CN112822038A - Resource allocation method for virtual optical network mapping cost facing data center - Google Patents

Resource allocation method for virtual optical network mapping cost facing data center Download PDF

Info

Publication number
CN112822038A
CN112822038A CN202011621746.6A CN202011621746A CN112822038A CN 112822038 A CN112822038 A CN 112822038A CN 202011621746 A CN202011621746 A CN 202011621746A CN 112822038 A CN112822038 A CN 112822038A
Authority
CN
China
Prior art keywords
virtual
optical network
data center
optical
spectrum
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011621746.6A
Other languages
Chinese (zh)
Other versions
CN112822038B (en
Inventor
揭水平
陈伯文
吴水清
符小东
马宗仰
房洪莲
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhongtian Communication Technology Co ltd
Jiangsu Zhongtian Technology Co Ltd
Zhongtian Broadband Technology Co Ltd
Original Assignee
Zhongtian Communication Technology Co ltd
Jiangsu Zhongtian Technology Co Ltd
Zhongtian Broadband Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhongtian Communication Technology Co ltd, Jiangsu Zhongtian Technology Co Ltd, Zhongtian Broadband Technology Co Ltd filed Critical Zhongtian Communication Technology Co ltd
Priority to CN202011621746.6A priority Critical patent/CN112822038B/en
Publication of CN112822038A publication Critical patent/CN112822038A/en
Application granted granted Critical
Publication of CN112822038B publication Critical patent/CN112822038B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0896—Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0893—Assignment of logical groups to network elements
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q11/00—Selecting arrangements for multiplex systems
    • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
    • H04Q11/0062—Network aspects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q11/00—Selecting arrangements for multiplex systems
    • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
    • H04Q11/0062—Network aspects
    • H04Q2011/0086—Network resource allocation, dimensioning or optimisation

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Algebra (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Optical Communication System (AREA)

Abstract

The invention discloses a resource allocation method of virtual optical network mapping cost facing a data center, which comprises the steps of initializing a network, establishing an optimization method taking the lowest virtual network mapping cost as an objective function, establishing constraint conditions meeting the objective function optimization method, and calculating the number of optical regenerators. The invention can greatly reduce the number of optical regenerators used in the mapping process of the virtual optical network, thereby reducing the mapping cost of the virtual optical network and enabling a network operator to obtain higher profit when the same virtual optical network is mapped.

