JP5364183B2 - Network resource management device - Google Patents

Description

  The present invention relates to a packet switching network such as IP (Internet Protocol), MPLS (Multiprotocol Label Switching), GMPLS (Generalized Multi-protocol label switching), a path using WDM (Wavelength Division Multiplexing), TDM (Time Division Multiplexing), and the like. The present invention relates to a resource management apparatus that manages physical resources of a network of a switching network.

  In recent years, with the increase of P2P (Peer to Peer) traffic or the rise of Consumer Generated Media (CGM) represented by YouTube (registered trademark) and the like, traffic is concentrated on specific popular sites. With the advent of specific popular sites like this, rapid changes in traffic patterns have occurred frequently. In particular, since the advent of YouTube (registered trademark), it has been reported that the amount of traffic on the international line from the US to Japan is more conspicuous than others. In the future, traffic patterns are expected to change significantly due to similar killer applications and killer services. Due to this rapid change in the traffic pattern, it is more difficult to construct an efficient network.

  Non-Patent Document 1 reports that the performance deteriorates due to the interaction between overlay networks, particularly in an infrastructure network that accommodates a plurality of overlay networks. In preparation for an era when various new services appear and traffic fluctuations that cannot be anticipated at the present time, networks need to flexibly use physical resources to adapt flexibly to environmental fluctuations. .

  As a technique for dealing with this problem, a network virtualization technique has attracted attention. The network virtualization technology is a technology for configuring a plurality of virtual networks by virtually dividing a physical resource of the network and allocating a part of the physical resource.

  PlanetLab described in Non-Patent Document 2 is an application layer virtualization technology developed in the United States, and is an overlay network technology deployed on an IP network. In PlanetLab described in Non-Patent Document 2, a participant brings a Linux (registered trademark) server as a physical resource, divides it into a plurality of virtual machines called Sliver, and configures a service network by combining these virtual machines.

  In PlanetLab described in Non-Patent Document 2, a virtual resource is a Linux server, and a virtual network is realized by software. Therefore, there is a limit to the communication speed that PlanetLab can provide. Furthermore, PlanetLab, which is a virtual network on the packet network, may cause interference because different virtual networks share physical resources related to the same packet transfer.

  Thus, in the existing network virtualization technology, a plurality of virtual networks are multiplexed on the IP packet network. For this reason, there is a possibility that a large change in the traffic pattern appears due to the interaction between the virtual networks, and the stability of the network is lowered.

  Non-Patent Document 3 describes an infrastructure network resource management technology that allows a configuration change of each virtual network while suppressing an interaction between the virtual networks in an infrastructure network that accommodates a plurality of virtual networks. The technique described in Non-Patent Document 3 avoids interference between virtual networks by allocating reserved resources to specific virtual networks.

W. Jiang et al., “On the interaction of multiple overlay routing”, Performance Evaluation 62, 10 August 2005, p.229-246 Akihiro Nakao, “PlanetLab: Towards Ubiquitous Networking, Possibility of Distributed Computing in Ubiquitous Society”, IEICE Technical Report, IEICE, 2005-10-13, NS Network System 105 (357), pp25-28 Takashi Miyamura et al., “Dynamic resource allocation mechanism for managed self-organization”, Network Operations and Management Symposium (APNOMS), 21-23 Sept 2011, 13th Asia-Pacific

  The technique described in Non-Patent Document 3 obtains AC traffic information of all virtual networks accommodated in the same physical network, calculates a correlation, and extracts a group of virtual networks to be controlled. As a result, the technique described in Non-Patent Document 3 may increase the calculation load when the number of virtual networks increases. Furthermore, the technique described in Non-Patent Document 3 may not be able to efficiently allocate physical resources when the number of virtual networks in which physical resources are in a competitive relationship increases.

  Accordingly, an object of the present invention is to provide a network resource management apparatus that can easily minimize the mutual interference between virtual networks even when the number of virtual networks increases.

In order to solve the above-described problem, in the invention described in claim 1, based on infrastructure network topology information relating to an infrastructure network and virtual network topology information relating to each virtual network constructed on the infrastructure network, An information collection unit that collects traffic demand information related to each virtual network, and a virtual network to be controlled is selected from each virtual network based on the traffic demand information, and each physical network related to the infrastructure network is selected based on the virtual network. Based on the traffic demand information collected by the information collection unit and a virtual network selection unit that selects a physical resource to be controlled from resources, the physical resource to be provided to the selected virtual network is calculated by a piecewise linear function A network resource management device characterized by comprising a reservation resource calculation unit.

