Control device, control method and program

Abstract

The purpose of the present invention is to enable the load for changing allocations in a virtual network to be reduced and enable virtual servers of which optimization calculation times are reduced and electricity charges are cheap to be allocated.　Thus, this network control device 100, which allocates a virtual network for providing a network service onto a physical network, comprises: an acquisition unit 102 which acquires information pertaining to the physical network; a reallocation candidate narrowing-down unit 103 which performs, on the basis of a change amount of a unit electricity charge for each node that constructs the physical network, narrowing down to prescribed virtual servers, which are reallocation candidates, among virtual servers that are respectively allocated to nodes; and an allocation unit 106 which allocates, onto the physical network on the basis of the information pertaining to the physical network acquired by the acquisition unit 102, the virtual network including a specific server among the prescribed virtual servers that are narrowed down by the reallocation candidate range narrowing-down unit 103.

Description

[Technical field]

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

[0002] A technology called Network Functions Virtualization (NFV) is known that enables the provision of various network services by flexibly combining virtual network functions (VNFs).

[0003] These network services are provided by allocating a virtual network (VN) built by a service provider to physical resources owned by a telecommunications carrier. Specifically, the virtual network is allocated to each physical resource while satisfying the requirements for each virtual network (e.g., the start point and end point of the virtual network, the required bandwidth, the allowable delay time, etc.) and the resource constraints of the physical network (e.g., the link bandwidth and the server capacity, etc.). The virtual network is built by virtual servers and virtual links.

[0004] Meanwhile, from the viewpoint of business continuity, the reduction of OPEX (Operating Expenses), such as business expenses and operating costs, is an important issue for telecommunications carriers. The electricity bill for operating servers and the air conditioners that cool them accounts for a large portion of OPEX, and there is a strong demand for a reduction in electricity bills. In response to this, a method for controlling a virtual network that minimizes power consumption has been proposed to reduce OPEX (Non-Patent Document 1). According to this method, the power consumption of servers and links is modeled, and the virtual network can be controlled to minimize power consumption.

[0005] On the other hand, it is known that the power consumption of servers and devices in a data center varies depending on the load factor (Non-Patent Document 2).In addition, it is known that the power consumption of a data center also depends on the outside temperature because the power consumption of air conditioners is closely related to the outside temperature (Non-Patent Document 3).

[0006] However, for example, the method described in Non-Patent Document 1 does not take into account both the outside temperature and the load factor (how busy the server is), and the accuracy of estimating power consumption is low.

[0007] In addition, electricity charges are calculated by multiplying the power consumption by the unit price of electricity, but the unit price of electricity varies depending on the region (i.e., the area under the jurisdiction of the power company) and the time of day. For this reason, even if a virtual network is allocated so that the electricity charge is minimized in a certain time period, the electricity charge is not necessarily minimized in another time period.

[0008] In view of this, a method has been proposed for optimizing a virtual network so as to minimize the electricity fee when the unit price of electricity changes, using a virtual network optimization method (Patent Document 1). [Prior art documents] [Non-patent literature]

[0009] Patent Document 1: JP 2020-127108 Summary of the Invention [Problem to be solved by the invention]

[0010] However, the method disclosed in Patent Document 1 has a problem in that the amount of calculation becomes enormous when the number of virtual servers or the scale of the physical network topology increases.

[0011] The present invention has been made in consideration of the above points, and aims to reduce the load of changing the allocation of virtual networks, reduce optimization calculation time, and enable allocation of virtual servers with low electricity costs. [Means for solving the problem]

[0012] In order to achieve the above objectives, The present disclosure relates to A control device that allocates a virtual network for providing a network service on a physical network, comprising: In a route connecting a start point and an end point of the virtual network, a reallocation candidate narrowing down means for narrowing down the virtual servers allocated to each node to a specific virtual server that is a reallocation candidate based on an amount of change in cost for each node constituting the physical network; and an allocation means for allocating the virtual network including a specific server out of the specific virtual servers narrowed down by the reallocation candidate narrowing down means to the physical network based on information about the physical network. the reallocation candidate narrowing means comprises: a change amount calculation means for calculating a change amount in cost for each node constituting the physical network; a ranking means for ranking the nodes in descending order of the change amount calculated by the change amount calculation means; a reallocation priority setting means for setting a reallocation priority of each virtual server allocated to each of the nodes based on the ranks assigned by the ranking means; and a reallocation candidate selection means for selecting the predetermined virtual servers as reallocation candidates in descending order of the reallocation priority set by the reallocation priority setting means. The control device is characterized by having the following features. Effect of the Invention