Description

Resource allocation method for virtual optical network mapping cost facing data center
Technical Field
The invention relates to a resource allocation method, in particular to a data center-oriented resource allocation method for virtual optical network mapping cost, and belongs to the field of optical communication.
Background
The 5G service, artificial intelligence, cloud computing, big data, intelligent home and the rapid development of the internet of things make the network resource efficiency a serious challenge problem. In addition, as network data storage and computing capabilities are increased, there is an urgent need to increase the storage capacity of data center networks to provide sufficient bandwidth access and data storage capacity. By considering the optical network virtualization technology facing the data center, the problem of resource allocation and scheduling rigidity of the optical network can be solved. This is because network virtualization technology provides an efficient way to share physical computing and network resources among multiple virtual network requests, and provides greater flexibility and efficiency in allocating resources. Virtual Optical Network (VON) mapping is one of the basic contents of network virtualization research, and the main purpose of the VON virtual optical network is to map the VON virtual optical network to a physical network on the basis of satisfying the node computing resource and bandwidth resource constraints, so as to realize the effective utilization of physical layer resources. Therefore, it can be seen that research on a network virtualization technology in a data center-oriented space division multiplexing flexible optical network (SDM-EON) is one of important approaches to solve the current network resource efficiency problem.
Virtual optical network mapping currently faces three major challenges: the difficulties of resource constraint, topology diversification and instantaneity of online requests lead to the difficulty of finding the optimal solution (NP-Hard) of the online request in polynomial time, namely the NP-Hard problem. Virtual network mapping is an NP-Hard problem, and an approximate solution is generally found by designing a heuristic algorithm. To date, many efforts have been made to map virtual networks on spectrally flexible optical networks. For example, how to reduce the mapping cost of the underlying physical network, how to reduce the blocking rate and improve the resource utilization rate, how to balance the load of the underlying network nodes and links, and the like. However, how to increase the total profit after virtual network mapping is a main objective for network operators, so optimizing the mapping cost of the network and increasing the resource utilization rate of the network will become a key technology to optimize the network operation.
Disclosure of Invention
The invention discloses a resource allocation method for mapping cost of a virtual optical network facing a data center, which realizes the lowest mapping cost of the virtual optical network in a space division multiplexing spectrum flexible optical network facing the data center and improves the total income of a network operator.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a resource allocation method for virtual optical network mapping cost facing to a data center is characterized by comprising the following steps:
the method comprises the following steps: initializing a network;
step two: establishing an optimization method taking the lowest virtual network mapping cost as an objective function;
step three: establishing a constraint condition meeting the objective function optimization method;
step four: the number of optical regenerators is counted.
Further, the step one is specifically
Given a physical network Gp(Vp，Ep，Cp，Fp) Wherein V ispRepresenting a group of data centers, EpRepresents a set of physical links, CpRepresenting the computing resources owned by each data center, FpRepresents the number of spectrum slots available on each link;
group of virtual optical networks Gv(Vv，Ev，Cv，Bv) Wherein V isv，Ev，CvAnd BvRespectively representing a set of virtual nodes, a set of virtual links, a set of node computing resource requirements and a set of link bandwidth requirements;
setting | Vp|、|Ep|、|Cp|、|FpThe number of the | respectively represents the number of data centers in the physical optical network, the number of optical fiber links, the number of available computing resources of each data center and the number of spectrum gaps of each fiber core;
while setting the number of cores per fiber.
Further, the second step is specifically
For the cost problem of virtual optical network mapping in the space division multiplexing spectrum flexible optical network facing the data center, an optimization objective function is expressed by the following sub-formula:
Figure BDA0002874104290000031
since the mapping cost of the virtual optical network is only related to the number of optical repeaters and optical regenerators, the objective of equation (2) is to use the minimum number of optical repeaters and optical regenerators while satisfying the relevant constraints of the virtual optical network mapping.
Further, the third step is specifically
When the virtual optical network is mapped, the nodes and the data center need to satisfy the following constraint conditions:
(1) optical switching node constraints;
(2) the data center calculates resource capacity constraints;
(3) a flow conservation constraint;
(4) a routing constraint;
(5) a bandwidth capacity constraint;
(6) a spectral continuity constraint;
(7) spectrum consistency constraint;
(8) the core is uniquely constrained.
Further, the optical switching node has the constraint condition of
Figure BDA0002874104290000041
Figure BDA0002874104290000042
Wherein,
Figure BDA0002874104290000043
the value is a binary variable, if the virtual node i is successfully mapped to the optical switching node k, the value is 1, otherwise, the value is 0;