In this way, even if the number of virtual networks increases, mutual interference between the virtual networks can be easily minimized. Further, even if the traffic demand is zero, the physical resource can be provided in advance, so that it can be prepared when the physical resource is newly used.
According to the second aspect of the present invention, based on the infrastructure network topology information relating to the infrastructure network and the virtual network topology information relating to each virtual network constructed on the infrastructure network, the traffic demand information relating to each virtual network. And a control target virtual network selected from each virtual network based on the traffic demand information, and based on the virtual network, a control target physical resource is selected from each physical resource related to the infrastructure network. Based on the traffic demand information collected by the virtual network selection unit to be selected and the information collection unit, the physical resources to be provided to the selected virtual network are increased in a synergistic manner with respect to the traffic demand information. A network resource management apparatus characterized by comprising a reservation resource calculation unit for calculation.
In this way, even if the number of virtual networks increases, mutual interference between the virtual networks can be easily minimized. Furthermore, many physical resources can be reserved in an important virtual network that consumes many physical resources.

In the invention according to claim 3 , based on the infrastructure network topology information related to the infrastructure network and the virtual network topology information related to each virtual network constructed on the infrastructure network, each physical resource related to the infrastructure network An information collection unit that collects traffic demand information related to each virtual network, and a physical resource to be controlled is selected from each physical resource based on the traffic demand information, and a control target is selected from each virtual network based on the physical resource A virtual network selection unit that selects the virtual network, and a reservation resource calculation unit that calculates the physical resource to be provided to the selected virtual network based on the traffic demand information collected by the information collection unit. The network resource management apparatus is characterized by this.

In this way, even if the number of virtual networks increases, mutual interference between the virtual networks can be easily minimized. Furthermore, when the number of virtual networks increases, the amount of calculation can be reduced as compared with the invention according to claim 1 or claim 2 .

The invention according to claim 6 , further comprising a resource distribution unit that reserves and distributes the physical resource calculated by the reservation resource calculation unit to the virtual network . The network resource management device described in any one of the items is used.

  In this way, the calculated physical resource can be reserved and allocated to the virtual network, and the virtual network can be dynamically reconfigured.

In the invention according to claim 7 , the reservation resource calculation unit calculates an allocation amount of the reservation resource based on the traffic demand information related to each virtual network, and the physical resource includes a shared attribute or reservation. to set one of the types of attributes, and a resource management device of the network according to any one of claims 1 to 6, characterized in that.

  In this way, by setting a reservation type attribute, it is possible to guarantee the bandwidth of the virtual network and avoid mutual interference between the virtual networks. Furthermore, by setting a shared attribute to the physical resource, the physical resource can be interchanged between virtual networks, the resistance to traffic fluctuation can be improved, and the resource utilization efficiency of the infrastructure network can be improved.

The invention according to claim 4 is characterized in that the reservation resource calculation unit calculates the physical resource to be provided to the selected virtual network from the traffic demand information by a piecewise linear function. The network resource management device described in 3 is used.

  In this way, even if the traffic demand is zero, the physical resource can be provided in advance, so that it can be prepared when the physical resource is newly used.

In the invention according to claim 5 , the reservation resource calculation unit calculates the traffic demand information so that the physical resources provided to the selected virtual network increase synergistically. The network resource management apparatus according to claim 3 .

  In this way, many physical resources can be reserved in an important virtual network that consumes many physical resources.

  According to the present invention, it is possible to provide a network resource management apparatus that can easily minimize mutual interference between virtual networks even when the number of virtual networks increases.

It is a schematic block diagram which shows the infrastructure network and virtual network in 1st Embodiment. It is a figure which shows the detail of the infrastructure network and virtual network in 1st Embodiment. It is a figure which shows the attribute of the physical resource in 1st Embodiment. It is a schematic block diagram which shows the resource management apparatus in 1st Embodiment. It is a figure which shows the example of the allocation amount calculation function in 1st Embodiment. It is a flowchart which shows the resource allocation amount calculation process in 1st Embodiment. It is a flowchart which shows the resource allocation amount calculation process in 2nd Embodiment.

Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings.
(Configuration of the first embodiment)

FIG. 1 is a schematic configuration diagram showing an infrastructure network and a virtual network in the first embodiment.
The infrastructure network 10 includes nodes 20a to 20d and links 21-1 to 21-4. The node 20a and the node 20b are connected by a link 21-1. The node 20b and the node 20d are connected by a link 21-2. The node 20d and the node 20c are connected by a link 21-3. The node 20c and the node 20a are connected by a link 21-4.
The infrastructure network 10 is, for example, a WDM network in which the nodes 20a to 20d are connected by optical fibers, and supports IP over WDM (Internet Protocol over wavelength division multiplexing).