[0013] As described above, according to the present invention, it is possible to reduce the load of changing the allocation of virtual networks, reduce the optimization calculation time, and enable the allocation of virtual servers with low electricity charges. [Brief description of the drawings]

[0014] [Figure 1] FIG. 2 is a diagram for explaining an example of a relationship between a physical network and an area under the jurisdiction of a power company. [Diagram 2] FIG. 11 is a diagram for explaining an example of a relationship between a time period and an electricity rate. [Diagram 3] FIG. 2 is a diagram illustrating an example of a functional configuration of a network control device according to an embodiment of the present invention. [Figure 4] 11 is a flowchart showing an example of a virtual network allocation process according to an embodiment of the present invention. [Diagram 5] 13 is a flowchart showing a process of narrowing down reallocation candidates. [Figure 6] FIG. 13 is a diagram showing electricity charges at each node at time t. [Figure 7] FIG. 13 is a diagram showing electricity charges for each node at time t+1. [Figure 8] FIG. 13 is a diagram showing the amount of change in electricity charges for each node. [Figure 9] FIG. 13 is a diagram illustrating the rank of the amount of change of each node. [Figure 10] FIG. 13 is a diagram showing the recalculation priority of each virtual server (VM). [Figure 11] 1 is a diagram illustrating an example of a hardware configuration of a network control device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] First Embodiment A first embodiment of the present invention will be described below in detail with reference to the drawings. In the embodiment of the present invention, electricity charges are assumed as an example of costs, and a case will be described in which virtual networks (virtual servers and virtual links) are allocated to physical resources in such a way that electricity charges are minimized while taking into consideration the outside temperature, load factor, time period, area (area under the jurisdiction of the power company), etc.

[0016] Therefore, in an embodiment of the present invention, power consumption models (a data center power consumption model and a link power consumption model) are defined that take into account the outside temperature (the higher the temperature, the more power is consumed for cooling) and the load factor (the busier the server is processing, the more power is consumed), and an objective function is formulated that multiplies the power consumption calculated by these power consumption models by the electricity rate taking into account the time period to calculate the total cost (i.e., the electricity fee). Then, the allocation of virtual networks is calculated so as to minimize this objective function. Note that allocating a virtual network to a physical resource may be expressed, for example, as "embedding a virtual network in a physical resource" or "placing a virtual network on a physical resource".

[0017] <Relationship between the physical network 200 and the jurisdiction area of ​​the electric power company> First, the relationship between physical network 200 and the area under the jurisdiction of a power company will be described with reference to Fig. 1. Fig. 1 is a diagram for explaining an example of the relationship between physical network 200 and the area under the jurisdiction of a power company.

[0018] As shown in Fig. 1, physical network 200 is constructed by network control device 100, nodes N1 to N24, and links connecting the nodes. The jurisdictional areas of electric power companies A to I are also shown in Fig. 1. Each of nodes N1 to N24 is, for example, a data center in which servers and various communication devices are installed. Hereinafter, nodes N1 to N24 of physical network 200 will also be described as a data center.

[0019] Here, for example, assuming that the electricity rate of all electric power companies is the same, when accessing node N8 where a user terminal is installed to node 16 where a service server that provides a network service is installed, the optimal route is node N8 → node N10 → node 13 → node N12 → node N16. However, for example, if the electricity rate of electric power company E that governs nodes N12 and N13 rises higher than that of other electric power companies, it is assumed that a route that bypasses the area governed by electric power company E, such as node N8 → node N10 → node N14 → node N18 → node N17 → node N16, will be the optimal route. Also, when a certain service server provides a network service, it is preferable to place a virtual server at node N12 or node N13, which is approximately in the center, in the case of FIG. 1, taking into consideration centrality (i.e., a position where the sum of distances to all service destinations is small), but it is assumed that there is a cost advantage in placing a virtual server at node N18 or node N17 in terms of rising electricity rates.

[0020] In this way, since the electricity unit price differs depending on the area, the network control device 100 changes the optimal virtual network allocation pattern taking the electricity price into consideration (i.e., the allocation pattern when allocating virtual networks to physical resources). Furthermore, as will be described later, since the electricity unit price differs depending on the time of day, the network control device 100 also changes the optimal virtual network allocation pattern depending on the time of day.

[0021] <Relationship between time of day and electricity rate> Here, the relationship between the time period and the electricity unit price will be described with reference to Fig. 2. Fig. 2 is a diagram for explaining an example of the relationship between the time period and the electricity unit price.