the data center computing resource capacity is constrained to
Figure BDA0002874104290000044
Wherein,
Figure BDA0002874104290000045
representing the computational resource requirement of virtual node i on the nth virtual optical network, CkRepresenting the amount of total computational resources of data center k on a data center oriented space division multiplexed spectrum flexible optical network.
Further, the flow conservation constraint is
Figure BDA0002874104290000046
Wherein,
Figure BDA0002874104290000047
is a binary variable, if the virtual link (i, j) on the nth virtual optical network occupies the f slot of the c core of the physical link (k, l), it is 1, otherwise it is 0,
Figure BDA0002874104290000048
the number of spectral slots required for the virtual link (i, j) is indicated.
Further, the routing constraints are
Figure BDA0002874104290000049
Wherein,
Figure BDA0002874104290000051
is a binary variable, if the virtual link (i, j) on the nth virtual optical network is successfully mapped to the data center pair (k, l), it is 1, otherwise it is 0;
the bandwidth capacity is constrained to
Figure BDA0002874104290000052
If is the number of spectral slots on each physical link.
Further, the spectral continuity is constrained to
Figure BDA0002874104290000053
Figure BDA0002874104290000054
The spectrum resources allocated on the physical path for a virtual link must be contiguous, in equation (9), if
Figure BDA0002874104290000055
And is
Figure BDA0002874104290000056
All spectral gaps with index values higher than f +1 are not used for mapping virtual links on node pair (k, l); in the formula (10), if
Figure BDA0002874104290000057
Then no spectrum slots with index values below f will be allocated to virtual link (i, j).
To reduce the complexity of the calculation, it is necessary to reduce the binary variables
Figure BDA0002874104290000058
Of the dimension space of (1), thus introducing additional binary variables
Figure BDA0002874104290000059
And
Figure BDA00028741042900000510
to decompose
Figure BDA00028741042900000511
Wherein
Figure BDA00028741042900000512
Whether the virtual link (i, j) on the nth virtual optical network occupies the physical link (k, l) or not is represented;
Figure BDA00028741042900000513
representing whether the fiber core c is occupied by a virtual link (i, j) on the nth virtual optical network;
Figure BDA00028741042900000514
representing whether the virtual link (i, j) on the nth virtual optical network occupies a spectrum gap with index f; binary variable
Figure BDA0002874104290000061
And
Figure BDA0002874104290000062
the relationship between them is:
Figure BDA0002874104290000063
Figure BDA0002874104290000064
Figure BDA0002874104290000065
further, the spectrum consistency is constrained to
Figure BDA0002874104290000066
The core is uniquely constrained to
Figure BDA0002874104290000067
Further, the fourth step is specifically
The cost is mainly related to the number of optical repeaters and optical regenerators, which is fixed since the virtual optical network is given and uses a single line rate, where the goal is to optimize the number of optical regenerators; after a virtual optical network has been successfully mapped, the number of optical regenerators required can be derived from equation (16), where R represents the maximum reachable distance of the optical regenerator at the selected line rate and modulation format, D(k,l)Refers to the distance of the physical link (k, l);
Figure BDA0002874104290000068
the method for allocating the resources of the route, the spectrum and the fiber core when the virtual optical network is mapped in the space division multiplexing spectrum flexible optical network facing the data center can be found out through the constraint conditions and the calculation formula of the optical regenerator.
Compared with the prior art, the invention has the following advantages and effects: the resource allocation method for the virtual optical network mapping cost oriented to the data center provides an evaluation mechanism for the virtual optical network mapping cost, then establishes an optimization method aiming at the lowest network mapping cost according to the mechanism, and realizes the problems of routing calculation, fiber core selection, spectrum allocation and the like of the virtual optical network in a multi-core optical fiber network by an integer linear programming method. A certain virtual optical network mapping request is generated in a space division multiplexing spectrum flexible optical network facing a data center, sufficient spectrum resources and computing resources are given to be allocated, and request blockage is avoided. And then, according to the constraint conditions and the optimization target, establishing a mapping method with the lowest cost, thereby finding the optimal mapping scheme and resource allocation method for all the virtual optical network mapping requests. The method can greatly reduce the number of optical regenerators used in the mapping process of the virtual optical network, thereby reducing the mapping cost of the virtual optical network, and can enable a network operator to obtain higher profit when the same virtual optical network is mapped. Meanwhile, the shortest physical path is found for each virtual link request, the waste of frequency spectrum resources in the multi-core optical fiber network is reduced, and the utilization rate of the frequency spectrum resources is improved.
Drawings
Fig. 1 is a flowchart of a resource allocation method for data center-oriented virtual optical network mapping cost according to the present invention.
Fig. 2 is a schematic diagram of a physical optical network according to an embodiment of the present invention.
FIG. 3 is a schematic view of a three-core optical fiber according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of mapping VON1 according to an embodiment of the invention.