  The nodes 20a to 20d are optical connect devices such as OXC (optical cross-connect). The nodes 20a to 20d are devices that can be switched to freely set an optical communication path for data transmission in an optical fiber. Hereinafter, when the nodes 20a to 20d are not particularly distinguished, they are simply referred to as nodes 20.

  The links 21-1 to 21-4 are optical fibers, for example, and connect the nodes 20 to enable optical communication. Hereinafter, when the links 21-1 to 21-4 are not particularly distinguished, they are simply referred to as links 21.

  On the infrastructure network 10, network virtualization is performed by combining layer 1 technology such as WDM and TDM, layer 2 technology such as Ethernet (registered trademark), and layer 3 technology such as packet switching. A virtual network “# 1” 11-1 and a virtual network “# 2” 11-2 are configured after the resources are virtually divided.

  The virtual network “# 1” 11-1 is a network that is virtually constructed on the infrastructure network 10. The virtual network “# 1” 11-1 is, for example, a virtual network defined by IEEE 802.1Q. The virtual network “# 1” 11-1 includes nodes 20 a to 20 d as in the infrastructure network 10, and includes a path 22 different from the infrastructure network 10.

  The virtual network “# 2” 11-2 is a network that is virtually constructed on the infrastructure network 10 like the virtual network “# 1” 11-1. The virtual network “# 2” 11-2 includes nodes 20 a to 20 d as in the infrastructure network 10, and includes a path 22 that is different from the infrastructure network 10.

  The path 22 of the virtual network “# 1” 11-1 is indicated by solid line arrows on the links 21-1 to 21-4 of the infrastructure network 10. The path 22 of the virtual network “# 2” 11-2 is indicated by broken-line arrows on the links 21-1 to 21-4 of the infrastructure network 10. Hereinafter, when the virtual network “# 1” 11-1 and the virtual network “# 2” 11-2 are not particularly distinguished, they are simply referred to as the virtual network 11.

  A resource management device 30 is connected to the infrastructure network 10. The resource management device 30 controls the infrastructure network 10 and the virtual network 11 by controlling the nodes 20a to 20d.

  Each virtual network 11 may construct a route or topology in a fixed manner depending on operating conditions, or may perform dynamic topology optimization control to change the topology of each virtual network 11 in response to environmental changes. . With the dynamic topology optimization control, each virtual network 11 can improve adaptability to environmental changes.

  When a plurality of virtual networks 11 are realized on the infrastructure network 10, a mechanism is required that selectively uses packets with fine bandwidth granularity and broadband wavelength paths according to the bandwidth required by the service. As such a mechanism, there is a concept called virtual network topology (VNT).

  The virtual network topology can be regarded as a packet network composed of routers and paths interconnecting them in the infrastructure network 10. By performing path control using GMPLS or the like, the virtual network topology can be dynamically reconfigured. The network quality is stabilized and the resource utilization efficiency and fault tolerance are improved by dynamically reconfiguring the virtual network topology following changes in traffic demands and failures of the virtual network 11. Can do.

  The resource management device 30 dynamically reconfigures the virtual network topology and allocates physical resources to each virtual network 11 so that each virtual network 11 is stably accommodated in the infrastructure network 10. .

  Regarding the mediation of competition between the virtual networks 11, the resources of the infrastructure network 10 so as to suppress the occurrence of physical resource competition from the traffic demand and performance information of each virtual network 11 and the accommodation relationship between the infrastructure network 10 and the virtual network topology. This is realized by designing the distribution amount for each virtual network 11.

The physical resources of the infrastructure network 10 include a link 21 using WDM, TDM, Ethernet (registered trademark), a node 20 such as a router or an optical cross-connect device, a port, a processor, a storage, and a computer.
In the present embodiment, a case is described in which each wavelength of a WDM link is allocated to each virtual network 11 as a physical resource of the infrastructure network 10.

  2A and 2B are diagrams showing details of the infrastructure network and the virtual network in the first embodiment.

FIG. 2A shows the structure of a frame 60 (packet) that flows through the infrastructure network 10 and the virtual network 11.
The frame 60 flowing through the infrastructure network 10 and the virtual network 11 conforms to IEEE 802.1Q. IEEE 802.1Q is a standard that transparently shares the same network link in a network connected by a bridge or the like.

