JP2013021678A - Transport line design method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an efficient line design method and program for minimizing the number of wavelengths in consideration of a permissible delay amount.SOLUTION: For each of all demands, a delay restriction amount is calculated from a straight-line distance between a start point node and a terminal point node and a permissible delay time, and a path is set to a demand having the delay restriction amount more than a constant value. Next, in the case that the permissible delay time is satisfied when a demand having the delay restriction amount equal to or less than the constant value is aggregated to the path that was set above, the demand is aggregated to the path. Finally, a network is divided into a plurality of sections, and a node to be a source point is decided to each of the divided sections. A path is set between the source point nodes, and a demand that was not aggregated above is aggregated to the path.

Description

  The present invention relates to an efficient circuit design method and program for minimizing the number of wavelengths in consideration of a maximum allowable delay required by a customer, and relates to an OTN (Optical Transport Network) or MPLS-TP (Multi Protocol Label Switching-Transport Profile). The transport network is composed of nodes that can accommodate and aggregate such lines.

  Conventionally, a technique for consolidating a plurality of paths into a single wavelength has been proposed for effective use of wavelengths (Non-Patent Documents 1-5). Non-Patent Document 2-5 is a topology limited to a ring-type network. Non-Patent Document 1 is a method that can be flexibly applied to a mesh network.

I.Yagyu et.al., "An EfficientHierarchical Optical Path Network Design Algorithm based on a Traffic DemandExpression in a Cartesian Product Space," IEEE Journalon Selected Areas In Communications, Vol.26, No.6, pp.22-31, August 2008. A..L.Chiu et al., "TrafficGrooming Algorithm for Reducing Electronic Multiplexing Costs in WDM RingNetworks," Journal of Lightwave Technology, Vol.18, No.1, pp.2-12, January2000. O.Gerstel, et.al., "Cost-Effective Traffic Grooming in WDM Rings," IEEE / ACM Transactionon Networking, Vol.8, No.5, pp.618-630, October 2000. X.Zhang et.al., "An Effective and Comprehensive Approach for Traffic Grooming and Wavelength Assignment in SONET / WDM Rings," Vol.8, No.5, pp.608-617, October 2000. J. M. Simmons et al., "Quantifying the Benefit of Wavelength Add-Drop in WDM Rings with Distance-Independent and Dependent Traffic," Journal of LightwaveTechnol.Vol.17, No.1, January 1999.

  However, these methods have a problem that the allowable delay amount in each traffic demand is not considered.

  Accordingly, an object of the present invention is to provide an efficient circuit design method and program for minimizing the number of wavelengths in consideration of an allowable delay amount.

  In order to achieve the above object, the circuit design method according to the present invention is a circuit design method for a demand having an allowable delay time. For all demands, the delay constraint amount is determined from the distance between the start node and the end node and the allowable delay time. A first step of setting a path for a demand whose delay constraint amount exceeds a certain value, and aggregating the demand whose delay constraint amount is a certain value or less into the path set in the first step When the same wavelength is used, if the allowable delay time is satisfied, the second step of consolidating the path, the network is divided into a plurality of partitions, and a starting node is determined in each divided partition, And a third step of setting a path between the origin nodes and aggregating demands that could not be aggregated in the second step in the path.

In the first step, the delay constraint amount is obtained from the distance between the start node and the end node and the allowable delay time for all demands, and the demand in which the delay constraint amount exceeds a certain value is registered in the list 1 A first sub-step, a second sub-step for determining a demand with the strictest delay constraint out of the demands registered in the list 1, and setting the shortest path for the demand, and registration in the list 1 Among the set demands, the demands where the start point node and the end point node exist within a certain distance from the passing node of the set path are aggregated to the path with the strictest delay constraint, and the same wavelength is used. In this case, it is confirmed whether or not the allowable delay time is satisfied, and if it is satisfied, the third substep is aggregated in the path with the strictest delay constraint and the third substep is aggregated. Demand registered in never been said list 1 as a new list 1, and a fourth sub-step of repeating said to said second sub-step the third sub-step the aggregation relationship is not generated,
In the second step, the start node and the end node are within a certain distance from the passing node including the start and end points of the path set in the first step in the demand in which the delay constraint amount is a predetermined value or less. When existing demand is aggregated in the path and the same wavelength is used, whether or not the allowable delay time is satisfied is confirmed, and if satisfied, the demand is aggregated in the path and the demand that cannot be aggregated is registered in the list 5 Is a substep of
In the third step, the network is divided so as to pass through as many lines as possible between the start and end nodes of demand registered in the list 2, and a starting node is determined for each divided section. Then, a path is set between the origin nodes, demand is aggregated in the path, and if the same wavelength is used, it is confirmed whether or not an allowable delay time is satisfied. If all demands cannot be aggregated by the sub-step and the sixth sub-step, draw a dividing line that intersects the line of the sixth sub-step, divide the network into four, and start from each divided section If the same wavelength is used, whether or not the allowable delay time is satisfied is confirmed, and if it is satisfied, the path is set to the path. 7th subs to be aggregated If all demands cannot be aggregated by the seventh sub-step, the dividing line intersecting with the line of the sixth sub-step is redrawn, and the number of divisions is increased while increasing the number of divisions to 6, 8 and It is also preferable to have an eighth sub-step for repeating the processing of the seventh sub-step.