Fig. 5 is a schematic diagram of mapping VON2 according to an embodiment of the invention.
Fig. 6 is a diagram of a mapping scheme of an embodiment of the invention.
Detailed Description
To elaborate on technical solutions adopted by the present invention to achieve predetermined technical objects, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, it is obvious that the described embodiments are only partial embodiments of the present invention, not all embodiments, and technical means or technical features in the embodiments of the present invention may be replaced without creative efforts, and the present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.
In order to solve the problem of the mapping cost of the virtual optical network in the space division multiplexing spectrum flexible optical network facing the data center, the invention provides an optimization method with the lowest mapping cost of the virtual optical network, and meets the constraint conditions of bandwidth flow conservation, node mapping uniqueness, spectrum consistency, spectrum continuity, spectrum resource occupation uniqueness, data center computing resource requirements and the like, thereby realizing the distribution method of space division multiplexing spectrum flexible optical network routing, spectrum and fiber core resources facing the data center.
The cost of mapping a virtual optical network in a space division multiplexed spectrally flexible optical network is mainly composed of the cost of optical repeaters and optical regenerators. Since the number of optical repeaters and optical regenerators required depends on the line rate, the spectral width and the modulation format of the optical channel, the invention uses a single line rate to transmit traffic in order to simplify the virtual optical network mapping problem. For each virtual link, a pair of optical repeaters needs to be configured on the source node and the sink node when it is mapped to the data center node pair. If the transmission distance of the bandwidth service of the virtual link in the channel exceeds the maximum reachable distance of the optical signal, a corresponding number of optical regenerators need to be configured on the intermediate optical switching node to meet the transmission quality of the optical signal. The cost model for mapping the virtual optical network is defined as shown in formula (1).
Figure BDA0002874104290000091
Wherein G isvIs a given set of virtual optical networks,
Figure BDA0002874104290000092
is a set of virtual links on the nth virtual optical network, TC represents the unit price of the optical repeater, RC represents the unit price of the optical regenerator;
Figure BDA0002874104290000093
and
Figure BDA0002874104290000094
respectively representing virtual links
Figure BDA0002874104290000095
The number of optical repeaters and optical regenerators required. It is an object of the present invention to minimize the virtual optical network mapping cost.
In order to solve the problem of the mapping cost of the virtual optical network in the space division multiplexing spectrum flexible optical network facing the data center, the invention provides a virtual optical network mapping integer linear programming model facing the data center on the basis of the cost evaluation mechanism, namely an optimization method aiming at the lowest cost is realized in the optical network facing the data center. As shown in fig. 1, the method for allocating resources to virtual optical network mapping cost for a data center of the present invention is characterized by comprising the following steps:
the method comprises the following steps: and (5) initializing the network.
Given a physical network Gp(Vp,Ep,Cp,Fp) In which V ispRepresenting a group of data centers, EpRepresents a set of physical links, CpRepresenting the computing resources owned by each data center; fpRepresenting the number of spectrum slots available on each link. Group of virtual optical networks Gv(Vv,Ev,Cv,Bv) In which V isv,Ev,CvAnd BvRespectively representing a set of virtual nodes, a set of virtual links, a set of node computing resource requirements, and a set of link bandwidth requirements. Setting | Vp|、|Ep|、|Cp|、|FpThe number of | represents the number of data centers, the number of optical fiber links, the number of available computing resources of each data center and the number of spectrum gaps of each fiber core in the physical optical network respectively. While setting the number of cores per fiber.
Step two: and establishing an optimization method taking the lowest virtual optical network mapping cost as an objective function.
The invention mainly solves the cost problem of virtual optical network mapping in the space division multiplexing spectrum flexible optical network facing a data center, and an optimization objective function can be expressed by the following sub-formula:
Figure BDA0002874104290000101
since the mapping cost of the virtual optical network is only related to the number of optical repeaters and optical regenerators, the objective here is to use as few optical repeaters and optical regenerators as possible while satisfying the relevant constraints of the virtual optical network mapping.
Step three: and establishing constraint conditions meeting the objective function optimization method.
When the virtual optical network is mapped, the nodes and the data center need to satisfy the following constraint conditions:
(1) optical switching node constraint conditions:
Figure BDA0002874104290000102
Figure BDA0002874104290000103
wherein,
Figure BDA0002874104290000104
is a binary variable, the value is 1 if the virtual node i maps successfully to the optical switching node k, otherwise it is 0. The constraint condition (3) ensures that one virtual node has and can only be mapped to one optical switching node, and the constraint condition (4) ensures that two different virtual nodes in the same virtual optical network cannot be mapped to the same optical switching node.