The frame 60 includes a destination address 61 (Destination Address), a source address 62 (Source Address), a tag 63 (Tag), an Ether type 67 (Ether Type), data 68 (Data), and an FCS (Frame Check). Sequence) 69. Further, the tag 63 has a priority 64 (Priority), a CFI (Canonical Format Indicator) 65, and a virtual network identifier 66 (VID: Vertial lan ID).
The destination address 61 stores a MAC (Media Access Control) address that the destination device of the frame 60 has.
The transmission source address 62 stores the MAC address of the transmission source device of the frame 60.

The priority 64 of the tag 63 is a 3-bit field, and stores the priority of the frame 60. The priority 64 indicates the priority of the frame 60 from 0 (lowest) to 7 (highest), and is used for prioritizing various types of traffic (voice, video, data, etc.).
The CFI 65 of the tag 63 is a 1-bit field, and indicates whether or not the MAC address of the frame 60 is an irregular format.

The virtual network identifier 66 of the tag 63 is a 12-bit field and stores the identifier of the virtual network 11 to which the frame 60 belongs. When the virtual network identifier 66 is 0, it means that the frame 60 does not belong to any virtual network 11. In the present embodiment, 1 is set in the virtual network identifier 66 of the frame 60 flowing through the virtual network “# 1” 11-1. In the frame 60 flowing to the virtual network “# 2” 11-2, 2 is set in the virtual network identifier 66.
The ether type 67 stores the protocol type transmitted by the data 68.
Data 68 is a payload of the frame 60.
The FCS 69 is for checking whether or not the frame 60 is broken.

FIG. 2B is a diagram showing details of the links and nodes of the infrastructure network 10 and the virtual network 11.
The infrastructure network 10 includes nodes 20a and 20b and IP routers 50a and 50b. The node 20a and the node 20b are connected via a link 21-1.

  The IP router 50a is connected to the node 20a. The IP router 50b is connected to the node 20b. Hereinafter, when the IP routers 50a and 50b are not particularly distinguished, they are simply referred to as the IP router 50.

  A frame 60 (packet) is transmitted from the virtual network “# 1” 11-1 and the virtual network “# 2” 11-2 to the IP router 50a. Similarly, a frame 60 (packet) is transmitted from the virtual network “# 1” 11-1 and the virtual network “# 2” 11-2 to the IP router 50b.

The frame 60 of each virtual network 11 transmitted from each IP router 50 is relayed to the destination device specified by the destination address 61 via each node 20 and the link 21.
Hereinafter, a method for distributing the reserved resource amount by adjusting the physical resource distribution amount of the infrastructure network 10 will be described.

  Each virtual network 11 uses a part of the resources of the infrastructure network 10 to configure a virtual network topology. Of the physical resources of the infrastructure network 10, the range in which each virtual network 11 can be used is managed by the resource management device 30 (FIG. 1). The resource management device 30 (FIG. 1) collects topology information of the infrastructure network 10, topology information of each virtual network 11, path route information, and traffic demand information, and provides them to each virtual network 11 based on the information. Calculate the amount of resources to be used.

  The resource management device 30 (FIG. 1) of the present embodiment uses a physical resource of a shared type (non-reserved type) and a reserved type attribute because each virtual network 11 flexibly uses physical resources of the infrastructure network 10. Physical resources are provided to each virtual network 11.

Here, the physical resource is a component of the infrastructure network 10 that can be allocated to the virtual network 11. The physical resource is, for example, each wavelength in the WDM link, and each TDM slot in the TDM link. The resource management device 30 presets either a shared attribute or a reserved attribute for these physical resources.
A plurality of virtual networks 11 can be used for physical resources with shared attributes. The reserved physical resource can only be used by the reserved specific virtual network 11. Each virtual network 11 constitutes a virtual network topology by combining a reservation-type physical resource having a use right and a shared-type physical resource.

  3A and 3B are diagrams showing attributes of physical resources in the first embodiment.

  FIG. 3A is a diagram illustrating wavelengths # 1 to # 10 of the link 21. FIG. All wavelengths # 1 to # 10 are shared by the virtual network “# 1” 11-1 and the virtual network “# 2” 11-2. That is, wavelengths # 1 to # 10 are unreserved and have a shared attribute.

  If the quality of one of the virtual networks 11 cannot be satisfied, such as when there is a large amount of interference, the resource management device 30 reserves a part of the physical resource to which the shared attribute is assigned to the virtual network 11 with the degraded quality. Assign as a resource.