  In the fifth sub-step, if an OTN node exists within a certain distance from the start point or end point of the demand and the start point or end point of the demand is accommodated in the OTN node, whether or not an allowable delay time is satisfied It is also preferable to further comprise aggregating the start point or end point of the demand to the OTN node when confirming and satisfying.

  It is also preferable to divide the seventh sub-step and the eighth sub-step so that the number of nodes is as equal as possible in each new partition.

  As another method of the third step, a demand with the longest distance between the start node and the end node is extracted from the demands registered in the list 2, and the start nodes of the demand are collected. A, the end node is included in the set B, and the other demand is divided into networks so that the end point (start node or end node) closer to the start point node of the demand belongs to the set A and the other to the set B. Determine the starting node for each divided section, set the path between the starting nodes, aggregate demand in the path, and check whether the allowable delay time is satisfied when the same wavelength is used, If all the demands cannot be aggregated by the sixth substep aggregated in the path and the sixth substep, the sets A and B of the sixth substep are the start points belonging to the same set. Node end point The network is divided into four so that the start node and the start node of the end node satisfy the conditions belonging to another same set, a node that is the starting point is determined in each divided partition, and a path is set between the starting nodes Then, when the demand is aggregated in the path and the same wavelength is used, it is confirmed whether or not the allowable delay time is satisfied, and if so, the seventh sub-step for consolidating the path and the seventh sub-step If all demands cannot be aggregated by the above or cannot be divided so as to satisfy the above conditions, re-divide into sections satisfying the conditions of the sixth sub-step, increasing the number of divisions to 6 and 8 while increasing the number of divisions. It is also preferable to be an eighth sub-step in which the processing of the seven sub-steps is repeated.

  The delay constraint amount is preferably a value obtained by dividing a straight line distance between a start node and an end node by an allowable delay time.

  The shortest path is preferably determined by the Dijkstra method with delay time as a path cost.

  The starting node is in the partition, and the sum of the straight line distances from the starting and ending nodes of demand registered in the list 2 is minimized, or the sum of the number of hops is minimized. A node is also preferable.

  In the seventh sub-step and the eighth sub-step, if the number of wavelengths after aggregation + the number of wavelengths that could not be aggregated <the number of wavelengths when all paths were aggregated, the path that could not be aggregated It is also preferable to assign a wavelength independently for.

Further, after the completion of the third step, when the total bandwidth after path aggregation is equal to or greater than the bandwidth of one interface,
Among all the demands scheduled to be aggregated, among the demands for which the delay constraint amount is not more than a certain value, there are passing nodes of the same existing wavelength within a certain distance from the starting point and the ending point of the demand. When accommodated in the path, it is also preferable to accommodate the demand satisfying the allowable delay time in the path of the passing node so that the total bandwidth after the path aggregation is less than the bandwidth of one interface.

  In addition, it is also preferable that the accommodation is performed in order from the demand with the largest delay constraint amount among the demands in which the delay constraint amount is a certain value or less until the total bandwidth after path aggregation becomes less than the bandwidth of one interface. .

  In order to achieve the above object, the program according to the present invention provides a computer for designing a line for a demand having an allowable delay time, and for all the demands, delay constraints are determined from the distance between the start node and the end node and the allowable delay time. Means for setting a path for a demand whose delay constraint amount exceeds a certain value, and aggregating the demand whose delay constraint amount is a certain value or less into the path set by the means, If the allowable delay time is satisfied, the network is divided into a plurality of sections, the node that is the starting point is determined for each divided section, and the path is set between the starting nodes. And it is made to function as a means to aggregate the demand that could not be aggregated by the means on the path.