(2) Data center computing resource capacity constraints:
Figure BDA0002874104290000105
wherein,
Figure BDA0002874104290000106
representing the computational resource requirement of virtual node i on the nth virtual optical network, CkRepresented in data orientedThe total amount of computational resources for data center k on a spatial division multiplexed spectrally flexible optical network. Constraint (5) ensures that the total computing resource requirement of all virtual nodes mapped to the same data center does not exceed the total number of computing resources of the data center.
When spectrum resources and data center computing resources in the space division multiplexing spectrum flexible optical network facing to the data center are allocated and optimized, the following constraint conditions are also required to be met:
(3) flow conservation constraint:
Figure BDA0002874104290000111
wherein,
Figure BDA0002874104290000112
is a binary variable, if the virtual link (i, j) on the nth virtual optical network occupies the f slot of the c core of the physical link (k, l), it is 1, otherwise it is 0,
Figure BDA0002874104290000113
the number of spectral slots required for the virtual link (i, j) is indicated. The constraint (6) ensures that the number of output spectral slots of the optical channel on each data center, except for the source node and the destination node, is equal to the number of input spectral slots, so-called traffic conservation.
(4) And (4) routing constraint:
Figure BDA0002874104290000114
wherein,
Figure BDA0002874104290000115
is a binary variable, which is 1 if the virtual link (i, j) on the nth virtual optical network is successfully mapped onto the data center pair (k, l), otherwise it is 0. Constraint (7) ensures that two end nodes of a physical link to which a virtual link is mapped must be mapped with a data center to which nodes at both ends of the virtual link are mappedAnd correspondingly. Equation (7) also ensures that the virtual link (i, j) can only occupy one optical channel.
(5) And (3) bandwidth capacity constraint:
Figure BDA0002874104290000116
and | F | is the number of the spectrum gaps on each physical link, and the constraint (8) ensures that the number of the spectrum gaps occupied on each link optical fiber does not exceed the total spectrum gaps.
(6) And (3) spectrum continuity constraint:
Figure BDA0002874104290000121
Figure BDA0002874104290000122
the spectrum resources allocated on the physical path for a virtual link must be contiguous. In the formula (9), if
Figure BDA0002874104290000123
And is
Figure BDA0002874104290000124
All spectral gaps with index values higher than f +1 are not used for mapping virtual links on node pair (k, l). In the formula (10), if
Figure BDA0002874104290000125
Then no spectrum slots with index values below f will be allocated to virtual link (i, j). Equations (9) and (10) ensure the constraint of spectral continuity.
To reduce the complexity of the calculation, it is necessary to reduce the binary variables
Figure BDA0002874104290000126
Is measured in the dimension space of (a). Thus introducing an extra binary variable
Figure BDA0002874104290000127
And
Figure BDA0002874104290000128
to decompose
Figure BDA0002874104290000129
Wherein
Figure BDA00028741042900001210
Whether the virtual link (i, j) on the nth virtual optical network occupies the physical link (k, l) or not is represented;
Figure BDA00028741042900001211
representing whether the fiber core c is occupied by a virtual link (i, j) on the nth virtual optical network;
Figure BDA00028741042900001212
it is represented whether the virtual link (i, j) on the nth virtual optical network occupies a spectral gap with index f. Binary variable
Figure BDA00028741042900001213
Figure BDA00028741042900001214
And
Figure BDA00028741042900001215
the relationship between them is:
Figure BDA00028741042900001216
Figure BDA00028741042900001217
Figure BDA0002874104290000131
Figure BDA0002874104290000132
(7) and (3) restraining the consistency of the frequency spectrum:
Figure BDA0002874104290000133
equation (14) ensures that for each virtual link mapping request, the same spectrum gap, i.e., the spectrum consistency constraint, is allocated on each segment of the link of the selected physical path. In other words, the number of indices selected over all spectrum slots can only be equal to the number of spectrum slots required for the virtual link.
(8) Core uniqueness constraint:
Figure BDA0002874104290000134
each virtual link mapping request occupies the same core index on different links of the physical path, and formula (15) ensures the uniqueness constraint of the core.
Step four: and (4) calculating the number of optical regenerators.
As can be seen from step two, the cost in the present invention is mainly related to the number of optical repeaters and optical regenerators. Since the virtual optical network is given and the present invention employs a single line rate, the number of optical repeaters is fixed, where the goal is to optimize the number of optical regenerators. After a virtual optical network has been successfully mapped, the number of optical regenerators required can be derived from equation (16), where R represents the maximum reachable distance of the optical regenerator at the selected line rate and modulation format, D(k,l)Refers to the distance of the physical link (k, l).
Figure BDA0002874104290000135
Through the constraint conditions and the calculation formula of the optical regenerator, the distribution method of the routing, the spectrum and the fiber core resources when the virtual optical network is mapped in the space division multiplexing spectrum flexible optical network facing the data center can be found out, so that the optimization objective function of the integer linear programming is realized.