FIG. 3B is a diagram illustrating the wavelengths # 1 to # 10 of the link 21.
The wavelengths # 1 to # 5 are reserved for the virtual network “# 1” 11-1. That is, the wavelengths # 1 to # 5 have reservation type attributes.
The wavelengths # 6 to # 10 are shared by the virtual network “# 1” 11-1 and the virtual network “# 2” 11-2. That is, wavelengths # 6 to # 10 are unreserved and have shared attributes.

  The node 20 preferentially sends the frame 60 of the virtual network “# 1” 11-1 to the wavelengths # 1 to # 5. When the link utilization rate of the wavelengths # 1 to # 5 exceeds the threshold value, the node 20 sends the frame 60 of the virtual network “# 1” 11-1 to the wavelengths # 6 to # 10.

  The resource management device 30 of this embodiment provides the wavelength of the WDM link, which is a physical resource of layer 1 independent to each virtual network 11, and constructs the virtual network 11. Thereby, the resource management apparatus 30 can avoid the mutual interference of the physical resources between the virtual networks 11, which is a problem of the network virtualization technology.

FIG. 4 is a schematic configuration diagram showing the resource management apparatus in the first embodiment.
The resource management device 30 includes a storage unit 31, a processing unit 35, and a communication unit 40. The storage unit 31 includes a network state information DB (Database) 32, a traffic demand information DB 33, and a virtual network path information DB 34. The processing unit 35 includes an information collection unit 36, a virtual network selection unit 37, a reserved resource calculation unit 38, and a resource distribution unit 39. The communication unit 40 is connected to the infrastructure network 10.
The storage unit 31 is, for example, a flash memory or an HDD (Hard Disk Drive), and stores information.
The network state information DB 32 is a database that stores information on each link 21 and each node 20 that configures the infrastructure network 10.
The traffic demand information DB 33 is a database that stores information on the traffic volume transferred in each virtual network 11.
The virtual network path information DB 34 is a database that stores information on the relay route of the path 22 set between the nodes 20 of each virtual network 11.

The processing unit 35 is realized, for example, by a CPU (Central Processing Unit) reading and executing a software program in a RAM (Random Access Memory).
The information collecting unit 36 periodically acquires AC traffic information of each virtual network 11 from each node 20 and accumulates it in the traffic demand information DB 33.

  The virtual network selection unit 37 selects a resource that is a distribution control target of a reserved resource and a virtual network 11 that is a distribution control target. The virtual network selection unit 37 determines whether or not each virtual network 11 is appropriately controlled based on the state of each virtual network 11 and selects the virtual network 11 to be subject to distribution control.

The reservation resource calculation unit 38 calculates the allocation amount of the reservation resource allocated to the nodes 20 and the links 21. The reserved resource calculation unit 38 avoids competition of physical resources between the virtual networks 11 by switching a shared physical resource to a reserved physical resource.
The resource allocation unit 39 allocates physical resources by notifying the node 20 and the link 21 of the allocation amount of the reserved resource.
The communication unit 40 communicates with each node 20 of the infrastructure network 10 and is, for example, a NIC (Network Interface Card).

FIGS. 5A to 5C are diagrams illustrating an example of a distribution amount calculation function in the first embodiment.
This allocation amount calculation function F (x) is a function used when the reservation resource calculation unit 38 calculates the allocation amount.
The vertical axis | shaft of Fig.5 (a)-(c) has shown the value of distribution amount calculation function F (x). 5A to 5C indicate the value of the traffic demand ratio x of the virtual network 11. Here, the traffic demand ratio x is the ratio of the traffic demand of the virtual network 11 to the total traffic demand of all the virtual networks 11 for the traffic demand of the virtual network 11.

  FIG. 5A shows an example in which the distribution amount is calculated linearly according to the traffic demand ratio x.

  FIG. 5B shows an example of calculating the distribution amount piecewise in accordance with the traffic demand ratio x. The distribution amount calculation function F (x) has a constant distribution amount when the traffic demand ratio x is C1 to C2 (C1 and C2 are 0 <C1, C1 <C2, and C2 <1). calculate. The allocation amount calculation function F (x) calculates the allocation amount in the range less than C1 or in the range exceeding C2 according to the traffic demand ratio x. At this time, since the predetermined distribution amount is calculated even if the traffic demand ratio x is 0, even when the predetermined traffic demand is newly generated in the physical resource where the traffic demand ratio x is 0. Can respond.

