires the information required to decide the line capacity between nodes. The above-mentioned processing sections 13, 14, 15 decide the line capacity between nodes of a communication network being the object of design by applying a prescribed algorithm to the information acquired by the processing data reception section 12. Thus, the line capacity form realizing the optimization of the line cost is decided efficiently within a practical processing time.

Description

[Detailed Description of the Invention] [Summary] This invention relates to a communication network line capacity determination method that determines a line capacity form that realizes optimization of the line cost of a communication network when the communication demand between nodes is given. The purpose of this method is to decompose the communication network into a plurality of subnetworks by decomposing the network structure of the communication network to be processed according to two connected components, with the aim of making decisions efficiently within a practical processing time. 1 processing process,
A second processing step in which the communication line capacity between the nodes of the partial network obtained in the first processing step is determined to give the optimum line cost; and a third processing step of determining the communication line capacity between nodes of the communication network to be processed by combining.

[Industrial application field]

The present invention provides a communication network line capacity that makes it possible to efficiently determine, within a practical processing time, a line capacity configuration that achieves optimization of the communication network line cost when communication demands between nodes are given. It concerns the method of determination.

With the recent development of computer networks, there is a demand for communication network designs that minimize line costs. When designing such a communication network, even if the communication network becomes complex, decisions must be made efficiently within a practical processing time.

[Prior art] For example, for a line between Tokyo and Nagoya (distance 11,269 km), a line of 384 kb/s (64 kb/s equivalent)
A 768kb/s line (12 lines at 64kb/s conversion) was used as the line between Tokyo and Kyoto (distance 374km), and a line between Nagoya and Kyoto (distance 1[101km) was used.
06 as a line of 384kb/s (64kb/s)
Suppose we use 6 pieces (in terms of s). When the monthly usage charges for the line are as shown in Figure 10, the communication demand between Tokyo and Nagoya is 5 lines at 64kb/s, and the communication demand between Tokyo and Tsukabe is 7 lines at 64kb/s. If the communication demand between Nagoya and Tsukabe is 4 lines at 64 kb/s, then
When adopting such a line capacity configuration, the total monthly line usage charge C is: C = 1,700 yen + 2,800 yen + 1, 100 yen = 5,
It will be 600 yen.

On the other hand, the line between Tokyo and Nagoya is 384.
Assuming that you use a kb/s line, a 384kb/3 line between Tokyo and Kyoto, and a 384kb/s line between Nagoya and Kyoto,
For the same communication demand as mentioned above, one line of communication demand between Tokyo and Tsukabe will be routed through Nagoya (there is room for one line between Tokyo and Nagoya, and two lines between Nagoya and Tsukabe). This detour is possible because there is a margin of too yen = 4
, it will be 900 yen.

Furthermore, the line between Tokyo and Nagoya is 768kb/s.
76 as a line between Nagoya and Kyoto.
If an 8kb/s line is used and all communication demand between Tokyo and Tsukabe is routed through Nagoya, the total monthly line usage charge C for the same communication demand as above is C-2, The cost will be 300 yen + 1,400 yen - 3,700 yen, allowing for a significant cost reduction.

As described above, the line cost of a communication network varies greatly depending on the type of line capacity of the communication network.

Conventionally, when the communication demand between nodes is given, in order to determine the line capacity form that optimizes the line cost of the communication network, possible solutions are counted and the line cost is the lowest among them. The methods used were to identify the line capacity type, and to identify the line capacity type using a heuristic method.

[Problem to be solved by the invention]

However, with methods such as the conventional technology, if the network structure of the communication network is relatively simple, it is possible to determine a line capacity configuration that achieves optimization of line costs within a practical processing time. When the network structure becomes complex, the search space becomes enormous, and there is a problem in that it is impossible to determine a line capacity form that optimizes line costs within a practical processing time.