In order to further understand the optimization method proposed in the present invention, the following detailed description of the specific implementation method of the present invention is provided in conjunction with the relevant examples, and the specific example steps are as follows:
the method comprises the following steps: taking the network topology shown in fig. 2 as an example, the network is a data center-oriented space division multiplexing spectrum flexible optical network with 6 nodes and 8 optical fiber links. In fig. 2, the data centers of the optical network are indicated by circles and the optical switching nodes are indicated by 0, 1, 2, 3, 4, 5; the number on the dashed circle adjacent to the optical switching node represents the number of computing resources provided by the data center; the solid links between the data centers represent optical fiber links, each of which is bidirectional and comprises a three-core optical fiber having a structure as shown in fig. 3, and the numbers beside them represent the transmission distance between the two data centers in kilometers (km); the spectral bandwidth of each optical fiber link is set to be 50GHz, and the bandwidth of each spectral gap is set to be 12.5GHz, namely, each fiber core has 4 spectral gaps.
Step two: mapping service requests VON1 and VON2 for two virtual optical networks are generated in a multi-core fiber network as shown in fig. 4 and 5. The virtual nodes are represented by regular hexagons, A, B, C respectively represent virtual node numbers of the virtual optical network, the number in a dashed circle represents the number of computing resources required by the virtual node, each virtual link is represented by a dashed line, and the number on the dashed line represents the number of required spectrum gaps between two different virtual nodes.
Step three: and establishing and executing an objective function with optimal virtual optical network mapping cost in the data center-oriented space division multiplexing spectrum flexible optical network, namely executing a formula (2).
Step four: and establishing and executing various constraint conditions of the optimization method of the virtual optical network mapping cost in the space division multiplexing spectrum flexible optical network facing the data center. In the process of allocating resources for each virtual optical network, the constraints of calculating resource capacity of the optical switching node and the data center, namely formula (3-5), flow conservation and routing constraints, namely formula (6-7), spectrum resource capacity constraints, namely formula (8), spectrum consistency and spectrum continuity constraints, namely formula (9-14) and the constraint condition of uniqueness of fiber core resource occupation, namely formula (15) are met.
Step five: the number of optical regenerators required for the selected mapping scheme is calculated according to equation (16) to determine the cost.
Through the steps, the VON1 and VON2 can be mapped and allocated with corresponding resources based on the target conditions. In order to achieve the purpose of optimal cost, the invention uses the optical repeaters as few as possible, and when the mapping nodes are searched for each virtual optical network, the path with the least hop count is selected as much as possible, so that the use of the optical repeaters is reduced. As shown in fig. 6, the white grid on the link represents the available spectral slots and the red grid represents the occupied spectral slots. VONs 1 and VONs 2 are mapped as shown in fig. 6 according to the above constraints and objective function, and the green grid on the link is the spectrum gap occupied by the virtual link. As can be seen from the figure, the above constraints can be satisfied, and neither virtual optical network uses an optical regenerator, so the cost of mapping must be the lowest.
The invention mainly solves the problem of the mapping cost of a virtual optical network in a space division multiplexing spectrum flexible optical network facing a data center. Because the network virtualization can well solve the problem of network rigidity, and the single-core optical fiber reaches the physical limit, the research of mapping the virtual optical network in the multi-core optical fiber network becomes an important way for solving the current network problem under the support of the space division multiplexing technology. The invention provides an evaluation mechanism of the mapping cost of a virtual optical network, then an optimization method which takes the lowest network mapping cost as a target is established according to the evaluation mechanism, and the problems of routing calculation, fiber core selection, spectrum allocation and the like of the virtual optical network in a multi-core optical fiber network are realized by an integer linear programming method. A certain virtual optical network mapping request is generated in a space division multiplexing spectrum flexible optical network facing a data center, sufficient spectrum resources and computing resources are given to be allocated, and request blockage is avoided. And then, according to the constraint conditions and the optimization target, establishing a mapping method with the lowest cost, thereby finding the optimal mapping scheme and resource allocation method for all the virtual optical network mapping requests.
The method can greatly reduce the number of optical regenerators used in the mapping process of the virtual optical network, thereby reducing the mapping cost of the virtual optical network, and can enable a network operator to obtain higher profit when the same virtual optical network is mapped. Meanwhile, the shortest physical path is found for each virtual link request, the waste of frequency spectrum resources in the multi-core optical fiber network is reduced, and the utilization rate of the frequency spectrum resources is improved.
Although the present invention has been described with reference to a preferred embodiment, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (10)