  FIG. 5C shows an allocation amount calculation function F (x) by a curve (secondary function, exponential function, etc.). The allocation amount calculation function F (x) in FIG. 5C calculates the allocation amount so that the physical resources provided to the virtual network 11 increase synergistically as the traffic demand ratio x increases. As a result, the virtual network 11 having a large traffic demand ratio x is preferentially allocated resources, and for example, the virtual network 11 centering on the video streaming service can view the video without stress.

(Operation of the first embodiment)

FIG. 6 is a flowchart showing the resource allocation amount calculation processing in the first embodiment.
When the processing is started, in step S10, the virtual network selection unit 37 of the resource management device 30 performs the performance status of each virtual network 11 in the virtual network path information DB 34 and the traffic demand of each virtual network 11 in the traffic demand information DB 33. From the above information, a set of virtual networks 11 to which the reservation resource is to be allocated is selected. That is, the virtual network selection unit 37 selects the virtual network 11 that is likely to cause a problem.

  In step S11, the virtual network selection unit 37 of the resource management device 30 configures the infrastructure network 10, and among the physical resources contained in the set of virtual networks 11 selected in step S10, competition is predicted. Therefore, a physical resource that requires a reservation type setting because it is likely to become a bottleneck is selected.

  In step S12, the reserved resource calculation unit 38 of the resource management device 30 calculates the traffic demand ratio x of the set of each virtual network 11 in the selected physical resource.

In step S <b> 13, the reserved resource calculation unit 38 of the resource management device 30 calculates the reserved resource amount to be allocated to each set of virtual networks 11 based on the traffic demand ratio x and the distribution amount calculation function F (x). When the process of step S13 ends, the reserved resource calculation unit 38 of the resource management device 30 ends the process of FIG.
Hereinafter, the process of each step S10-S14 is demonstrated in detail.

<< Process of Step S10 >>
The process of step S10 is a process of the selection method of the virtual network 11. There are the following methods for selecting a virtual network to be controlled. The virtual network 11 to be controlled can be selected by combining one or more of the following methods.

  The first virtual network selection method is to select the virtual network 11 whose performance is degraded. In the first virtual network selection method, the virtual network selection unit 37, for example, keeps the maximum link utilization rate (performance index) of the virtual network 11 continuously in advance for a certain period of time (for example, 1 hour). It is determined whether or not a set threshold value (for example, 80%) is exceeded. If the determination condition is satisfied, the virtual network 11 is selected.

  The second virtual network selection method is a method of selecting the virtual network 11 having a problem that traffic demand is large. In the second virtual network selection method, the virtual network selection unit 37, for example, has a preset threshold value (for example, 10 Gbps) for the total traffic demand of the virtual network 11 or the total demand between specific grounds. It is determined whether or not the virtual network 11 is exceeded. If the determination condition is satisfied, the virtual network 11 is selected.

  The third virtual network selection method is a method of selecting the virtual network 11 using resources that are likely to become bottlenecks. The third virtual network selection method is important for improving the performance of other virtual networks 11. In the third virtual network selection method, the virtual network selection unit 37 is a physical resource in which a predetermined number or more of virtual networks 11 are accommodated when, for example, the communication path of each virtual network 11 is designed to be accommodated by the shortest path or the like. And the virtual network 11 accommodating the physical resource is selected.

<< Process of Step S11 >>
The process of step S11 is a process of selecting a physical resource to be allocated to the set of virtual networks 11.
In the process of step S11, the virtual network selection unit 37 of the resource management device 30 causes a physical resource in which contention occurs among the physical resources accommodated by the set of virtual networks 11 or occurrence of contention. Set a reserved attribute for the expected physical resource. For example, if the usage rate of the physical resource exceeds a threshold value (for example, 90%), the virtual network selection unit 37 of the resource management device 30 determines that occurrence of physical resource contention is expected.

<< Step S12 >>
The process of step S12 is a process of calculating each traffic demand ratio x for the set of virtual networks 11 selected in step S10.
The reserved resource calculation unit 38 of the resource management device 30 calculates the traffic demand ratio x in units of physical resources that are allocation targets of reserved resources. The reserved resource calculation unit 38 of the resource management device 30 calculates the traffic demand ratio of the virtual network 11 with respect to the total traffic demand after calculating the total traffic demand of all the virtual networks 11 passing through the physical resource. .

<< Process of Step S13 >>
The process of step S13 is a process of calculating a reservation resource to be allocated to the virtual network 11 based on the traffic demand ratio x of the virtual network 11. The reserved resource calculation unit 38 of the resource management device 30 calculates the distribution amount by the distribution amount calculation function F (x) according to the following calculation formula in units of physical resources to be calculated. The reserved resource allocation amount of the virtual network 11 is calculated by the following formula 1.

  The distribution amount calculation function F (x) is shown in FIG. The reservation resource calculation unit 38 of the resource management device 30 can calculate the distribution amount of the reservation resource to the virtual network 11 using Equation 1. Among the physical resources other than those for which reservation type attributes are set, shared type attributes are set.

  The reserved resource calculation unit 38 of the resource management device 30 performs the calculation of Equation 1 for all the physical resources selected in the process of step S11. Here, the virtual network 11 to be calculated is the virtual network 11 to be controlled that is selected in the process of step S10.

When a finite physical resource of the infrastructure network 10 is shared by a plurality of virtual networks 11, there is a risk that physical resource contention may occur between the virtual networks 11.
In the present embodiment, instead of directly arbitrating physical resource contention by communication of each virtual network 11, when the resource management device 30 performs network virtualization of the infrastructure network 10, Coordinate physical resources to be allocated and indirectly arbitrate physical resource contention. Thereby, each virtual network 11 can avoid the overhead by communication and can improve the robustness of the whole system.

The resource management apparatus 30 according to the present embodiment sets two types, a reserved type and a shared type, as physical resource attributes when defining the physical resource allocation amount. The resource management device 30 can guarantee the bandwidth of the virtual network 11 and avoid mutual interference between the virtual networks 11 by setting a reservation type attribute for the physical resource. Further, the resource management device 30 sets a shared attribute for the physical resource, thereby allowing the physical resource to be interchanged between the virtual networks 11, improving resistance to traffic fluctuation, and improving the resource utilization efficiency of the infrastructure network 10. Can be improved.
(Effects of the first embodiment)
The first embodiment described above has the following effects (A) to (C).

(A) The resource management device 30 sets reservation-type attributes for physical resources that can be used by each virtual network 11, and changes the configuration of each virtual network 11 in an adaptive manner. Thereby, even if the number of virtual networks 11 increases, the mutual interference between the virtual networks 11 can be easily minimized.

(B) The resource management device 30 can guarantee the bandwidth of the virtual network 11 and avoid mutual interference between the virtual networks 11 by setting a reservation type attribute in the physical resource.

(C) The resource management apparatus 30 sets a shared attribute for the physical resource, thereby allowing the physical resource to be interchanged between the virtual networks 11 and improving the resistance to traffic fluctuation, and improving the resource utilization efficiency of the infrastructure network 10. Can be improved.

(D) The reserved resource calculation unit 38 calculates a physical resource to be provided to the selected virtual network 11 from the traffic demand information by using a piecewise linear function. Thus, even if the traffic demand information is zero, Can be reserved for use of the physical resource.

(E) The reserved resource calculation unit 38 calculates so that the physical resources provided to the selected virtual network 11 increase synergistically. As a result, many physical resources can be reserved in an important virtual network 11 that consumes many physical resources.

(Configuration of Second Embodiment)
The resource management device 30 according to the second embodiment is characterized in that the processing load is reduced by further reducing the amount of calculation than the resource management device 30 according to the first embodiment.
The configurations of the infrastructure network 10 and each virtual network 11 of the second embodiment are the same as the configurations of the infrastructure network 10 and each virtual network 11 (FIG. 1) of the first embodiment.
The configuration of the resource management device 30 of the second embodiment is the same as the configuration of the resource management device 30 (FIG. 4) of the first embodiment.

(Operation of Second Embodiment)
FIG. 7 is a diagram illustrating a resource allocation amount calculation process in the second embodiment. The same symbols are assigned to the same elements as the resource allocation amount calculation processing of the first embodiment shown in FIG.
When the process is started, in step S10a, the virtual network selection unit 37 of the resource management device 30 selects a physical resource that is likely to become a bottleneck based on the traffic demand information DB 33 among the physical resources that constitute the infrastructure network 10. . That is, the virtual network selection unit 37 selects a physical resource for which a conflict is particularly predicted and reservation type setting is required.

In step S11a, the virtual network selection unit 37 of the resource management device 30 accommodates the physical resource selected from the virtual network 11 that accommodates the physical resource selected in the process of step S10a. The virtual network 11 that is likely to become is selected. That is, the virtual network selection unit 37 has a performance state of each virtual network 11 in the virtual network path information DB 34, traffic demand information of each virtual network 11 in the traffic demand information DB 33, and the virtual network 11 to be allocated as a reservation resource. Select a set.
The processes in steps S12 and S13 are the same as the processes in steps S12 and S13 (FIG. 6) of the first embodiment.