[0022] Fig. 2 shows an example of the electricity rate unit price of a certain electric power company. As shown in Fig. 2, the electricity rate unit price generally differs depending on the time of day. In the example shown in Fig. 2, the electricity rate unit price is α1 during the nighttime (0:00-8:00, 22:00-24:00), the electricity rate unit price is α2 during the daytime (8:00-13:00, 16:00-22:00), and the electricity rate unit price is α3 during the peak time period (13:00-16:00).

[0023] In this way, electricity rates generally vary depending on the time of day. However, the example shown in Figure 2 is just one example, and some electric power companies may have a uniform electricity rate regardless of the time of day, or may have more detailed time periods than the example shown in Figure 2.

[0024] <Formulation of objective function> Next, in the physical network 200 where the electricity unit price varies depending on the region (area under the jurisdiction of the electric power company) and the time period, the network control device 100 formulates an objective function for calculating the allocation of virtual networks that will result in the lowest electricity price.

[0025] First, the power consumption model P DCi and the power consumption model of the jth link P Linkj is defined by functions f and g as shown in the following equations (1) and (2).

[0026]

number

[0027]

number

[0028] It is considered that the relationship between the outside temperature and power consumption is an exponential function. In addition, even if the load factor is zero, power consumption will not be zero because idle power, standby power, and power for air conditioning equipment to keep the servers cool are necessary. Taking these factors into consideration, the power consumption model P DCi As an example, it can be expressed as the following formula (3).

[0029]

number

[0030] Virtual link power consumption model P Linkj Similarly, as an example, it can be expressed as the following formula (4).

[0031]

number

[0032] The constant D in the above equations (3) and (4) i and D. j It may be derived, for example, by regression analysis from actual measurement data of power consumption, load factor, and outside air temperature.

[0033] Next, the CPU (Central Processing Unit) demand required by the a-th virtual network is calculated as C VNa , the CPU processing power of the i-th data center is C imax, the jth virtual link bandwidth is C jmax In this case, l i and b j can be expressed by the following equations (5) and (6).

[0034]

number

[0035]

number

[0036] In addition, since it is not possible to allocate a virtual network that exceeds the processing capacity of the CPU of each data center and the link bandwidth of each virtual link, the constraints shown in the following formulas (7) and (8) are obtained.

[0037]

number

[0038]

number

[0039]

number

[0040] Therefore, by calculating a virtual network allocation pattern (i.e., a route connecting the start point and end point of each virtual network) that minimizes the objective function h shown in the above formula (9) while satisfying the constraint conditions shown in the above formulas (7) and (8), it is possible to realize allocation of virtual networks with minimized costs (i.e., electricity charges). Note that, for example, in the case of service function chaining in which a firewall (FW) or an intrusion detection system (IDS) is placed along the route, in addition to the route, which data center should have the functions of the FW, IDS, etc. is also calculated.

[0041] In addition, the constraints shown in the above equations (7) and (8) are the CPU processing capacity of each data center and the link bandwidth of each link. In addition, the following constraints may also be considered depending on the required quality of the network service provided by the virtual network.

[0042] - The bandwidth of each link must not be exceeded - The delay time does not exceed the set upper limit. Each virtual network is assigned to one physical machine (i.e., virtual networks are indivisible). -Do not exceed the contracted power of each data center Here, the objective function h shown in the above formula (9) can be applied to the case where the load factor and the outside air temperature are in a steady state (i.e., when ti, li, and bj are constants), but can also be applied to the case where ti, li, and bj change with time (i.e., when ti, li, and bj are variables). By making ti, li, and bj variables, it is possible to calculate an optimal virtual network allocation pattern for each hour, taking into consideration all of the time changes in the outside air temperature, the load factor (i.e., the traffic volume), and the electricity rate. For example, the outside air temperature ti can be data obtained from a weather information service, and the load factors li and bj can be calculated from the traffic volume predicted by a known traffic prediction technique. The traffic prediction technique is, for example, a technique for predicting future traffic volume from past traffic volume by a technique such as machine learning. Here, the prediction technique can be any time series prediction technique, such as an ARIMA model or LSTM (Long Short-Term Memory). For the ARIMA model, for example, a method described in "Feng, Huifang, and Yantai Shu. "Study on network traffic prediction techniques.", Wireless Communications, Networking and Mobile Computing, 2005. Proceedings. 2005 International Conference on. Vol. 2. IEEE, 2005." can be used. For the LSTM, for example, a method described in "Hochreiter, Sepp, and Jurgen Schmidhuber. "Long short-term memory.", Neural computation 9.8 (1997): 1735-1780." can be used.

[0043] Hereinafter, a description will be given of a network control device 100 that realizes optimal virtual network allocation by calculating a virtual network allocation pattern that minimizes the objective function h shown in the above formula (9) while satisfying the constraint conditions shown in the above formulas (7) and (8).

[0044] <Functional configuration of the network control device 100> Next, a functional configuration of the network control device 100 in the embodiment of the present invention will be described with reference to Fig. 3. Fig. 3 is a diagram showing an example of the functional configuration of the network control device 100 in the embodiment of the present invention.

[0045] 3, the network control device 100 in the embodiment of the present invention has a schedule management unit 101, an acquisition unit 102, a reallocation candidate narrowing down unit 103, an allocation unit 106, and a setting command unit 107 as functional units realized by commands from a processor 304 shown in Fig. 11 described later. The network control device 100 in the embodiment of the present invention also has a model storage unit 108, an electricity rate unit price storage unit 109, and an acquired information storage unit 110 as storage units constructed by at least one of a RAM 302 and an auxiliary storage device 306 described later and shown in Fig. 11 described later.

[0046] The model storage unit 108 stores a power consumption model (i.e., the power consumption model shown in the above formulas (1) and (2)). The electricity rate unit price storage unit 109 stores the electricity rate unit price for each time period in each region. The electricity rate unit price is updated each time the electricity rate unit price is changed, for example, due to a revision of the power company's rate unit price, a change in the contract with the power company, a change to another power company, etc. The acquired information storage unit 110 stores various information acquired by the acquisition unit 102.

[0047] The schedule management unit 101 calculates an optimal virtual network allocation pattern, and manages a schedule (i.e., execution conditions for virtual network allocation) for executing a process (virtual network allocation process) for allocating the virtual network of this allocation pattern to physical resources. Here, examples of execution conditions for virtual network allocation include "a predetermined time interval has elapsed since the previous execution" and "traffic volume has increased by a certain threshold or more."

[0048] Next, the acquiring unit 102 acquires information related to the physical network and information related to the allocation request of the virtual network.

[0049] The information relating to the physical network includes at least one of information indicating the outside temperature of each node included in the physical network, information indicating the predicted traffic volume of the physical network, and physical resource information of the physical network.

[0050] The information indicating the outside temperature of each node included in the physical network is information indicating the outside temperature at the location of each data center (hereinafter, also referred to as "outside temperature information"). Here, the acquisition unit 102 may acquire the outside temperature information, for example, from a weather information service or the like, or may acquire the information from an outside temperature sensor if an outside temperature sensor or the like is installed in each data center.

[0051] Moreover, the information indicating the predicted traffic volume of the physical network is information indicating the traffic volume at the time when the virtual network is allocated (i.e., the predicted traffic volume). As described above, such a traffic volume is predicted by the traffic prediction technique. Note that the acquiring unit 102 may acquire the traffic volume predicted by another device connected to the network control device 100 via the network, or may acquire the traffic volume predicted by the network control device 100.

[0052] Furthermore, the physical resource information of the physical network is information indicating, for example, the processing capacity of the CPU of each data center, the link bandwidth of each link, and the like.

[0053] On the other hand, the information on the allocation request of the virtual network is information indicating the allocation request of the virtual network from the service provider that provides the network service. The allocation request includes the required conditions of the virtual network (for example, the start point and end point of the virtual network, the required bandwidth, the allowable delay, etc.).

[0054] Next, the reallocation candidate narrowing unit 103 narrows down the virtual servers allocated to each node to predetermined virtual servers that are candidates for reallocation, based on the amount of change in the electricity unit price for each node that constitutes the physical network.

[0055] Here, a detailed description will be given of the reallocation candidate narrowing unit 103. The reallocation candidate narrowing unit 103 includes a change amount calculation unit 131, a ranking unit 132, a reallocation priority setting unit 133, and a reallocation candidate selection unit .

[0056] Of these, the change amount calculation unit 131 calculates the amount of change in the electricity unit price for each node constituting the physical network, based on information indicating the electricity rate for each time period in each region stored in the electricity unit price storage unit 109. The ranking unit 132 ranks the nodes in descending order of the amount of change calculated by the change amount calculation unit 131. The reallocation priority setting unit 133 sets reallocation priorities for each virtual server allocated to each node, based on the ranks assigned by the ranking unit 132. The reallocation candidate selection unit 134 selects predetermined virtual servers as reallocation candidates in descending order of the reallocation priorities set by the reallocation priority setting unit 133. The predetermined virtual servers are a predetermined number or a predetermined ratio of virtual servers that are reallocation candidates among the virtual servers.

[0057] Next, the allocating unit 106 allocates, to the physical network, a virtual network including a specific server out of the predetermined virtual servers narrowed down by the reallocation candidate narrowing down unit 103, based on the information about the physical network acquired by the acquiring unit 102. More specifically, based on the information about the physical network acquired by the allocating unit 106 and the acquiring unit 102, the virtual network including a specific server out of the predetermined virtual servers selected by the reallocation candidate selecting unit 134, to the physical network.

[0058] Furthermore, in response to an instruction from the schedule management unit 101, the allocation unit 106 calculates an optimal allocation pattern of a virtual network including a specific server onto a physical network using the objective function h shown in the above formula (9). That is, the allocation unit 106 first creates constraint conditions using information on a service request (here, a virtual network allocation request) and physical resources. Then, the allocation unit 106 calculates an optimal virtual network allocation pattern using a power consumption model, electricity rates for each region and time period, outside temperature, and a load factor. The load factor is calculated, for example, from a predicted traffic volume and a CPU processing capacity (or link bandwidth).

[0059] The configuration command unit 107 transmits a virtual network configuration command to each physical resource (e.g., a server, various communication devices, etc.) so as to realize the optimal virtual network allocation pattern calculated by the allocation unit 106. This realizes optimal virtual network allocation.

[0060] <Processing or Operation of the Present Embodiment> Next, the allocation process of a virtual network (virtual servers and virtual links) in the embodiment of the present invention will be described with reference to Fig. 4 to Fig. 10. Fig. 4 is a flowchart showing an example of the allocation process of a virtual network in the embodiment of the present invention.

[0061] Step S101: First, the schedule management unit 101 judges whether or not the execution condition of the virtual network allocation is satisfied. As described above, examples of the execution condition of the virtual network allocation include "a predetermined time interval has passed since the previous execution" and "traffic volume has increased by a certain threshold or more." If it is judged that the execution condition of the virtual network allocation is satisfied, the process proceeds to step S102.

[0062] Here, for example, if the condition for allocating a virtual network is set to "when traffic volume increases beyond a certain threshold," a daily operational plan (i.e., a plan for what kind of virtual network to allocate) can be formulated based on the traffic volume predicted by traffic prediction technology, and then a virtual network can be allocated when an unexpected increase in traffic occurs that exceeds a certain predetermined threshold.

[0063] For example, if the execution condition for allocating a virtual network is "a specified time interval has elapsed since the last execution," then since electricity charges are generally calculated every 30 minutes, it may be possible to set the specified time interval (this time interval is also referred to as the "control interval") to 30 minutes.

[0064] On the other hand, if the control interval is set to 30 minutes or more, the electricity unit price may change between the control intervals. In such a case, the time average of the electricity unit price may be used in the objective function h shown in the above formula (9). For example, if the electricity unit price changes between a certain control interval T1 and T2, the time average unit price F avr k is defined as the following equation (10), and the F of the objective function h shown in the above equation (9) is Ti and F Tj Instead of F avr i and F avr j can be used.

[0065]

number

[0066] Step S102: The acquisition unit 102 acquires outside air temperature information indicating the outside air temperature at the location of each data center, and stores the information in the acquired information storage unit 110.

[0067] Step S103: The acquiring unit 102 acquires the predicted traffic volume at the time when the virtual network is to be allocated, and stores the acquired traffic volume in the acquired information storage unit 110.

[0068] Step S 104 : The acquiring unit 102 acquires information related to physical resources, and stores the information in the acquired information storage unit 10 .

[0069] Step S105: The acquiring unit 102 acquires information indicating a service request (a request for allocation of a virtual network), and stores the information in the acquired information storage unit 10.

[0070] It should be noted that steps S102 to S105 may be performed in any order.

[0071] Step S106: The reallocation candidate narrowing unit 103 narrows down the virtual servers allocated to each node to predetermined virtual servers that are candidates for reallocation, based on the amount of change in the electricity unit price for each node that constitutes the physical network.

[0072] (Reassignment candidate narrowing down process) Here, the process of narrowing down reallocation candidates will be described in detail with reference to Fig. 5. Fig. 5 is a flowchart showing the process of narrowing down reallocation candidates. Fig. 6 is a diagram showing the electricity charges of each node at time t. Fig. 7 is a diagram showing the electricity charges of each node at time t+1. Fig. 8 is a diagram showing the amount of change in the electricity charges of each node. Fig. 9 is a diagram showing the rank of the amount of change of each node. Fig. 10 is a diagram showing the reallocation priority of each virtual server (VM). Note that Figs. 6 to 8 show four nodes N6, N9, N11, and N12 as an example.

[0073] Step S201: First, the change amount calculation unit 131 calculates the change amount of the electricity unit price for each node. Specifically, the change amount calculation unit 131 reads information indicating the electricity price of each node at time t from the electricity unit price storage unit 109. As shown in Fig. 6, at time t, the electricity unit price of node N6 is "15", the electricity unit price of node N9 is "18", the electricity unit price of node N11 is "11", and the electricity unit price of node N12 is "12".

[0074] Then, the change amount calculation unit 131 reads information indicating the electricity rate of each node at time t+1 from the electricity rate unit price storage unit 109. As shown in Fig. 7, at time t+1, the electricity rate unit price of node N6 is "18", the electricity rate unit price of node N9 is "18", the electricity rate unit price of node N11 is "10", and the electricity rate unit price of node N12 is "20".

[0075] Furthermore, the change amount calculation unit 131 calculates the amount of change in the electricity unit price at time t+1 from the electricity unit price at time t. As shown in Fig. 8, the amount of change in the electricity unit price at node N6 is "+3", the amount of change in the electricity unit price at node N9 is "0", the amount of change in the electricity unit price at node N11 is "-1", and the amount of change in the electricity unit price at node N12 is "+4".

[0076] Step S202: Next, the ranking unit 132 ranks the nodes in descending order of the amount of change. Specifically, as shown in Fig. 9, the rank of node N6 is "2", the rank of node N9 is "3", the rank of node N11 is "4", and the rank of node N12 is "1".

[0077] Step S203: The reallocation priority setting unit 133 sets a reallocation priority for each virtual server assigned to each ranked node. At this time, the reallocation priority setting unit 133 sets a reallocation priority for each virtual server in descending order of the rank of the assigned node. Specifically, as shown in Fig. 9 and Fig. 10, the recalculation priority of virtual server VM1 is "3", the recalculation priority of virtual server VM2 is "4", and the recalculation priority of virtual servers VM3 and VM4 is "1". Note that, as shown in Fig. 9, virtual server VM1 is assigned to node N6, and therefore virtual servers have already been assigned to nodes, so the reallocation priority setting unit 133 sets the reallocation priority of the virtual server, not the reallocation priority of the node.

[0078] Step S204: The reallocation candidate selection unit 134 selects predetermined virtual servers as reallocation candidates in descending order of reallocation priority.

[0079] The specified number of virtual servers as candidates for reallocation may be selected as a predetermined number, for example, by running two virtual servers in advance, or may be selected as a specified percentage of the total, such as the top 25% of all virtual servers in terms of recalculation priority.

[0080] This completes the process of narrowing down the reallocation candidates. In this way, before step S107 described later is executed, the reallocation candidate narrowing down unit 103 narrows down the reallocation candidates, so that it is possible to reduce the load of changing the settings of the virtual network, reduce the optimization calculation time, and enable allocation of virtual servers with low electricity charges.

[0081] Step S107: Next, the allocation unit 106 creates constraint conditions using the virtual network allocation request and information on physical resources. Then, the allocation unit 106 calculates an optimal virtual network allocation pattern using a power consumption model, the electricity rate for each time period in each area, the outside temperature, and the load factor.

[0082] Step S108: Next, the configuration command unit 107 transmits information indicating a virtual network configuration command to each physical resource so as to realize the optimal virtual network allocation pattern calculated in the above step S107.

[0083] The above-mentioned steps S101 to S105, S107, and S108 may be executed by other methods.

[0084] <Main Effects of This Embodiment> As described above, according to the embodiment of the present invention, the concept of a virtual server is used, which is highly effective in recalculating and reconfiguring from the viewpoint of minimizing electricity charges. When the electricity unit price of each link and node changes, the network control device 100 does not recalculate and reconfigure the entire virtual network, but refers to the amount of change in each electricity unit price and narrows down to a specific virtual server (group) for which load allocation is to be recalculated, and applies optimization calculation only to this specific virtual server (group). This makes it possible to reduce the load of reassignment of the virtual network, and also reduces the optimization calculation time, resulting in the effect of enabling virtual server load allocation at a low electricity charge.

[0085] In addition, the present invention has the effect of solving the problem that reconfiguring the entire virtual network every time the electricity rate changes may lead to instability in the network.

[0086] Furthermore, it is possible to realize allocation of virtual networks that minimizes electricity charges by taking into consideration the outside temperature, load factor, time period, region, etc. Therefore, by using the network control device 100 according to the embodiment of the present invention, for example, telecommunications carriers can realize further reductions in OPEX.

[0087] In the embodiment of the present invention, electricity charges are assumed as an example of costs, but this is not limited to this, and the present invention can be similarly applied to any objective function derived by multiplying a variable that can change for each region at each time, like the electricity rate, by power consumption.

[0088] For example, another example of cost is CO2 emissions. CO2 emissions are calculated by multiplying power consumption by a CO2 emission coefficient, but the CO2 emission coefficient, like the unit price of electricity, has a different value for each electric power company and there are regional differences. In most cases, the CO2 emission coefficient does not change over time and is a fixed value, but in the future, it is considered that CO2 emission coefficients will be defined for each time period due to the increasing need to reduce environmental loads and the fact that power generation facilities operating differ depending on the time of day. In that case, it is considered that the application of the present invention will make it possible to minimize CO2 emissions and provide services that appeal to reducing environmental loads.

[0089] <Hardware Configuration of the Network Control Device 100> Finally, the hardware configuration of the network control device 100 in the embodiment of the present invention will be described with reference to Fig. 11. Fig. 11 is a diagram showing an example of the hardware configuration of the network control device 100 in the embodiment of the present invention.

[0090] 11, the network control device 100 according to the embodiment of the present invention has, as hardware, an external I / F 301, a RAM (Random Access Memory) 302, a ROM (Read Only Memory) 303, a processor 304, a communication I / F 305, and an auxiliary storage device 306. Each of these pieces of hardware is connected to each other via a bus B so as to be able to communicate with each other.

[0091] The external I / F 301 is an interface with an external device. The external device may be a recording medium 301a. The network control device 100 can read and write data from and to the recording medium 301a via the external I / F 301.

[0092] The recording medium 301a may be, for example, a flexible disk, a CD (Compact Disc), a DVD (Digital Versatile Disk), an SD memory card (Secure Digital memory card), or a USB (Universal Serial Bus) memory card.

[0093] The RAM 302 is a volatile semiconductor memory that temporarily stores programs and data. The ROM 303 is a non-volatile semiconductor memory that can store programs and data even when the power is turned off. The ROM 303 stores, for example, setting information related to the OS (Operating System) and setting information related to the communication network.

[0094] The processor 304 is, for example, a CPU or a GPU (Graphics Processing Unit), and is an arithmetic device that reads out programs and data from the ROM 303, the auxiliary storage device 306, etc. onto the RAM 302 and executes processing. Each functional unit (for example, the schedule management unit 101, the acquisition unit 102, the reallocation candidate narrowing down unit 103, the allocation unit 106, and the setting command unit 107) of the network control device 100 in the embodiment of the present invention is realized by, for example, processing in which one or more programs stored in the auxiliary storage device 306 or the like are executed by the processor 304. Note that these one or more programs may be stored in, for example, the recording medium 301a.

[0095] The communication I / F 305 is an interface for connecting the network control device 100 to the physical network 200 .

[0096] The auxiliary storage device 306 is, for example, a hard disk drive (HDD) or a solid state drive (SSD), and stores various programs and data. The programs and data stored in the auxiliary storage device 306 include, for example, an OS, application software that realizes various functions on the OS, and one or more programs that realize each functional unit of the network control device 100 in the embodiment of the present invention.

[0097] The network control device 100 according to the embodiment of the present invention has the hardware configuration shown in Fig. 11, and thereby can realize the above-mentioned various processes. In the example shown in Fig. 11, the network control device 100 according to the embodiment of the present invention is realized by one device (computer), but the present invention is not limited to this, and may be realized by multiple devices (computers). One device (computer) may include multiple processors 304 and multiple memories (RAM 302, ROM 303, auxiliary storage device 306, etc.). In addition to the above hardware, the network control device 100 may have, for example, a display device such as a display and an input device such as a keyboard or a mouse.

[0098] Second Embodiment Next, a second embodiment of the present invention will be described. Note that this embodiment is different from the first embodiment only in the processing of the ranking unit 132, and the rest of the configuration is the same, so only the different processing will be described.

[0099] In the above embodiment, the network control device 100 ranks the nodes using the amount of change in the electricity unit price (see S201, S202), but this is not limited to this. For example, the ranking unit 132 may rank the nodes taking into consideration information such as the centrality of the topology of the physical network or the availability of resources of the physical server of each node, in addition to the amount of change in the electricity unit price. In this case, the acquired information storage unit 110 stores information indicating the topology of the physical network in advance, and the ranking unit 132 refers to the information indicating this topology. In addition, the acquiring unit 102 periodically acquires information such as the availability of resources of the virtual server from each node.

[0100] In this way, when using the amount of change and centrality, the following indexes can be used for ranking.

[0101] (index)=(change) / (centrality) Since "low centrality" is considered to mean "being placed at the edge of the topology," it is expected that the distance between the user terminal and the virtual server will be greater. By using the above index, it is possible to move virtual servers on nodes with high electricity unit prices to nodes with lower prices, while also allocating virtual servers that tend to be farther away from the user terminal to the center of the topology, which is expected to reduce communication delays.

[0102] 〔supplement〕 The present invention is not limited to the above-described embodiments specifically disclosed, and various modifications and variations are possible as described below without departing from the scope of the claims.

[0103] That is, in the above embodiment, the electricity rate is taken as an example of the additional information that each node and each link has, but the present invention is not limited to this. For example, it is also possible to apply the present invention to a situation where each node is equipped with a solar power generation system, and by selecting a virtual server to be rearranged while observing the change in the amount of power generated, a virtual network that effectively utilizes renewable energy can be realized. [Explanation of symbols]

[0104] 100 Network control device (an example of a control device) 101 Schedule Management Department 102 Acquisition unit (an example of an acquisition means) 103 Reallocation candidate narrowing down unit (an example of reallocation candidate narrowing down means) 106 Allocation unit (an example of allocation means) 107 Setting command part 108 Model Memory Section 109 Electricity rate unit price memory section 110 Acquired information storage unit 131 Change amount calculation unit (an example of a change amount calculation means) 132 Ranking unit (an example of a ranking means) 133 Reallocation priority setting unit (an example of reallocation priority setting means) 134 Reallocation candidate selection unit (an example of reallocation candidate selection means) 200 Physical Network 301 External I / F 301a Recording medium 302 RAM 303 ROM 304 Processor 305 Communication I / F 306 Auxiliary storage device

Claims

1. A control device that allocates a virtual network for providing a network service on a physical network, comprising: a reallocation candidate narrowing down means for narrowing down the virtual servers allocated to each node in a path connecting a start point and an end point of the virtual network to a predetermined virtual server that is a reallocation candidate based on a change in cost for each node that constitutes the physical network; an allocation means for allocating, to the physical network, the virtual network including a specific server among the predetermined virtual servers narrowed down by the reallocation candidate narrowing means, based on information about the physical network; having The reallocation candidate narrowing down means a change amount calculation means for calculating a change amount of a cost for each node constituting the physical network; a ranking means for ranking the nodes in descending order of the amount of change calculated by the amount of change calculation means; a reallocation priority setting means for setting a reallocation priority of each virtual server allocated to each of the nodes based on the ranks assigned by the ranking means; a reallocation candidate selection means for selecting the predetermined virtual servers as reallocation candidates in the order of the reallocation priorities set by the reallocation priority setting means; A control device comprising:

2. 2. The control device according to claim 1, wherein the reallocation candidate narrowing down unit narrows down the virtual servers to a predetermined number or a predetermined percentage of the virtual servers that are candidates for reallocation.

3. 2. The control device according to claim 1, wherein the ranking means ranks the nodes using the amount of change calculated by the amount of change calculation means and taking into consideration centrality in the topology of the physical network.

4. 2. The control device according to claim 1, wherein the ranking means ranks the nodes using the amount of change calculated by the amount of change calculation means and taking into consideration an available resource situation of each of the nodes.

5. The control device according to claim 1, characterized in that the information relating to the physical network includes at least one of information indicating the outside temperature of each node included in the physical network, physical resource information of the physical network, and information indicating a predicted traffic volume of the physical network.

6. A control method executed by a control device that allocates a virtual network for providing a network service on a physical network, comprising: The control device, a reallocation candidate narrowing down step of narrowing down the virtual servers allocated to each node in a path connecting a start point and an end point of the virtual network to a predetermined virtual server that is a reallocation candidate based on a change in cost for each node that constitutes the physical network; an allocating step of allocating, to the physical network, the virtual network including a specific server among the predetermined virtual servers narrowed down by the reallocation candidate narrowing step, based on information about the physical network; Run The reallocation candidate narrowing down step includes: a change amount calculation step of calculating a change amount of cost for each node constituting the physical network; a ranking step of ranking the nodes in descending order of the amount of change calculated by the amount of change calculation step; a reallocation priority setting step of setting a reallocation priority of each virtual server allocated to each of the nodes based on the ranks assigned by the ranking step; a reallocation candidate selection step of selecting the predetermined virtual servers as reallocation candidates in the order of the reallocation priorities set by the reallocation priority setting step; A control method comprising:

7. A program causing a computer to execute the method according to claim 6.

Images