The present invention has been made in view of the above circumstances, and is a method for determining the line capacity of a communication network, which enables the line capacity form to be efficiently determined within a practical processing time to realize the optimization of line costs. It is intended for the purpose of providing.

[Means to solve the problem]

FIG. 1 is a diagram showing the principle configuration of the present invention.

In the figure, Pl is the first treatment process, P2 is the second treatment process,
P3 is the third processing step.

In the first processing step P1, the network structure of the communication network to be processed is divided according to two connected components, thereby dividing the communication network into a plurality of partial networks.

In the second processing step P2, processing is performed to determine the communication line capacity between the nodes of the partial network obtained in the first processing step PL so as to give the optimum line cost.

In the third process P3, the communication line capacities obtained in the second process P2 are combined to determine the communication line capacity between the nodes of the communication network to be processed.

[Effect]

In the present invention, in the first processing step P1, a node (hereinafter referred to as a joint point node) that will cause the communication network to be disassembled into two or more if the node is removed from among the nodes that constitute the communication network is searched. At the same time, the communication network is divided into partial networks according to the searched joint point nodes. When the network is divided into partial networks in this way, the communication demand given between node α and node β, which are a communication node pair, is
If one or more joint point nodes that exist between these nodes are represented by r = (i-1 to n), then the communication demand between node α and joint point node T1, and the joint point node T8
and the communication demand between the joint point node T and the node β, and the communication demand between each rR1fi point node is as follows. The communication demand will be met by adding power to all communication node pairs.

In the second processing step P2, by allocating the given communication demand to each bus of the divided partial network in such a way that the line cost is minimized, the line capacity configuration of the divided partial network is Determine. This process of determining the line capacity type is executed for a small partial network, unlike the original communication network, so it can be realized in an extremely short processing time.

In the third processing step P3, a bus that realizes communication between a pair of communication nodes of the communication network to be processed is obtained, and a second processing step P2 is applied to this bus, for example, to a bus constituting this bus. The line capacity form of the communication network to be processed is determined by allocating the line capacity of the minimum value among the line capacities found in .

As described above, according to the present invention, instead of processing the communication network to be processed as it is and determining its line capacity form, the line capacity form of the divided partial network is obtained and the calculation is performed. Since the configuration is configured such that the bandwidth configuration is determined by synthesizing the bandwidth configurations obtained, it becomes possible to determine the bandwidth configuration of a complex communication network within a practical processing time.

〔Example] Hereinafter, the present invention will be explained in detail according to examples.

FIG. 2 shows an embodiment of the present invention. In the figure, IO is a communication network development support device implementing the present invention, which determines the line capacity between nodes of the communication network to be designed; Reference numeral 20 denotes a man-machine interface included in the communication network development support device 10, which executes interaction processing with the user. This communication network development support device 10 includes a human output control unit, an IL processing data reception unit 12, a division processing unit 13, an optimization processing unit 14, and a composition processing unit 15. This input/output control section 11 is
Executing the interface processing with the man-machine interface 20, the processed data reception unit 12 obtains information required to determine the line capacity between nodes,
Division processing unit 13, optimization processing unit 14, and composition processing unit 15
determines the line capacity between nodes of the communication network to be designed by applying a predetermined algorithm to the information obtained by the processing data reception unit 12.

Hereinafter, the processing contents of each processing section will be described in detail according to the order of processing data receiving section 12, division processing section 13, optimization processing section 14, and composition processing section 15.

The processed data receiving unit 12 obtains information necessary for determining the line capacity between nodes by interacting with the user. FIG. 3 illustrates the information obtained at this time. As shown in this figure, first, there is a set V of nodes U that serve as communication bases of a communication network, and a link j between the nodes.
! Network structure information of a communication network expressed by a graph G-(V, E) consisting of a set E of (u + v) (u + vεV) is obtained. Here, the graph G is given as a simple graph that does not include any self-cycles or multiple links. Also, link 1 (u, v)
is a selectable line connecting nodes uv, and is a variable whose value is the line capacity (unit: ch), which is an item of the line. A zero value is given as the initial value of this line capacity.

Next, a pair of nodes d(u, v
) (u, vEV), the distance between node leaves r(u, v) (in -), and the cost function of the line cost between nodes u and v as shown in Figure 1O. c(j!
(u, v), r(u, v)). Here, d(
u, v) is the communication demand (unit: ch, 1 ch =
This is a constant having a value of 64 kbps), and this demand value is given from design materials.

Next, by searching the information on the obtained graph G=(V, E), a set P(u→V) of paths f(U→V) connecting node leaves is obtained. Here, the same node is
The set P (u −> v) is obtained by searching for a path (elementary path) that does not vary more than a degree. This path f(u
-+v) is a variable having the amount of communication (unit: ch) as a value, and a zero value is given as the initial value of this amount of communication. Note that f(u−+v) may be written as f(u, . . . , v) by arranging the passing nodes in order.

In this way, the set V of nodes, the set E of links, the path node set P (u, v), the node open circuit 11r (uV), the cost function c (j! (u, v), r (u, V)) and the set of demands between nodes are obtained, the communication network line cost optimization problem can be formulated as the following problem N. In other words, [interval BN] When the graph G-(V, E) and the set of demands of communication node pairs are given, satisfy the following conditions and minimize the value of the objective function of path value (traffic value)
seek.

■Demand conditions d(u,v)#Σf(u-+v) However, ・Vd(u,v)εD ・K:f(u-+v)ep(u-+v)■Line conditions j! (u, v) = u (u, v) + δ (u, v) However,・
Vj! (u + v) εE・δ(u, v) is the line capacity margin, and δ(u, v)>0°・u(u+V) is the capacity of all the lines passing through Il(u + V). Total amount of communication on the path.

■Objective function C = Σc (j! (u, v), r (u, v)) However, ・L
ff1(u,v)εE Here, this "■demand condition J is defined as It describes the condition that the sum of the amounts must be the amount of communication demand, and "■ line conditions" is the link l (u, v
) describes the condition that the line capacity of ``■ The objective number is expressed as the graph G-(V
E) represents the total line cost C.

When the processing data reception unit I2 obtains the information necessary for determining the line capacity between nodes, the division processing unit 13
In order to solve the communication network line cost optimization problem in a practical processing time, by dividing the communication network graph C-(V, E) into two connected components,
During this time! Divide IN into subproblems.

If there is an elementary cycle that passes through any two distinct links in a graph, then the graph is 2-connected (as a special case, a graph with only one node or a complete graph with 2 nodes is also 2-connected). ), such that the graph G = (
A subgraph CJI-(Vl, E) which is two-connected
1) It is known that it can be arbitrarily divided into (i = I + 1 is a set of integers), and this subgraph C,i is
It is called the two connected precepts (non-redistribution component). and,
The whole of the two connected precepts of the graph G = (V, E) Gi-(Vi
, Ei) (i E I ), when VinVj1≦l (i, jer, i+j) holds, and VinVj= (v) (i, jεI, i+j),
This V is called the joint point of G. In other words, a joint point means that if you remove that joint point from graph G, graph G becomes 2.
This is the point where it ends up being broken down into more than one thing. Note that for a 2-connected graph with 3 or more nodes, any different 2
There are at least two elementary buses connecting two nodes that share no intermediate nodes.

The division processing unit 13 firstly receives the processing data reception unit 12.
The graph of communication network G=(V.

E) is given, for example, in the document "Co-translated by Nozaki Noshita, Algorithm Design and Analysis I, pp, 166169.
, Science Inc. (1979) This given graph G = (V, E) according to the algorithm described in J.
Divide into two connected precepts.

According to this division process, the communication network graph (:,=(V,E) is divided into two connected precepts, as illustrated in FIG. 4.Hereinafter, for convenience of explanation, nodes U and A bus f (y −1−W ) connecting W is connected to two buses f (u → V) and f (v → W) by a relay node V.
When decomposing into
), and conversely, when two buses are connected by a relay node, it is also written in the same way.

Therefore, the graph G-(V, E) is divided into two connected disciplines Gt=(V
t, Et) (i=I+ I is a set of integers), when uEVi, vεVk, ink, any bus f(u-+v) between u and v can be divided into an appropriate joint point ac(cε■ )
using f (u-+v)= f (u-+a1)+ f (a
It will be written as l→a2)+...+f (ak→V).

Corresponding to this division process, the communication demand d(uV) can be redefined for each divided two-linked precept. That is, the demand condition of the original problem N is d (u, v)
=Σ[(u-+y) However, Kmi f(u-+v)εP(u
-4v), then d(u,v)=d'(u,a1)=d'(aLa2)=
...=d'(ak, v), the demand condition can be redefined as follows.

d'(ak, v)=Σf(ak-+v) However, -Klmif(u=a1)εP(u-+a1)・K2mif(
al-a2)εP(at→a2)・Kvmif(ak-
+ v ) E P (ak−+ v ) Through this division process, if the number of elements included in the set of buses P(u−+v) is expressed as IP(u−”v)l, then the number of buses to be searched is The total number is P (u −+a1) I X I P (al−+
a2) From f X...X P(ak-+v), after division,
a2) l +...10 P(ak-+v).

The division processing unit 13 divides all communication node pairs d(u,v)
This division process is executed for ED.

When dividing into two-connected commands in this way, the demands of multiple communication node pairs may share the same set of buses, but the bus demand is divided into two-connected commands without distinguishing them. Define. Therefore, every communication node pair d
We will decompose the demand one after another for (u, v) εD, superimpose it on the previous results, and define a new demand for each two-connected precept by taking the sum of the decomposed demands. Become.

At this time, the paths shared within the two connected precepts are defined without distinction, so the number of paths to be searched is further reduced than the above.

That is, the division processing unit 13 performs processing to obtain new demand conditions according to the following algorithm.

for all d(u, v)εD ・f (u-+v)= f (u-+a1)+ f (
It is decomposed as al-+a2)+...+f(ak-*v). However, ac(cεI) is the joint point.

・d' (u, a1) = d' (u, a1) + d (u, v)
・d'(al, a2)=d'(al, a2)+d(u,
v)-d'(ak, v)←d'(ak, v)+d(u,
v) FIG. 5 shows a flowchart of the processing of the division processing unit 13 described above.

In this way, we can convert the graph G-(V, E) into two connected precepts
i = (Vi, Ei), and the communication node pair Di = (d(ui, vi); ui, vi6:Vil
If we define, the interval I3! N can be redefined as inter-part RNi below. That is, [Subproblem Ni] When the graph Gi = (Vi, Ei) and the set Di of demands of communication node pairs are given, the following condition is satisfied,
Find the path value (communication amount value) that minimizes the value of the objective function of (1) and (2).

■ Demand conditions d 1 (ui, vi) = Σfi (ui - + vi) However, ・di (ui, vi) εDi ・Ki = fi (ui - + vi) EP (ui + vi) ■ Line conditions 1 (ui, vi ) = u (ui, vi) 10 δ (u t +
V t) However, ・VIl(ui, vi)EEi ■Objective function Ci=Σc(j!(ui, vi), r(ui, vi))
However, Li = ffi (ui, vi) εEi When the optimization processing unit 14 is given this subproblem Ni according to the processing of the division processing unit 13, as shown in the flowchart shown in FIG. Demand di(ui, vi)εDi
(i is omitted in the figure for simplicity) and all paths fi(ui-”vi)EP(ui-
+vi), when the value of the communication amount of the path is increased by a unit amount (for example, 1ch), select one path that minimizes the increase in the objective function, and increase the communication amount of that path by that unit amount. increase only. Then, by performing this process on all demands d i (u i + v i) εD, each subproblem Ni is solved. Since this subproblem Ni targets the divided graph Gi = (Vi, Ei), the communication of each path can optimize the line cost with a significantly shorter processing time than the problem N before division. It will be possible to determine the amount.

The composition processing unit 15 is a part that determines the path value for the original inter-part INN by combining the path values for the inter-part MIN i calculated by the optimization processing unit 14, and Determine the communication amount of the path in the original problem for each node pair demand d (u, v). It is necessary to combine the communication amounts of the zero-divided paths at the joint points to determine the communication amount of the original path. However, there are multiple ways to allocate traffic to the original path, and the result is the same no matter which way you use it, so here we will use the seventh method.
As shown in the flowchart of the figure, select the unprocessed demand d (u + v ) εD and all f(u
→V) For εp (u-+ y), the demand condition d(u v) = Σf(u-+v) However, until K: f(u-+v)EP(u-v) is satisfied , for example, f (u-+v)= f (u-+a1)+ f (a
When divided into l-+a2)+...10f (ak→■), f(u-+v)=min[f(u-+a1)s=,
f (ak-+v)]f (u-*a1)←f (u
-+a1) -f (u-+v)f (al-+a2)
←f (al-+a2) -f (u -> v)f
(ak-+v)←f (ak-+v)-f (u-+
v) and repeat this process for all demands d(u
, v) εD to determine the path value for the original problem N.

Next, the above-described processing of the present invention will be specifically explained with reference to FIG. Here, as shown in FIG. 8(a), in this explanation, V= (1, 2, 3, 4, 5,
6,71E= [I! (1,2),! (1,3), 11
．． 7).

12.3), 12.7), I! , (3,4).

13.5), j! (3,6), 13.7)ffi(4,
'5), j! (4,6), f(5,6)), and assume a graph G = (V, E), and as a set of demands of communication node pairs, D-(d(1,5), d(1, 4)) Assume.

Problem N in the case where it is not divided into two connected precepts is shown in Figure 9. In the conventional technology, this problem N was to be solved.

In the present invention, according to the processing of the division processing unit 13, this graph G-tension (V, E) is expressed as follows: ), 11.3),! ! , (1
, 7).

l(2,3),! ! , (2,7), j! (3,7)1■
G2=(V2.E2) V2=(3,4,5,6) E2=+1(3,4),13.5), l(3,6).

fi(4,5), l(4,6), l(5,6)), G1=(VLE1),! :G2-(V
2. E2). The set of communication node pairs in each two-connected precept is D1=(dHl, 3), respectively.
l, D2-(d2(3,5Ld2(3,4)).

Then, according to the processing of the division processing unit 13, d (1, 5
), any path f between node 1 and node 5
(1 → 5) is divided into f (1 → 5) = f (1 → 3) + f (3 → 5), and d (1) (1, 3) - d (1, 5) d (3 , 5)=d(1, 5), the following demand conditions regarding the partial problem are obtained as shown in FIG. 8(b).

d(1)(L3)=Σf(1→3) =f(L2,3)+f(1,2,7,3)+f(1,3
) + f (1, 7, 2, 3) + f (1, 7, 3) d (3, 5) = Σf (3 → 5) = f (3, 4, 5) + f (3, 4, 6, 5)+f(3,
5)+f(3,6,4,5)+f(3,6,5) On the other hand, for d(1,4), any path f(1→4) between node 1 and node 4 But f (1→4)= f
(1 → 3) + f (3-4), d" (1, 3) - d (1, 4) d (3, 4) = d (1, 4), and as shown in Figure 8 ( As shown in C), the following demand conditions regarding the subproblem are obtained.

a'' (1, 3) - Σf (1 → 3) - f (1, 2, 3) + f (1, 2, 7, 3) + f (1,
3) +f(1,7,2,3)+f(1,7,3) d(3,4)=Σf(3→4) =f(3,4)−+4(3,5,4) +f (3, 5, 6, 4) + f (3, 6, 4) + f (
3, 6, 5, 4) Since the set of paths for d(1)(1,3) and the set of paths for d"(1,3) are the same, the division processing unit 13 calculates the We do not treat them separately in this section.From this, we obtain the following two subproblems Nl, 2 according to this division into two connected precepts.

[Subproblem N1] ■Demand condition dHl, 3) = d”+(1,3)+d”(13), di(1,3)=Σf1(1→3) =fl(1,2,3) ) + fH1, 2, 7, 3) 10 fH1
,3)+fl(1,ma,2.3)+fH1,7,3) ■Line conditions j! (ul, v1) = u (ul, v1) + δ (ul, v
1) However, −V l (ul, v1)εE1・δ(
ul, v1)>0 ■ Minimization of the value of the objective function C2=Σ c (1(ul, v1), r (ul,
v1)) -+min However, -L1=4! (uLv1)ε
E1 [between parts 11iN2) ■ Demand conditions d2 (3, 5) = d (3, 5), d2 (3, 4) = d (
3,4), d2(3,5)=Σ[2(3→5) =f2(3,4,5)+f2(3,4,6,5)+f2
(3,5) ten f2(3,6,4,5)+f2(36,5
) d2 (34) = Σf2 (3 → 4) = f2 (3, 4) + f2 (3, 5, 4) + f2 (3, 5
,6,4)+f2(3,6,4)−+−r2(3,6,
5, 4) ■Line conditions j! (u2. v2) = u (u2. v2) + δ
(u2.v2) However, -V l (u2.v2) εE
2・δ(u2.v2)>0 ■ Minimization of objective function value CI=Σc (i! (u2. v2), r (u2
．． v2)) −+厘fnHowever,・C2mif(u2.v2
) εE2 optimization processing unit 14 calculates d(1→5)=10. For the design condition of d(1→4)=5, f(1,2,3)=3. f (1, 2, 7, 3)
=5゜f(1,3)=L f(1,7,2,3
)=3f(1,7,3)=3. f(3,4,5
)=Lf(3,4,6,5)=3. f(3,5)=
4°f(3,6,4,5)=1. f(3,6,5)
=Lf(3,4)=0. f(3,5,4)=
3°f (3, 5, 6, 4) = 1. f(3,6,4)
Assume that a path value of =0°f(3,6,5,4)=1 is obtained.

The synthesis processing unit 15 converts all paths f(1→5)εP(1
→5), until d(1,5)=Σf(1→5) is satisfied, for the division of f(1→5)=f(1→3)+f(3→5), f (1→5)=sjn[f (1→3n, f(3→5
)1f(1→3)←f(1→3)−f(1→5)f(3
→5) Execute the calculation of ← f (3 → 5) - f (1 → 5),
Also, until all paths f(1→4)6P(1-4) satisfy d(1,4)=Σf(1→4), f(1→4)=f(1-3 )+f(3-+4), f(1-4)=min[f(1-3), f(3-
4)]f(1→3)←f(1→3)−f(1→5)f(
By executing the calculation 3→4)←f(3→4)−f(1→5), the path value for the original interval aN is determined.

That is, if f(1→5) is explained using the path value obtained by the optimization processing unit 14 described above, ■・f(
1,2,3,4,5) =4in[f(1,2,3),f(3,4,5))--
in[3,1]=1 ・f (1,2,3)-f (1,2,3)~1-2-
f(3,4,5) ←f(3,4,5)-1=0■・
f(C2,3,4,6,5) -min[f(1,2,3),f(3,4,6,5)]
=win[2,3]=2 ・f(1,2,3)←f(1,2,3)−2=0−
f(3,4,6,5) ←f(3,4,6,5)-2
=1 ■ f (1, 2, 3, 5) = -4n [f (1, 2, 3), f (3, 5) 1
-min[0,4]=0 ■・f (1,2,3,6,4,5) =min[f(1,2,3),f(3,6,4,5)]
=0 ■・ f(1,2,'3,6.5) -min[f(1,2,3),f(3,6,5)]=
0 ■・ f (1, 2, 7, 3, 4, 5) = minlf (1, 2, 7, 3), f (3, 4, 5)]
- Garden in [5, O] = 0 By executing such processing also for r (1 → 4), the path value f (1, 2
, 3, 4, 5) = 1 ° f (1, 2, 3, 4, 6, 5) = 2 ° f (1, 2, 7, 3, 4, 6, 5) = 1 ° f (1, 2,
7, 3, 5) = 4°f (1, 3, 6, 4, 5) = 1°f (1, 7, 2, 3, 6, 5) - 1°f (1, 7, 2, 3 , 5, 4) = 2°f (1, 7, 3, 5, 4) = 1°f (1, 7, 3, 5, 6, 4) = 1°f (1, 7, 3
, 6, 5, 4) = 1 (other path values are zero).

If this specific example problem N (the problem shown in Figure 9) is solved without dividing, the number of paths to be searched is the demand d(
25 pieces of 5×5 for 1,5) and demand d(1,4
), we have to explore a search space made up of 50 paths, which are the sum of 25 5×5 values for ), while thus inter-part i! When dividing into N1 and 2, in subproblem N1, the set of paths used for demands d(1,5) and d(1,4) are the same, so there is no need to distinguish between them, so the search space created by five paths is In subproblem N2, the number of paths used for demand d(1,5) is 5, and the number of paths used for demand d(1,4) is 5, which is the total value of 10 paths. Since it is necessary to search the created search space, it is necessary to search the search space created by a total of 15 paths. In this way, by using the present invention,
This allows the search space to be significantly compressed.

Although the illustrated embodiment has been described, the present invention is not limited thereto. For example, in the embodiment, a link is expressed as a pair of nodes ff1(u,v), but if a branch line is used, the link is expressed as a list of nodes)l(u,...
V), and the present invention can be directly applied according to a similar formulation. Furthermore, in the embodiment, the line is described as a two-way communication, but it may be a one-way communication, and an effective graph (a graph in which the direction of the link is defined)
By formulating using , the present invention can be applied as is.

(Effects of the Invention) As explained above, according to the present invention, a line capacity configuration that realizes optimization of the line cost of a communication network within a practical processing time when the communication demand between nodes is given. be able to decide. From now on, it will be possible to significantly reduce the man-hours required to design communication networks.

[Brief explanation of drawings]

Fig. 1 is a diagram of the principle configuration of the present invention, Fig. 2 is an embodiment of the present invention, Fig. 3 is an explanatory diagram of information obtained by the processing data reception unit, and Fig. 4 is an explanation of division into two connected components. 5 is a flowchart executed by the division processing section, FIG. 6 is a flowchart executed by the optimization processing section, FIG. 7 is a flowchart executed by the composition processing section, and FIG. 8 explains the processing contents of the present invention. Figure 9 is an explanatory diagram for explaining problem N when dealing with the graph in Figure 8 using conventional technology.
The explanatory diagram, Figure 10, is an example of line usage charges. In the figure, IO is a communication network development support device, 20 is a man-machine interface, 11 is a human output control unit, and 12
13 is a processing data reception section, 13 is a division processing section, 14 is an optimization processing section, and 15 is a composition processing section.

Claims (1)

[Claims] Line capacity of a communication network for determining the line capacity between each node that will achieve optimization of line cost when communication demand between nodes configuring the communication network is given. A determination method, comprising: a first processing step (P
1), a second processing step (P2) in which the communication line capacity between the nodes of the partial network obtained in the first processing step (P1) is determined so as to give an optimal line cost; and a third processing step (P3) for determining the communication line capacity between nodes of the communication network to be processed by composing the communication line capacities obtained in the second processing step (P2). A method for determining the line capacity of a communication network.