  By applying the path aggregation method of the present invention, it is possible to design the metro core optical network in consideration of the maximum allowable delay amount for each service and to minimize the number of wavelengths. It contributes to the reduction of equipment cost.

  Further, in the present invention, by designing the route so that a plurality of demands merge at the start point and the end point of the demand path, it is possible to suppress an increase in the number of OTN nodes and an increase in cost. Furthermore, when the total bandwidth after path aggregation is equal to or greater than the bandwidth of one interface, it is possible to reduce the number of wavelengths by concentrating demand on other wavelengths.

2 shows a flowchart according to the invention. An example of implementation of step 1 is shown. An example of implementation of step 2 is shown. An example of implementation of step 3 is shown. An example of implementation of step 4 is shown. An example of implementation of step 5-1 is shown. An example of implementation of step 5-2 is shown. An example of the implementation of step 5-3 will be shown. An example of implementation of step 6 is shown. The example of implementation of step 7-1 is shown. An example of the implementation of step 7-2 is shown. An example of the implementation of step 7-3 is shown. Another example of implementation of step 6 is shown. Another example of implementation of step 7-1 is shown. The other example of implementation of step 7-2 is shown. The other example of implementation of step 7-3 is shown. The flowchart of the method of suppressing the increase in the number of OTN nodes is shown. An embodiment of a method for suppressing an increase in the number of OTN nodes will be described. 2 shows a flowchart of a method for reducing the number of wavelengths. An example of a method for reducing the number of wavelengths will be described.

  The best mode for carrying out the present invention will be described in detail below with reference to the drawings. FIG. 1 shows a flowchart according to the invention. Hereinafter, description will be given based on this flowchart.

Step 1: Select all demands with severe delay constraints. At that time, first for all demand _{n n,} start node (S: Source) and the end node (D: Destination) obtains a linear distance (d (Sn, Dn)) between the maximum allowable delay the linear distance delay constraint amount is divided by the l _n a _{(x n)} is calculated. If the value exceeds a certain level α, the corresponding path is registered in the list 1-1 as a path with severe delay constraints.

FIG. 2 shows an example of step 1 implementation. In this example, it is assumed that there are 13 demands from demand n ₁ to demand n ₁₃ . For all demand,
x _n = (d (Sn, Dn)) / l _n (n = 1 to 13)
Is calculated. In this example, since x ₁ to x ₅ exceed a certain level α, these demands are registered in the list 1-1. FIG. 2 shows the linear distance of these demands.

Step 2: In list 1-1, a demand that maximizes _xn is selected, and a path for the demand is determined by the Dijkstra method. The path cost at that time is a delay amount.

FIG. 3 shows an example of step 2 implementation. In this example, the demand for list 1-1, those maximum of x _n are assumed to be x _1. In other words,
x ₁ = max (x ₁ , x ₂ , x ₃ , x ₄ , x ₅ )
And In this case, a path is determined for the demand n ₁ by the Dijkstra method. FIG. 3 shows the determined paths from S1 to D1.

Step 3: Among demands n _i registered in Step 1, k is a constant threshold value for the demands n _m set in Step 2.
d (Si, Sm) <k and d (Di, Dm) <k,
Or d (Si, Dm) <k and d (Di, Sm) <k
Demand n _i is selected, and it is confirmed whether or not each allowable delay time can be satisfied when it is collected in the path determined in step 2.

  Demand that can satisfy the allowable delay time is aggregated in the path determined in step 2 and cannot satisfy the allowable delay time, and d (Si, Sm) <k and d (Di, Dm) <k, or d Demands that do not correspond to either (Si, Dm) <k and d (Di, Sm) <k are put on hold, and these demands are registered in the list 1-2.

  However, even if d (Si, Sm) <k and d (Di, Dm) <k or d (Si, Dm) <k and d (Di, Sm) <k are not satisfied, Si and Di are stepped. Demand that exists on the path determined in 2 and can satisfy the allowable delay time can be aggregated.

FIG. 4 shows an example of step 3 implementation. The demand set in step 2 corresponds to demand n ₁ . Among the demands n ₂ to n _{5 in} the list 1-1, the demand n ₂ is d (S2, S1) <k and d (D2, D1) <k, and satisfies the allowable delay time. to accommodate the path of the demand n _1. Moreover, the demand _{n 5} is, S5, D5 together in the path of the demand _{n 1.} Assuming that the allowable delay time can be satisfied, it is accommodated in the path of demand n ₁ . Demand n ₃ and demand n ₄ are put on hold and registered in list 1-2.