1. A resource allocation method for virtual optical network mapping cost facing to a data center is characterized by comprising the following steps:
the method comprises the following steps: initializing a network;
step two: establishing an optimization method taking the lowest virtual network mapping cost as an objective function;
step three: establishing a constraint condition meeting the objective function optimization method;
step four: the number of optical regenerators is counted.
2. The method for resource allocation of virtual optical network mapping cost for data center according to claim 1, wherein: the step one is specifically
Given a physical network Gp(Vp,Ep,Cp,Fp) Wherein V ispRepresenting a group of data centers, EpRepresents a set of physical links, CpRepresenting the computing resources owned by each data center, FpRepresents the number of spectrum slots available on each link;
group of virtual optical networks Gv(Vv,Ev,Cv,Bv) Wherein V isv,Ev,CvAnd BvRespectively representing a set of virtual nodes, a set of virtual links, a set of node computing resource requirements and a set of link bandwidth requirements;
setting | Vp|、|Ep|、|Cp|、|FpThe number of the | respectively represents the number of data centers in the physical optical network, the number of optical fiber links, the number of available computing resources of each data center and the number of spectrum gaps of each fiber core;
while setting the number of cores per fiber.
3. The method for resource allocation of virtual optical network mapping cost for data center according to claim 1, wherein: the second step is specifically that
For the cost problem of virtual optical network mapping in the space division multiplexing spectrum flexible optical network facing the data center, an optimization objective function is expressed by the following sub-formula:
Figure FDA0002874104280000021
since the mapping cost of the virtual optical network is only related to the number of optical repeaters and optical regenerators, the objective of equation (2) is to use the minimum number of optical repeaters and optical regenerators while satisfying the relevant constraints of the virtual optical network mapping.
4. The method for resource allocation of virtual optical network mapping cost for data center according to claim 1, wherein: the third step is specifically that
When the virtual optical network is mapped, the nodes and the data center need to satisfy the following constraint conditions:
(1) optical switching node constraints;
(2) the data center calculates resource capacity constraints;
(3) a flow conservation constraint;
(4) a routing constraint;
(5) a bandwidth capacity constraint;
(6) a spectral continuity constraint;
(7) spectrum consistency constraint;
(8) the core is uniquely constrained.
5. The method for resource allocation of virtual optical network mapping cost for data center according to claim 4, wherein: the optical switching node has the constraint condition of
Figure FDA0002874104280000022
Figure FDA0002874104280000023
Wherein,
Figure FDA0002874104280000031
the value is a binary variable, if the virtual node i is successfully mapped to the optical switching node k, the value is 1, otherwise, the value is 0;
the data center computing resource capacity is constrained to
Figure FDA0002874104280000032
Wherein,
Figure FDA0002874104280000033
representing the computational resource requirement of virtual node i on the nth virtual optical network, CkRepresenting the amount of total computational resources of data center k on a data center oriented space division multiplexed spectrum flexible optical network.
6. The method for resource allocation of virtual optical network mapping cost for data center according to claim 4, wherein: the conservation of flow is constrained to
Figure FDA0002874104280000034
Figure FDA0002874104280000035
Wherein,
Figure FDA0002874104280000036
is a binary variable, if the virtual link (i, j) on the nth virtual optical network occupies the f slot of the c core of the physical link (k, l), it is 1, otherwise it is 0,
Figure FDA0002874104280000037
the number of spectral slots required for the virtual link (i, j) is indicated.
7. The method for resource allocation of virtual optical network mapping cost for data center according to claim 4, wherein: the routing constraint is
Figure FDA0002874104280000038
Wherein,
Figure FDA0002874104280000039
is a binary variable, if the virtual link (i, j) on the nth virtual optical network is successfully mapped to the data center pair (k, l), it is 1, otherwise it is 0;
the bandwidth capacity is constrained to
Figure FDA0002874104280000041
If is the number of spectral slots on each physical link.
8. The method for resource allocation of virtual optical network mapping cost for data center according to claim 4, wherein: the spectral continuity constraint is
Figure FDA0002874104280000042
Figure FDA0002874104280000043
Figure FDA0002874104280000044
The spectrum resources allocated on the physical path for a virtual link must be contiguous, in equation (9), if
Figure FDA0002874104280000045
And is
Figure FDA0002874104280000046
All spectral gaps with index values higher than f +1 are not used for mapping virtual links on node pair (k, l); in the formula(10) In, if
Figure FDA0002874104280000047
Then no spectrum slots with index values below f will be allocated to virtual link (i, j).
To reduce the complexity of the calculation, it is necessary to reduce the binary variables
Figure FDA0002874104280000048
Of the dimension space of (1), thus introducing additional binary variables
Figure FDA0002874104280000049
And
Figure FDA00028741042800000410
to decompose
Figure FDA00028741042800000411
Wherein
Figure FDA00028741042800000412
Whether the virtual link (i, j) on the nth virtual optical network occupies the physical link (k, l) or not is represented;
Figure FDA00028741042800000413
representing whether the fiber core c is occupied by a virtual link (i, j) on the nth virtual optical network;
Figure FDA00028741042800000414
representing whether the virtual link (i, j) on the nth virtual optical network occupies a spectrum gap with index f; binary variable
Figure FDA00028741042800000415
And
Figure FDA00028741042800000416
the relationship between them is:
Figure FDA0002874104280000051
Figure FDA0002874104280000052
Figure FDA0002874104280000053
9. the method for resource allocation of virtual optical network mapping cost for data center according to claim 4, wherein: the spectrum consistency constraint is
Figure FDA0002874104280000054
The core is uniquely constrained to
Figure FDA0002874104280000055
10. The method for resource allocation of virtual optical network mapping cost for data center according to claim 1, wherein: the fourth step is specifically that
The cost is mainly related to the number of optical repeaters and optical regenerators, which is fixed since the virtual optical network is given and uses a single line rate, where the goal is to optimize the number of optical regenerators; after a virtual optical network has been successfully mapped, the number of optical regenerators required can be derived from equation (16), where R represents the maximum reachable distance of the optical regenerator at the selected line rate and modulation format, D(k,l)Refers to the distance of the physical link (k, l);
Figure FDA0002874104280000056
the method for allocating the resources of the route, the spectrum and the fiber core when the virtual optical network is mapped in the space division multiplexing spectrum flexible optical network facing the data center can be found out through the constraint conditions and the calculation formula of the optical regenerator.
CN202011621746.6A 2020-12-30 2020-12-30 Resource allocation method for virtual optical network mapping cost facing data center Active CN112822038B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011621746.6A CN112822038B (en) 2020-12-30 2020-12-30 Resource allocation method for virtual optical network mapping cost facing data center