(Effect of 2nd Embodiment)
The second embodiment described above has the following effect (F).

(F) The virtual network selection unit 37 of the resource management device 30 selects a physical resource that is predicted to be particularly competitive, and then selects a virtual network from which the reserved resource is allocated from the virtual network 11 that accommodates the physical resource. Eleven sets are selected. Thereby, even when the number of virtual networks 11 increases, the processing load can be reduced by reducing the amount of calculation by selecting the physical resources first and narrowing down the virtual networks 11.

(Modification)
The present invention is not limited to the above embodiment, and can be modified without departing from the spirit of the present invention. For example, there are the following forms (a) to (b) as usage forms and modifications.

(A) In the first embodiment, each wavelength of the WDM link is distributed to each virtual network 11 as a physical resource of the infrastructure network 10. However, the present invention is not limited to this, and each TDM slot, node 20, port, processor, storage, or computer of the TDM link may be allocated to each virtual network 11.

(B) The virtual network 11 of the first embodiment is a virtual network (Virtual Private Network). However, the present invention is not limited to this, and any general logical network in which each node is connected by a path may be used.

10 Infrastructure Network 11 Virtual Network 20 Node 21 Link 22 Path 30 Resource Management Device 31 Storage Unit 32 Network Status Information DB (Infrastructure Network Topology Information)
33 Traffic demand information DB (Traffic demand information)
34 Virtual network path information DB (virtual network topology information)
35 Processing Unit 36 Information Collection Unit 37 Virtual Network Selection Unit 38 Reserved Resource Calculation Unit 39 Resource Distribution Unit 40 Communication Unit 50 IP Router 60 Frame

Claims (7)

An information collection unit that collects traffic demand information related to each virtual network based on infrastructure network topology information related to the infrastructure network and virtual network topology information related to each virtual network constructed on the infrastructure network;
A virtual network selection unit that selects a control target virtual network from each of the virtual networks based on the traffic demand information, and selects a control target physical resource from each physical resource related to the infrastructure network based on the virtual network;
A reservation resource calculation unit that calculates the physical resource to be provided to the selected virtual network by a piecewise linear function based on the traffic demand information collected by the information collection unit;
A network resource management apparatus comprising: An information collection unit that collects traffic demand information related to each virtual network based on infrastructure network topology information related to the infrastructure network and virtual network topology information related to each virtual network constructed on the infrastructure network;
A virtual network selection unit that selects a control target virtual network from each of the virtual networks based on the traffic demand information, and selects a control target physical resource from each physical resource related to the infrastructure network based on the virtual network;
A reservation resource calculation unit that calculates the physical resource to be provided to the selected virtual network based on the traffic demand information collected by the information collection unit so as to increase synergistically with the traffic demand information ;
A network resource management apparatus comprising: Based on infrastructure network topology information related to the infrastructure network and virtual network topology information related to each virtual network constructed on the infrastructure network, traffic demand information related to each virtual network of each physical resource related to the infrastructure network An information collection unit that collects
A virtual network selection unit that selects a physical resource to be controlled from each physical resource based on the traffic demand information, and selects a virtual network to be controlled from each virtual network based on the physical resource;
A reservation resource calculation unit that calculates the physical resource to be provided to the selected virtual network based on the traffic demand information collected by the information collection unit;
A network resource management apparatus comprising: The reserved resource calculation unit calculates the physical resource to be provided to the selected virtual network from the traffic demand information by a piecewise linear function.
The network resource management apparatus according to claim 3 , wherein the network resource management apparatus is a network resource management apparatus. The reservation resource calculation unit calculates the traffic demand information so that the physical resources provided to the selected virtual network increase synergistically.
The network resource management apparatus according to claim 3 , wherein the network resource management apparatus is a network resource management apparatus. A resource distribution unit that reserves and distributes the physical resource calculated by the reservation resource calculation unit to the virtual network;
Furthermore the resource management device of the network according to any one of claims 1 to 5, characterized in that it comprises. The reservation resource calculation unit
Based on the traffic demand information relating to each of the virtual networks, the allocation amount of the reservation resource is calculated,
Set either a shared attribute or a reserved attribute on the physical resource.
Resource management device of the network according to any one of claims 1 to 6, characterized in that.

Images