  Step 4: The demand in list 1-2 is changed to list 1-1 in step 2, and the procedure from step 2 to step 3 is repeated. These are repeated until no more aggregation relationship occurs.

FIG. 5 shows an example of step 4 implementation. Listing 1-2 is entered in step 2,
x ₄ = max (x ₃ , x ₄ )
As shown in FIG. 5, the path is determined in step 2 for the demand n ₄ . Thereafter, in step 3, d (S4, S3) <k and d (D4, D3) <k, and the demand n ₃ is finally accommodated in the path of the demand n _{4 on} the assumption that the allowable delay time can be satisfied. In addition, the demands n ₁ , n ₂ , and n ₅ are concentrated on the same wavelength λ 1, and the demands n ₃ and n ₄ are concentrated on the same wavelength λ 2. In this example, no further aggregation relationship occurs, so the process proceeds to the next step 5.

Step 5: It is examined whether or not the demand satisfying x _i ≦ α in Step 1 can be aggregated in the path determined up to Step 4. First of all, if the start and end points correspond to the passing nodes of the path determined up to step 4, aggregation is performed. Next, from the passing node (Sn, Dn) including the start and end points of the path determined in step 2 where Si, Di of demand satisfying x _i ≦ α, d (Si, Sn) <k and d (Di, Dn) < The determination is made based on whether k or d (Si, Dn) <k and d (Di, Sn) <k. Those that do not meet the above conditions, or those that cannot be aggregated into the existing path with an allowable delay time even if they meet, are registered in the list 2.

6, 7 and 8 show an example of the implementation of step 5. Of the 13 items from the demand n ₁ to the demand n ₁₃ , the demands that satisfy x _i ≦ α were the demand n ₆ to the demand n ₁₃ . As shown in FIG. 6, since the demand n ₆ is a passing node on the path having the start / end point (S6, D6) of the wavelength λ1, the demand n ₆ is aggregated to the wavelength λ1 if the allowable delay time is satisfied. Further, as shown in FIG. 7, since the demand n ₇ is located within k from the passing node of the path of wavelength λ1, if the allowable delay time is satisfied, the demand n ₇ is collected at wavelength λ1.

However, as shown in FIG. 8, the demand n ₈ to the demand n ₁₃ are not within the pass nodes of the path of wavelength λ1 and the path of wavelength λ2, but within k from the pass nodes of the path of wavelength λ1 and the path of wavelength λ2. These are registered in list 2 because they are not located.

  Step 6: The network graph is divided so as to pass as many straight lines as possible connecting the S / D nodes of the demand listed in List 2. However, lines that divide the graph (hereinafter referred to as dividing lines) are uniquely drawn downward (or upward) so that the dividing lines do not intersect. Here, the dividing line obtained by dividing the entire network into two separates the demands S and D so that they belong to different sets. The node that becomes the starting point of aggregation is determined in the divided sections. The starting node is a node in which the sum of the linear distances from the demand S / D nodes in the list 2 in each section is minimized or the sum of the number of hops is minimized. When a path is set between the nodes of this starting point, demand is concentrated on this path, and the same wavelength is used, if the allowable delay time is satisfied, the path is aggregated on this path. As a result of the aggregation, if all the paths can satisfy the constraints such as delay, the process ends here.

FIG. 9 shows an example of step 6 implementation. In this example, the graph is divided by dotted dividing lines. Here, it is divided so as to pass through all straight lines connecting the S / D nodes of demand n ₈ to demand n ₁₃ . However, in this example, it is impossible to aggregate paths into one wavelength within the same section only by dividing into two.

  Step 7: If all the paths in Step 6 fail to satisfy the constraints such as delay, the second division in the direction intersecting with the dividing line in Step 6 (that is, the direction that is never parallel to the dividing line in Step 6) Draw a line and divide the graph into four. When dividing into four, it divides | segments so that the number of nodes may become as equal as possible in each new division. In the newly created four sections, the starting node is searched in the same manner as in step 6 and aggregation is performed. If all the paths still do not satisfy the constraints, redraw the dividing line that intersects the first dividing line and increase the number of divisions so that the entire graph becomes 6 divisions, 8 divisions, and so on. The same processing is repeated until all the paths can be collected. However, if “number of wavelengths after aggregation + number of wavelengths that could not be aggregated <number of wavelengths when all paths are aggregated” in step 6 to step 7, the number of wavelengths further decreases. Therefore, the processing is immediately stopped without exception, and a wavelength is assigned independently for a path that cannot be aggregated.

  10, 11, and 12 show an example of implementation of step 7. In FIG. 10, the second dividing line is drawn so as to intersect the first dividing line. In this example, even if it is divided into four sections, the demands S and D in the same section are too far apart, so that they cannot be aggregated while satisfying each other's allowable delay time. Therefore, as shown in FIG. 11, two dividing lines intersecting with the first dividing line are drawn again and divided into six sections. When divided into six sections, a plurality of demands can be aggregated to the same wavelength while satisfying the allowable delay time of all demands as shown in FIG.

Specifically, a node indicated by a black dot in FIG. 12 is a starting node in each section. The paths set in the starting node of section 1 and the starting node of section 2 aggregate demand n ₉ and demand n ₁₁ , and the paths set in the starting node of section 3 and the starting node of section 4 are demand n _10. And the demand n ₁₂ are aggregated, and the paths set in the origin node of the section 5 and the origin node of the section 6 aggregate the demand n ₈ and the demand n ₁₃ .

  Regarding Step 6 and Step 7, it is also possible to divide the graph by other methods to aggregate demand. Steps 6 and 7 according to other methods are shown below.

  Step 6: Extract a demand with the longest distance between S / Ds of the demands listed in List 2. If the start point node of the demand is included in the set A and the end point node is included in the set B, for other demands, the end point (start point node or end point node) closer to the start point node of the demand is set A, and the other is B Belonging to. If the aggregated demands pass on the same path, the starting and ending nodes of the longest path are set as starting points.

  As an exception, those having the same distance within the range of error A from S and D are excluded from aggregation targets. For these, the wavelength is individually assigned to each of them, or these are collectively collected as described above, or the number of wavelengths is reduced.

  In each of the sets A and B, a node as a starting point of a path to be aggregated is determined, and a node having a minimum sum of distances from nodes belonging to each set or a minimum number of hops is selected as the starting node. Shall. When demand is concentrated on this path and the same wavelength is used, if the allowable delay time is satisfied, it is aggregated on this path. As a result of the aggregation, if all the paths can satisfy the constraints such as delay, the process ends here.

FIG. 13 shows an example of step 6 implementation. In this example, as in section 1 and section 2, the start point of demand n ₁₂ that is the farthest distance between the start and end points is divided into set A (section 1) and the end point is divided into set B (section 2), and demand n ₈ by dividing start and end points to the partition 1 and 2 also demand other than the above-demand n _13, it shall constitute the set a, the B.

Step 7: If the constraints such as delay are not satisfied for all paths when divided and aggregated into two sections in Step 6, the sets A and B in Step 6 are divided to form n node sets. (N: even number). At this time, the condition is that the end point (start point) node of the start point (end point) node belonging to the same set belongs to another same set. That is, in the example shown in FIG 14, since the start and end points of the demand n ₁₀ and n ₁₂ belong to different sets, is inadequate, and since the distance between demand apart, aggregation is difficult. On the other hand, in FIG. 15, the classification satisfies the above-described conditions. Based on the division of FIG. 15, the starting node in each section (set) is determined as shown in FIG. 16, and the paths are aggregated. Here, if all demands satisfy the delay constraint, the path is completely set. If the condition is not satisfied even if it is divided into 4 sections as in the examples of FIGS. 14 to 16, the number of sections is increased to 6 sections, 8 sections,... Repeat until all the paths can be aggregated. However, if “number of wavelengths after aggregation + number of wavelengths that could not be aggregated <number of wavelengths when all paths are aggregated” in step 6 to step 7, the number of wavelengths further decreases. Therefore, the processing is immediately stopped without exception, and a wavelength is assigned independently for a path that cannot be aggregated.

Specifically, a node indicated by a black dot in FIG. 16 is a starting node in each section. The paths set in the starting node of section 1 and the starting node of section 2 aggregate demand n ₉ and demand n ₁₁ , and the paths set in the starting node of section 3 and the starting node of section 4 are demand n _10. And the demand n ₁₂ are aggregated, and the paths set in the origin node of the section 5 and the origin node of the section 6 aggregate the demand n ₈ and the demand n ₁₃ .

  In the above step, it is assumed that an OTN cross-connect or a network device similar to the above that can aggregate lower speed lines into high speed lines is inserted at the start and end points of all paths. Also, ROADM or WXC can be applied as the layer 0 optical switching node, but WXC is preferentially used for optical switching nodes having four or more routes.

  Further, when the paths are aggregated in the above step, the paths are aggregated in consideration of the upper limit of the number of wavelengths per link. As an example, it is assumed that demands exceeding the upper limit of λn are aggregated and another wavelength λm is allocated. At this time, the path of λm may pass through exactly the same path as λn. However, the path may be set by determining the starting node between the demands aggregated in λm as in steps 6 and 7.

  In addition, after aggregating paths in the above step, if there is a demand that does not satisfy the constraint condition of OSNR (Optical Signal to Noise Ratio), 3R or OTN-XC (universal switch) is set to a node that does not correspond to the start and end points. So that the path satisfies the OSNR constraint.

  In the above embodiment, only one path is displayed for each demand. However, when it is necessary to set both the working path and the protection line for a single demand, the working line is used. The route of the protection line should be determined so that it does not pass the link connected to the relay node (not including the start and end points) and the link connected to the relay node (the path cost is infinite). And

  In addition, when the demands are merged into a single route and aggregated to the same wavelength in Step 5 of the above-described embodiment, it is necessary to consider which station should install the OTN node for each wavelength. However, although an OTN node can be shared among a plurality of wavelengths, the installation of an OTN for each wavelength results in an increase in the number of OTN nodes, resulting in an increase in cost.

  As described above, since all the nodes corresponding to the start and end points of the demand path are OTN nodes, the number of OTN nodes can be determined by designing the route so that the demands merge at the start and end points of the demand path. It is possible to suppress the increase in cost and the increase in cost. FIG. 17 shows a flowchart of a method for suppressing an increase in the number of OTN nodes.

Here, _{x i} ≦ α become demand _{n i} of Si, transit node that Di includes start and end points of the path determined in step 2 (Sn, Dn) from d (Si, Sn) <k and d (Di, Dn ) <K, or d (Si, Dn) <k and d (Di, Sn) <k.

Step 51: An OTN node list (Oi: i = 1 to n, n is the number of OTN nodes) is created in order to grasp the node where the OTN exists. Initially, at least the start and end nodes of all demands are included in the OTN node list.
Step 52: From the starting point Si demand _{n i} to aggregate, 'as certain threshold, d (Si, Oj) <k' k OTN nodes Oj is to be searched for satisfying. If it exists, the process proceeds to step 53. If it does not exist, the process proceeds to step 54.
Step 53: aggregating the starting point of the demand _{n i} in the OTN node Oj.
Step 54: Convert the nearest passing node to an OTN node and include the node in the OTN node list.
Step 55: From the end point Di demand _{n i} to aggregate, 'as certain threshold, d (Di, Oj) <k' k OTN nodes Oj is to be searched for satisfying. If it exists, the process proceeds to step 56; otherwise, the process proceeds to step 57.
Step 56: aggregating endpoint demand _{n i} in the OTN node Oj.
Step 57: Convert the nearest passing node to an OTN node and include the node in the OTN node list.

FIG. 18 shows an embodiment of a method for suppressing an increase in the number of OTN nodes. Si demand _{n i} is assumed to be d (Si, Sn) <k . At the start point Si, there exists an OTN node Oj that satisfies d (Si, Oj) <k ′. For this reason, Si is accommodated in Oj, not Sn.

  In addition, after the demand paths are aggregated in Steps 6 and 7 of the above embodiment, when the total band after the path aggregation is equal to or greater than the band of one interface, another wavelength is assigned in the same route. However, at different wavelengths that are not related to the above, there is a possibility that the sum of the bands after aggregation is less than or equal to the band of one interface and other demands can be aggregated.

For this reason, out of all the demands that were planned to be aggregated, if the quantity of demands with loose delay constraints (demand that was x _i ≦ α in step 1) was grasped and all of these were assigned to different wavelengths, per wavelength It is confirmed whether or not the total sum of the bandwidths is less than the bandwidth of one interface. If the bandwidth is less than one interface band, the strictest delay constraints are aggregated in the order of wavelengths having a passing node within a predetermined distance P from the start / end points of the demand.

FIG. 19 shows a flowchart of a method for reducing the number of wavelengths. The method is performed for all demand paths after steps 6 and 7 have been performed and all demands have been aggregated.
Step 71: It is confirmed whether or not the sum of the bandwidths of demand scheduled to be aggregated exceeds the bandwidth of one interface. If it is less than the band, the process ends because there is no need for wavelength reassignment.
Step 72: When only demands with loose delay constraints are allocated to different wavelengths among demands scheduled to be aggregated, it is confirmed whether or not the sum of the remaining bandwidths of demand is below the bandwidth of one interface. . If it falls below, it is terminated because reassignment to another wavelength is impossible.
Step 73: Numbers (in ascending order) are assigned to the demands with the slow delay constraints in the above steps in the order of severe delay constraints.
Step 74: It is confirmed whether there is a passing node having the same existing wavelength within a predetermined radius P from the start point and end point of the demand in order from the smaller number assigned in the above step (that is, the demand path with severe delay constraints). If none exists, it is terminated because reassignment to another wavelength is impossible.
Step 75: When the demand confirmed in the above step is accommodated in the demand path of the transit node, it is confirmed whether or not the allowable delay time can be satisfied (if all the wavelengths correspond to a plurality of wavelengths, check all the wavelengths with a large free band) Housed in). If there is no route that does not satisfy the allowable delay time, the process ends because reassignment to another wavelength is impossible.
Step 76: The demands confirmed in the above step are removed from the smaller number to which the demand is allocated until the total sum of the bandwidths of all demands becomes less than the bandwidth of one interface, and the wavelength reassignment is completed and ends.

  FIG. 20 shows an embodiment of a method for reducing the number of wavelengths. In steps 6 and 7, the demand d (Sn, Dn) is collected in the demand path λ1. However, when the demand d is aggregated, the sum of bands per wavelength is equal to or greater than the band of one interface. The demand d is a demand with loose delay constraints, and there are passing nodes having the same existing wavelength λn within a predetermined radius P from the start point and the end point of the demand d. Furthermore, if the allowable delay time of the demand d can be satisfied when accommodated in λn, it is possible to reduce the number of wavelengths of the demand path λ1 by accommodating the demand d in λn.

  The transport line design method according to the present invention can be realized by a program that causes a computer to function as the steps described above. These computer programs can be stored in a computer-readable storage medium or distributed via a network. Furthermore, the present invention can be realized by a combination of hardware and software.

  Moreover, all the embodiments described above are illustrative of the present invention and are not intended to limit the present invention, and the present invention can be implemented in other various modifications and changes. Therefore, the scope of the present invention is defined only by the claims and their equivalents.

Claims (12)

In a circuit design method for a demand having an allowable delay time,
A first step of determining a delay constraint amount from the distance between the start node and the end node and an allowable delay time for all demands, and setting a path for the demand in which the delay constraint amount exceeds a certain value;
A second step of consolidating the demand whose delay constraint amount is a certain value or less into a path set in the first step and using the same wavelength and satisfying an allowable delay time when the same wavelength is used. When,
A network is divided into a plurality of partitions, a node as a starting point is determined in each divided partition, a path is set between the starting nodes, and demand that cannot be aggregated in the second step is aggregated in the path. 3 steps,
A circuit design method characterized by comprising: The first step includes
A first sub-step for obtaining a delay constraint amount from the distance between the start node and the end node and an allowable delay time for all demands, and registering a demand in which the delay constraint amount exceeds a certain value in the list 1;
A second sub-step of obtaining a demand with the strictest delay constraint among the demands registered in the list 1, and setting a shortest path for the demand;
Among the demands registered in the list 1, the demands having the start node and the end node within a certain distance from the passing node of the set path are aggregated into the path path having the strictest delay constraint, If the same wavelength is used, it is confirmed whether or not the allowable delay time is satisfied, and if it is satisfied, a third sub-step for concentrating on the path path having the strictest delay constraint,
The demand registered in the list 1 that has not been aggregated in the third sub-step is newly set as the list 1, and the second sub-step and the third sub-step are repeated until no aggregation relationship occurs. Substeps of
Have
The second step includes
Among the demands in which the delay constraint amount is equal to or less than a certain value, for the demands in which the start node and the end node exist within a certain distance from the passing node including the start and end points of the path set in the first step, the path If the same wavelength is used, it is confirmed whether or not the allowable delay time is satisfied.
The third step includes
The network is divided so as to pass through as many lines as possible between the start and end nodes of demand registered in the list 2, and a starting node is determined in each divided section, and between the starting nodes When a path is set, demand is aggregated in the path, and the same wavelength is used, it is confirmed whether or not the allowable delay time is satisfied, and if so, a sixth sub-step for aggregating in the path;
If all demands cannot be aggregated by the sixth substep, draw a dividing line that intersects the line of the sixth substep, divide the network into four, and set the starting node in each divided section. Determining, setting a path between the origin nodes, concentrating demand on the path, and using the same wavelength, confirming whether or not the allowable delay time is satisfied. Substeps of
If all demands cannot be aggregated by the seventh sub-step, the dividing line intersecting the line of the sixth sub-step is redrawn, and the seventh sub-step is increased while increasing the number of divisions to 6 divisions and 8 divisions. An eighth sub-step that repeats the processing of the step;
The circuit design method according to claim 1, further comprising:   In the fifth sub-step, when an OTN node exists within a certain distance from the start point or end point of the demand and the start point or end point of the demand is accommodated in the OTN node, it is confirmed whether or not an allowable delay time is satisfied. 3. The circuit design method according to claim 2, further comprising aggregating a start point or an end point of the demand in the OTN node when the demand is satisfied.   The circuit design method according to claim 2, wherein the seventh substep and the eighth substep are divided so that the number of nodes in each new partition is as equal as possible. As another method of the third step,
From the demands registered in the list 2, the demand having the longest distance between the start node and the end node is extracted, the start node of the demand is included in the set A, the end node is included in the set B, and other demands are The network is divided so that the end point (start point node or end point node) closer to the demand start point node belongs to the set A and the other belongs to B, and the node that is the starting point is determined for each divided block, A sixth sub-step of setting a path between origin nodes, aggregating demand in the path, and confirming whether or not the allowable delay time is satisfied when the same wavelength is used, and aggregating in the path if satisfied When,
When all the demands cannot be aggregated by the sixth substep, the set A and B of the sixth substep are different from each other in the end point node of the start point node and the start point node of the end point node belonging to the same set. The network is divided into four so as to satisfy the conditions belonging to the set, a node as a starting point is determined in each divided section, a path is set between the starting nodes, demand is concentrated on the path, and the same wavelength is used. And confirming whether or not the allowable delay time is satisfied.
If all demands cannot be aggregated by the seventh sub-step or cannot be divided so as to satisfy the conditions, the division is re-divided into sections that satisfy the conditions of the sixth sub-step, and the number of divisions is 6 and 8 An eighth sub-step that repeats the processing of the seventh sub-step while increasing
The circuit design method according to claim 2, wherein:   6. The circuit design method according to claim 2, wherein the delay constraint amount is a value obtained by dividing a linear distance between a start point node and an end point node by an allowable delay time.   The circuit design method according to claim 2, wherein the shortest path is determined by a Dijkstra method with a delay time as a path cost.   The starting node is a node that is in the partition and has the minimum sum of the straight line distances from the demand start point node and the end point node registered in the list 2, or the sum of the number of hops is the minimum. 8. The circuit design method according to claim 2, wherein the circuit design method is provided.   In the seventh sub-step and the eighth sub-step, when the number of wavelengths after aggregation + the number of wavelengths that could not be aggregated <the number of wavelengths when all paths were aggregated, the paths that could not be aggregated 9. The circuit design method according to claim 2, wherein the wavelength is assigned independently. After the completion of the third step, when the total bandwidth after path aggregation is equal to or greater than the bandwidth of one interface,
Among all the demands scheduled to be aggregated, among the demands for which the delay constraint amount is not more than a certain value, there are passing nodes of the same existing wavelength within a certain distance from the starting point and the ending point of the demand. 9. When accommodated in a path, a demand satisfying an allowable delay time is accommodated in a wavelength of the passing node, so that a total band after path aggregation is less than a band of one interface. The circuit design method according to any one of the above.   The accommodation is performed in order from the demand having the largest delay constraint amount among the demands in which the delay constraint amount is equal to or less than a predetermined value until the total bandwidth after path aggregation becomes less than the bandwidth of one interface. The circuit design method according to claim 10. A computer for designing a line for demand with an acceptable delay time,
Means for determining a delay constraint amount from the distance between the start node and the end node and the allowable delay time for all demands, and setting a path for the demand in which the delay constraint amount exceeds a certain value;
A means for consolidating the demands in which the delay constraint amount is equal to or less than a predetermined value in a path set by the means and, when using the same wavelength, satisfying an allowable delay time;
Means for dividing the network into a plurality of partitions, determining a node as a starting point in each divided partition, setting a path between the starting nodes, and aggregating demand that could not be aggregated by the means in the path;
A program characterized by designing a circuit.

Images