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011621746.6A CN112822038B (en) 2020-12-30 2020-12-30 Resource allocation method for virtual optical network mapping cost facing data center

Publications (2)

Publication Number Publication Date
CN112822038A true CN112822038A (en) 2021-05-18
CN112822038B CN112822038B (en) 2023-02-17

Family

ID=75856061

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011621746.6A Active CN112822038B (en) 2020-12-30 2020-12-30 Resource allocation method for virtual optical network mapping cost facing data center

Country Status (1)

Country Link
CN (1) CN112822038B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114614899A (en) * 2022-03-29 2022-06-10 中天宽带技术有限公司 Data center virtual optical network mapping method and device and electronic equipment

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110636394A (en) * 2019-10-17 2019-12-31 中天通信技术有限公司 Virtual optical network mapping method, device, equipment and medium

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110636394A (en) * 2019-10-17 2019-12-31 中天通信技术有限公司 Virtual optical network mapping method, device, equipment and medium

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
BOWEN CHEN等: "《Cost-Effective Survivable Virtual Optical Network Mapping in Flexible Bandwidth Optical Networks》", 《JOURNAL OF LIGHTWAVE TECHNOLOGY》 *
BOWEN CHEN等: "《Minimum-cost survivable virtual optical network mapping in flexible bandwidth optical networks》", 《2014 IEEE GLOBAL COMMUNICATIONS CONFERENCE》 *
QI CHEN等: "《A Profit-Maximized Approach Based on Link Importance Degree for Virtual Optical Networks Mapping》", 《2019 ASIA COMMUNICATIONS AND PHOTONICS CONFERENCE (ACP)》 *
QI CHEN等: "《Virtual Optical Network Mapping Approaches in Space-Division-Multiplexing Elastic Optical Data Center Networks》", 《JOURNAL OF LIGHTWAVE TECHNOLOGY》 *
陈伯文等: "《在SDON中面向成本优化的虚拟光网络映射方法》", 《中兴通讯技术》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114614899A (en) * 2022-03-29 2022-06-10 中天宽带技术有限公司 Data center virtual optical network mapping method and device and electronic equipment
CN114614899B (en) * 2022-03-29 2024-06-11 中天宽带技术有限公司 Data center virtual optical network mapping method and device and electronic equipment

Also Published As

Publication number Publication date
CN112822038B (en) 2023-02-17

Similar Documents

Publication Publication Date Title
CN110035336B (en) Routing fiber core frequency spectrum allocation method of space division multiplexing elastic optical network
CN108834004B (en) Routing calculation, fiber core selection and frequency spectrum allocation method and system based on cross crosstalk perception
CN105141517B (en) A kind of flexible optical-fiber network Anycast business energy-saving routing method based on resource perception
WO2021088230A1 (en) Resource allocation method in multi-core optical fiber network
WO2013037570A1 (en) Allocation of spectral capacity in a wavelength-division multiplexing optical network
CN108199881B (en) Ultra-low loss optical fiber replacement scheduling method and system in backbone network
CN113099328B (en) Resource allocation method of multi-core elastic optical network based on node and crosstalk perception
CN108377220B (en) Transparent virtual optical network collaborative mapping method for sensing node importance
Beyragh et al. IF-RSCA: intelligent fragmentation-aware method for routing, spectrum and core assignment in space division multiplexing elastic optical networks (SDM-EON)
Yuan et al. A RMSA algorithm for elastic optical network with a tradeoff between consumed resources and distance to boundary
CN113489617B (en) Minimum network energy consumption optimization method and system based on traffic grooming
CN111866623A (en) High-efficiency virtual optical network survivability mapping method for service reliability
CN112822038B (en) Resource allocation method for virtual optical network mapping cost facing data center
CN111010627B (en) Route distribution method and device applied to mixed grid optical network
WO2019008718A1 (en) Optical path control device and optical path control method
CN113015040B (en) Resource allocation method based on fragment and domain matching degree in multi-core elastic optical network
CN112995805B (en) Routing and spectrum allocation method based on path idleness in EON
CN111385680B (en) Spectrum allocation method based on mixed spectrum conversion resource pool in elastic optical network
Liu et al. Resource efficiency improved approach for shared path protection in EONs
CN117459859A (en) Multi-dimensional fragment-aware dynamic routing and resource allocation method
JP5759636B2 (en) Method for allocating bandwidth in an optical network
WO2023108716A1 (en) Data center resource awareness-oriented virtual network mapping method
Lv et al. Load-balanced adaptive routing flexible grouping spectrum and core assignment in SDM-EONs based on mixed super-channel
Lan et al. A fragmentation-aware load-balanced RMSCA algorithm in space-division multiplexing elastic optical networks
CN115209249B (en) Virtual network collaborative mapping method based on optical path adjacent link fragment awareness

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant