CN104320343A - Electric energy perception cross-domain virtual network mapping method and system - Google Patents
Electric energy perception cross-domain virtual network mapping method and system Download PDFInfo
- Publication number
- CN104320343A CN104320343A CN201410589835.5A CN201410589835A CN104320343A CN 104320343 A CN104320343 A CN 104320343A CN 201410589835 A CN201410589835 A CN 201410589835A CN 104320343 A CN104320343 A CN 104320343A
- Authority
- CN
- China
- Prior art keywords
- physical
- node
- virtual
- mapping
- msub
- 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.)
- Pending
Links
- 238000013507 mapping Methods 0.000 title claims abstract description 200
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000008447 perception Effects 0.000 title abstract description 10
- 238000009826 distribution Methods 0.000 claims abstract description 30
- 238000005265 energy consumption Methods 0.000 claims description 35
- 238000004422 calculation algorithm Methods 0.000 claims description 26
- 230000005611 electricity Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000003860 storage Methods 0.000 claims description 7
- 235000008694 Humulus lupulus Nutrition 0.000 claims description 6
- 230000001174 ascending effect Effects 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims 1
- 230000007774 longterm Effects 0.000 abstract description 18
- 230000008901 benefit Effects 0.000 description 10
- 239000004793 Polystyrene Substances 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 238000006243 chemical reaction Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 241000721662 Juniperus Species 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920002223 polystyrene Polymers 0.000 description 1
- 238000005295 random walk Methods 0.000 description 1
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides an electric energy perception cross-domain virtual network mapping method and system. The method includes the steps of calculating electric energy expenses, produced for mapping virtual routing nodes, of all physical routing nodes, obtaining the NR priority of each physical routing node through the combination of the degree and bandwidth resources of the physical routing node, mapping the virtual routing nodes to the physical routing node with the highest NR priority, calculating average distribution values of resource utilization rates of all physical host nodes in a physical host node set, and mapping virtual host nodes to the physical host node with the smallest average distribution value of the resource utilization rates. By means of the method and system, the electric energy expenses of a physical network can be reduced in the cross-domain virtual network mapping process while the physical network long-term operation incomes are kept.
Description
Technical Field
The invention relates to the technical field of network virtualization in the field of computer networks, in particular to a cross-domain virtual network mapping method and a system thereof for power perception.
Background
The Internet has become an important information infrastructure supporting the economic development and technological innovation of modern society. However, it still faces serious technical challenges in terms of security, mobility, and quality of service. Due to the great success of the Internet, researchers often address this challenge by means of patching. This concept severely hampers the creation, deployment, and evaluation of innovative network architectures and technologies, making it difficult to fundamentally solve the problems inherent in the current internet itself.
In recent years, network virtualization technology has received much attention from the industry and academia. Network virtualization technology allows multiple heterogeneous virtual networks to coexist on top of a shared physical network infrastructure. Each virtual network is a slice of resources of a physical network, which is made up of virtual nodes (e.g., virtual routers) and virtual links. Each virtual network may run a personalized routing protocol in a specified topology. Therefore, the network virtualization technology makes it possible to deploy new network architectures, protocols, and applications without affecting existing networks, thereby effectively supporting innovations in network technologies. The method not only provides a feasible path for the evolution from the current Internet to the future network, but also is one of the key characteristics which the future Internet should have.
In a network virtualization environment, an infrastructure provider manages an operating physical network, and a service provider makes a request to the infrastructure provider to lease a virtual network. The problem of mapping virtual network requests with node and link resource constraints onto a physical network is referred to as the virtual network mapping problem. As one of the key problems in the field of network virtualization, the problem has received widespread attention from both academic and industrial circles. The existing achievements mainly focus on how to design an efficient virtual network mapping algorithm to map more virtual network requests, thereby improving the revenue of infrastructure providers and neglecting the energy consumption problem of physical networks. However, statistically, the proportion of energy consumption overhead in the operating costs of network operators has reached 50%. Therefore, it is of great significance for infrastructure providers to research how to reduce the energy consumption of physical networks while maintaining the operational benefits of the physical networks. In addition, the research result also meets the strategic requirements of national green economy, and has wide application prospect.
Disclosure of Invention
Features and advantages of the invention will be set forth in part in the description which follows, or may be obvious from the description, or may be learned by practice of the invention.
In order to overcome the problems of the prior art, the invention provides a power-aware cross-domain virtual network mapping method and a power-aware cross-domain virtual network mapping system, which can reduce the power overhead of a physical network while maintaining the long-term operation benefit of the physical network in the cross-domain virtual network mapping process.
The technical scheme adopted by the invention for solving the technical problems is as follows:
according to one aspect of the present invention, there is provided a power-aware cross-domain virtual network mapping method, including mapping a virtual node onto a physical node and mapping a virtual link onto a physical link, wherein:
when the virtual node is mapped to the physical node, respectively mapping a virtual routing node and a virtual host node in the virtual node to a corresponding physical routing node and a corresponding physical host node in the physical node; wherein:
when mapping the virtual routing node to the corresponding physical routing node, the method comprises the following steps:
a1, constructing a physical routing node list, wherein each physical routing node in the physical routing node list can meet the node requirement of the virtual routing node;
a2, calculating the electric energy cost generated by mapping the virtual routing node for each physical routing node, and obtaining the NR priority of each physical routing node according to the degree and bandwidth resource of each physical routing node;
a3, mapping the virtual routing node to the physical routing node with the highest NR priority;
when mapping the virtual host node to the corresponding physical host node, the method comprises the following steps:
b1, searching a physical host node set connected with the physical routing node mapped by all the virtual routing nodes;
b2, calculating the average distribution value of the resource utilization rate of each physical host node in the physical host node set;
b3, mapping the virtual host node to the physical host node with the smallest average distribution value of the resource utilization rate.
According to an embodiment of the present invention, the step a2 specifically includes:
when mapping virtual routing node r, using a formulaAnd <math>
<mrow>
<mi>Δ</mi>
<msub>
<mi>PN</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfenced open='{' close='}'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>P</mi>
<mi>fix</mi>
</msub>
<mo>+</mo>
<mi>ΔL</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>lc</mi>
</msub>
<mo>+</mo>
<mi>ΔP</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>port</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>if</mi>
<mo>,</mo>
<msub>
<mi>PS</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mi>ΔL</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>lc</mi>
</msub>
<mo>+</mo>
<mi>ΔP</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>port</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>otherwise</mi>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
</math> calculating the electric energy cost generated by mapping the virtual routing node for each physical routing node i; wherein Pri(t) represents the electricity price of the area where the physical routing node i is located in the current time window t, PfixRepresenting fixed energy consumption, P, within chassis, etc., of a physical routing nodelcIndicating energy consumption of line cards, PportRepresenting the energy consumption of each port, L and P respectively representing the number of line cards and ports, PSi0 represents that the physical routing node is in an inactive state, and the other represents that the physical routing node is in an active state;
sorting all physical routing nodes in the physical routing node list in ascending order according to the corresponding electric energy cost, and using NRE(i) Indicating the sequence number of the physical routing node i in the ascending order;
calculating Noderank values of all physical routing nodes in the physical routing node list by using a Noderank node measurement method, wherein the Noderank values are used for representing the degree and bandwidth resources of each physical node, sequencing the physical routing nodes from large to small according to the Noderank values, and using NRR(i) Representing the sequence number of the physical routing node i in the sequence;
using the formula NR ═ α NRE(i)+(1-α)NRR(i) Obtaining the NR priority of each physical routing node, wherein alpha satisfies the condition of 0 ≦α≤1。
According to an embodiment of the present invention, when mapping the virtual routing node to the corresponding physical routing node, the method further includes step a 4: if the mapping is successful in step a3, the physical routing node with the highest NR priority to which the virtual routing node is mapped is marked, and the physical routing node list does not include the marked physical routing node.
In this step B2, according to one embodiment of the present invention, a formula is utilized <math>
<mrow>
<msub>
<mi>δ</mi>
<mi>j</mi>
</msub>
<mo>=</mo>
<msqrt>
<mfrac>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>C</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>M</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>D</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>B</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
<mn>4</mn>
</mfrac>
</msqrt>
</mrow>
</math> Calculating the average distribution value of the resource utilization rate of each physical host node j in the physical host node set; wherein, Cj、MjAnd SjRespectively representing available CPU resources, available memory resources and available hard disk storage resources after the virtual host node is pre-mapped to j; b isjRepresents the link l to which the virtual host node is connectedhPre-mapping to j connected link ljThe next available bandwidth resource; u. ofjRepresents the average resource utilization, u, of the physical host node jj＝(Cj+Mj+Sj+Bj)/4. According to an embodiment of the present invention, when mapping the virtual home node to the corresponding physical home node, the method further includes step B4: if the mapping is successful in step B3, the virtual host node is mapped to the physical host node with the average distribution value of the minimum resource utilization, and the physical host node set in step B1 does not include the physical host node that has been marked.
According to an embodiment of the present invention, when mapping the virtual link onto the physical link, the method comprises the steps of:
c1, calculating the distance between every two physical routing nodes in advance according to a Flouard algorithm;
c2, searching a set of all physical links with a distance length within [ LEN, MAXLEN ] capable of meeting the bandwidth requirement of the virtual link from two physical host nodes, wherein LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is the maximum number of hops that can be accepted by a physical loop-free path to which the virtual link is mapped;
c3, calculating the weighted value of all physical links in the set by weighting the number of unopened nodes and ports;
c4, mapping the virtual link to the path with the least good and bad weight value.
According to an embodiment of the present invention, in the step C3, the formula L ═ β N is usedf+(1-β)NpCalculating the quality weighted value of all physical links in the set; wherein N isfAnd NpThe number of the unopened forwarding nodes and the number of the unopened ports on the current evaluation path are respectively represented, and beta represents the proportion of the energy consumption of one unopened forwarding node and the energy consumption of one unopened port.
According to another aspect of the present invention, there is provided a power-aware cross-domain virtual network mapping system, including a node mapping unit that maps virtual nodes onto physical nodes and a link mapping unit that maps virtual links onto physical links, wherein:
the node mapping unit is composed of a routing node mapping subunit and a host node mapping subunit, and is respectively used for mapping the virtual routing node to the corresponding physical routing node and mapping the virtual host node to the corresponding physical host node; wherein:
the routing node mapping subunit includes:
the list construction module is used for constructing a physical routing node list, and each physical routing node in the physical routing node list can meet the node requirement of the virtual routing node;
the priority calculation module is used for calculating the electric energy cost generated for mapping the virtual routing node for each physical node, and obtaining the NR priority of each physical routing node by combining the degree of each physical node and the bandwidth resource;
a routing node mapping module for mapping the virtual routing node to the physical routing node with the highest NR priority;
the host node mapping subunit includes:
a host node set searching module, configured to search a set of physical host nodes connected to the physical routing node to which all the virtual routing nodes are mapped;
the resource utilization rate average distribution value calculation module is used for calculating the resource utilization rate average distribution value of each physical host node in the physical host node set;
and the host node mapping module is used for mapping the virtual host node to the physical host node with the average distribution value of the minimum resource utilization rate.
According to an embodiment of the present invention, the node mapping unit further includes a marking module, configured to mark the physical routing node with the highest NR priority to which the virtual routing node is successfully mapped, or mark the physical hosting node with the average distribution value with the smallest resource utilization to which the virtual hosting node is successfully mapped.
According to an embodiment of the present invention, the link mapping unit includes:
the distance calculation module is used for calculating the distance between every two physical routing nodes in advance according to a Flouard algorithm;
a link set searching module, configured to search, from between two physical host nodes, a set of all physical links having a distance length within [ LEN, MAXLEN ], where LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is a maximum number of hops that can be accepted by a physical loop-free path to which a virtual link is mapped;
the quality weighted value calculation module is used for calculating the quality weighted values of all the physical links in the set by using a method of weighting the number of unopened nodes and ports;
and the link mapping module is used for mapping the virtual link to the path with the minimum quality weighting value.
The features and content of these solutions will be better understood by those skilled in the art from reading the present description.
Drawings
The advantages and realisation of the invention will be more apparent from the following detailed description, given by way of example, with reference to the accompanying drawings, which are given for the purpose of illustration only, and which are not to be construed in any way as limiting the invention, and in which:
fig. 1 is a schematic flowchart of node mapping in the power-aware cross-domain virtual network mapping method according to the embodiment of the present invention.
Fig. 2 is a schematic flow chart of link mapping in the power-aware cross-domain virtual network mapping method according to the embodiment of the present invention.
Fig. 3 is a schematic structural diagram of a power-aware cross-domain virtual network mapping system according to an embodiment of the present invention.
Fig. 4 is a statistical chart of long-term average operation revenue of a physical network according to an embodiment of the present invention.
Fig. 5 is a statistical chart of the physical network power overhead according to the embodiment of the present invention.
Fig. 6 is a statistical diagram of the number of opened nodes of the physical network according to the embodiment of the present invention.
Fig. 7 is a schematic diagram of cross-domain virtual network mapping according to an embodiment of the present invention.
Detailed Description
As shown in fig. 1, the present invention provides a power-aware cross-domain virtual network mapping method, which includes mapping a virtual node to a physical node and mapping a virtual link to a physical link, and is characterized in that:
when mapping a virtual node to a physical node, mapping a virtual routing node and a virtual host node in the virtual node to a corresponding physical routing node and a corresponding physical host node in the physical node respectively; wherein:
when mapping the virtual routing node to the corresponding physical routing node, the method comprises the following steps:
a1, constructing a physical routing node list, wherein each physical routing node in the physical routing node list can meet the node requirement of the virtual routing node;
a2, calculating the electric energy cost generated by mapping the virtual routing node for each physical routing node, and obtaining the NR priority of each physical routing node according to the degree and bandwidth resource of each physical routing node;
a3, mapping the virtual routing node to the physical routing node with the highest NR priority;
when mapping the virtual host node to the corresponding physical host node, the method comprises the following steps:
b1, searching a physical host node set connected with the physical routing node mapped by all the virtual routing nodes;
b2, calculating the average distribution value of the resource utilization rate of each physical host node in the physical host node set;
b3, mapping the virtual home node to the physical home node with the smallest average distribution value of the resource utilization.
It should be noted that, in the physical network, the topology of the physical network may be marked as a weighted undirected graph Gs＝(Ns,Ls) In which N issRepresenting the set of all physical nodes of a physical network, LsRepresenting a collection of physical network physical links. In the present invention, physical nodes are further divided into two categories: physical routing nodes and physical host nodes, i.e. Ns＝(Nsr,Nsh). Wherein N issrRepresenting the set of all physical routing nodes in a physical network, NshRepresenting the set of all physical host nodes in the physical network.
For a routing node, the present invention considers the following three node attributes:
the first node attribute is the maximum number of virtual routers which can be supported by each routing node and used for deploying and running the personalized network protocol;
the second routing node attribute is the position attribute of the routing node, in the deployment environment in the actual environment, the routing node is often distributed in a plurality of geographic areas, and for the routing node i, the invention uses a two-dimensional coordinate loc (i) ═ xi,yi) Indicating the position of the node i;
the third routing node attribute is an electricity price attribute, and in the north american electric energy market, electric energy markets in different regions are operated by different Regional Transmission Organizations (RTOs) or Independent electric energy System operators (ISO). For example, New York-ISO is responsible for managing the operation of the electric energy supply system in New York State. Due to differences in the cost of production materials for the regions for which different RTOs are responsible, electricity prices can exhibit regional differences. Furthermore, in a region, such as the houston region of texas, a user with a demand for electrical energy may be free to choose between multiple electrical energy suppliers. Therefore, the electricity prices in the area also exhibit time-varying characteristics. In order to describe regional difference and time-varying difference of electrovalence, the invention utilizes Pri(t) represents the electricity price of the area where the routing node i is located in the current time window t, and t is usedaIndicating the arrival time of the virtual network request, denoted tdRepresenting the duration of the virtual network request in the physical network, the termination time of the virtual network request is te＝ta+td。
For virtual networks, a weighted undirected graph G may also be usedv＝(Nv,Lv) Is represented by the formula, wherein NvRepresenting a set of virtual nodes, LvRepresenting a set of virtual links. Also, in the present invention, virtual nodes are further divided into two categories: virtual routing nodes and virtual host nodes, i.e. Nv＝(Nvr,Nvh)，NvrAnd NvhRepresenting a set of virtual routing nodes and virtual host nodes, respectively.
For a given virtual network GvAnd physical network GsPerforming virtual network mapping M to Gv→GsTime, including node mapping MnAnd a link map Ml。
It can be seen that MnRefers to the slave set NvTo set NsThe mapping may be further divided into two sub-mappings: mnrAnd Mnh(ii) a Wherein M isnrRefers to the slave set NvrTo set NsrOne-to-one mapping of subsets, MnhRepresenting a slave set NvhTo set NshOne-to-one mapping of subsets, that is, Mnr:Nvr→NsrMapping means mapping a virtual routing node to a corresponding physical routing node, and Mnh:Nvh→NshIt means that the virtual home node is mapped onto the corresponding physical home node.
At Mnr:Nvr→NsrIn the mapping, i.e. for steps A1 to A3, for any virtual routing node nvr∈NvrAnd physical routing node Mnr(nvr)∈Nsr，Mnr(nvr) Needs to satisfy nvrThe position of the router and the number of virtual router nodes.
At Mnh:Nvh→NshIn the mapping, i.e. for steps B1-B3, for any virtual host node nvh∈NvhAnd a physical host node Mnh(nvh)∈Nsh，Mnh(nvh) Need to satisfy node nvhCPU power, memory power, and hard disk storage power.
The invention also provides an electric energy overhead model of the physical node, the electric energy overhead of the physical node mainly comprises three parts of electric energy overhead of the physical routing node and the physical host node and converted electric energy overhead, and the electric energy overhead model specifically comprises the following steps:
(a) physical routing node power consumption
The energy consumption model of the physical routing node is as follows:
PNi＝Pfix+L·Plc+P·Pportequation (1)
Wherein, PfixRepresenting fixed energy consumption, P, of devices such as chassis of routing nodelcIndicating energy consumption of line cards, PportIndicating the energy consumption of each port, L and P indicating the number of line cards and ports, respectively.
Thus, to map virtual routing node r, the additional power consumed by physical routing node i can be calculated by:
Wherein PSiState representing physical routing node i: PS (polystyrene) with high sensitivityi0 indicates that the physical routing node is in an inactive state, and another indicates that the physical routing node is in an active state.
(b) Physical host node power consumption
Because the power consumption of the host node is almost linear with the load of the CPU, the power consumption including the memory and the hard disk is less changed. Thus, the present invention approximates the power consumption of the physical host node j by:
PNj＝Pb+Plu, formula (3)
Wherein, PbRepresents the power of the node at empty load, referred to as baseline power consumption; u denotes currentThe load of the CPU; plA linear parameter representing the power of the node as a function of u.
Based on the energy consumption model, virtual host node h (C for CPU demand) is mappedhRepresentation), the additional power consumption required by the physical host node j is:
Based on the energy consumption models of the physical routing nodes and the physical host nodes, the total node electric energy expenditure consumed by the physical network is as follows:
(c) Conversion energy consumption
In addition, in physical routing and host nodesIn the process of converting the point from an inactive state to an active state, a fixed amount of energy consumption is consumed, which is called conversion energy consumption in the invention. The conversion energy consumption of the physical routing node and the host node is respectively EsrAnd EshAnd (4) showing. The total conversion power overhead of the physical network is as follows:
The invention provides an electric energy perception cross-domain virtual network mapping method (CA-VNE for short), which is a two-stage virtual network mapping algorithm and comprises two stages of node mapping and link mapping, wherein a virtual node is required to be mapped to a physical node in the node mapping stage, and a virtual link is required to be mapped to a physical link in the link mapping stage.
In the node mapping stage, the mapping of the virtual routing nodes, i.e., steps a1 to A3, is performed first, and then the mapping of the virtual host nodes, i.e., steps B1 to B3, is performed.
When mapping the virtual routing node, the following two goals need to be achieved: 1) optimizing the electric energy expenditure of the virtual routing node; 2) because the mapping of the virtual routing node may affect the subsequent mapping of the virtual host node and the virtual link, at this stage, the subsequent stage still needs to be considered to improve the success probability of the virtual network request.
To achieve the first objective, the present invention proposes a ranking strategy for electricity price and energy consumption perception. The present invention assumes that, for the virtual network request to be mapped currently, the electricity price is known in advance from the current arrival time to the expiration time of the virtual network for the following reasons:
first, through various effective prediction techniques, such as a fuzzy inference system-based method or a similar day and artificial neural network-based technique, the RTO/ISO can predict electricity prices more accurately in the next day or even a week according to historical rules;
second, for virtual network services, such as Voice Over Internet Protocol (VOIP), online games, and Internet Protocol Television (IPTV) services, it often takes several minutes or hours, and the time length is far from one day or one week.
Based on the above assumptions, when mapping virtual routing node r, the formula is utilizedAnd the above equation (2) calculates the power cost generated for mapping the virtual routing node for each physical routing node i. Then, sorting all physical routing nodes in the physical routing node list in ascending order according to the corresponding electric energy cost, and using NRE(i) Indicating the sequence number of physical routing node i in the ascending order.
In order to achieve the second objective, the invention designs the worst suitable strategy, and preferentially selects the physical routing node with larger degree and rich bandwidth resources, so as to achieve the purpose of improving the successful mapping probability of subsequent host nodes and links. For how to evaluate the degree of a node and the abundance degree of bandwidth resources, the invention utilizes a Noderank node measurement method considering the self resources of the node and the surrounding neighbor node resources of the node based on the random walk theory. The Noderank value is calculated according to the above, and the physical routing nodes are sorted from large to small according to the Noderank value, and NR is usedR(i) Representing the sequence number of physical routing node i in the ordering. The Noderank node measuring method not only considers the degree and bandwidth resources of the physical routing node, but also considers the degree and bandwidth resources of the neighboring nodes, so the physical routing node with larger Noderank value is selected,the method is beneficial to improving the successful mapping probability in the following virtual host node mapping and virtual link mapping process, thereby being beneficial to improving the long-term operation income of the physical network.
In order to make a trade-off between the above two conflicting objectives of minimizing power overhead and maximizing revenue, the present invention utilizes the formula NR ═ α NRE(i)+(1-α)NRR(i) Obtaining the NR priority of each physical routing node; wherein alpha and 1-alpha respectively represent the proportion of the minimized electric energy expenditure and the maximized operation profit in the objective function (alpha is more than or equal to 0 and less than or equal to 1).
And according to the comprehensive sequencing, selecting the physical routing node with the highest comprehensive sequencing for each virtual routing node to map. The advantages of the comprehensive sorting method are as follows: firstly, the electric energy expenditure of an infrastructure provider can be reduced by using the regional difference and the time-varying difference of the electricity price and the starting state of the physical routing node; secondly, the difficulty of subsequent virtual host node and virtual link mapping is reduced, the successful probability of subsequent mapping is improved, and the long-term operation benefit of the physical network is improved.
In this embodiment, when mapping the virtual routing node to the corresponding physical routing node, the method further includes step a 4: if the mapping is successful in step A3, the physical routing node with the highest NR priority to which the virtual routing node is mapped is marked, and the marked physical routing node is not included in the physical routing node list in step a 1. In specific implementation, a step may be added after the step a1 and before the step a 2: and deleting the marked physical routing nodes in the physical routing node list.
After the mapping of the virtual routing node is completed, the mapping of the virtual host node is performed, where the virtual host node mapping refers to mapping a virtual host node with a node requirement, such as a CPU capacity requirement, a memory capacity requirement, or a hard disk storage capacity requirement, onto a physical host node. Since the mapping of a virtual host node may affect the mapping of subsequent virtual links, it is consideredWhile considering the above three node requirements, link bandwidth resources also need to be considered to improve the probability of success in the virtual link phase. Thus, the problem can be viewed approximately as a classical four-dimensional binning problem. In order to solve the problem, the invention designs the following optimal suitable strategy, namely mapping the virtual host node to the nodes of the four physical hosts with more evenly distributed resource utilization rates. Specifically, first, an average utilization u of four resources of each physical host node j is definedj，uj＝(Cj+Mj+Sj+Bj) /4, wherein Cj、MjAnd SjRespectively representing available CPU resources, available memory resources and available hard disk storage resources after the virtual host node h is pre-mapped to the physical host node j; b isjRepresenting the link l to which the virtual host node h is connectedhPre-mapping to j connected link ljThe available bandwidth resources thereafter. Then, the standard deviations of the four resources for each physical host node j are defined as follows:
the standard deviation of the four resources of the physical host node j is the average distribution value of the resource utilization rate, and finally, the virtual host node h is mapped to the physical host node with the minimum standard deviation, namely the physical host node with the average distribution value of the minimum resource utilization rate.
The best fit strategy has the following advantages: 1) the starting of a new host node is avoided, the baseline and the conversion energy consumption are avoided, and therefore the electric energy expenditure of a physical network is reduced; 2) the method is beneficial to reasonably and evenly utilizing the physical network resources, and avoids causing the shortage of certain one-dimensional resources, thereby improving the utility of the physical network.
In this embodiment, when mapping the virtual host node to the corresponding physical host node, the method further includes step B4: if the mapping is successful in step B3, the virtual host node is mapped to the physical host node with the average distribution value of the minimum resource utilization, and the physical host node set in step B1 does not include the physical host node that has been marked. In specific implementation, the following step B1 may be added before step B2: deleting the marked physical host nodes in the physical host node set; in step B3, it is also necessary to determine the average distribution value with the minimum resource utilization before mapping, and the physical host node is able to satisfy the requirement of the virtual host node.
The following describes the mapping method of the virtual link, specifically, the virtual link luvIs that at the physical host node huAnd hvFind a loop-free path between the virtual links l and the physical links l, wherein the available bandwidth resources of each physical link on the path need to satisfy the virtual link luvThe bandwidth requirement of (c). In this process, the following two issues need to be considered: 1) the problem of resource waste caused by mapping the virtual link to a long path of a physical network is avoided; 2) the starting states of the physical nodes and the ports of the physical network are considered, and the waste of baseline energy consumption and conversion energy consumption is avoided.
Based on the above analysis, when mapping the virtual link onto the physical link, the present invention designs a shortest path algorithm for sensing node and port states, please refer to fig. 2, which includes the following steps:
c1, calculating the distance between every two physical routing nodes in advance according to a Flouard algorithm;
c2, searching a set of all physical links with a distance length within [ LEN, MAXLEN ] capable of meeting the bandwidth requirement of the virtual link from two physical host nodes, wherein LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is the maximum number of hops that can be accepted by a physical loop-free path to which the virtual link is mapped;
c3, calculating the weighted value of all physical links in the set by weighting the number of unopened nodes and ports;
c4, mapping the virtual link to the path with the least good and bad weight value.
Step C3 includes: using the formula L ═ beta Nf+(1-β)NpCalculating the quality weighted value of all physical links in the set; wherein N isfAnd NpThe number of the unopened forwarding nodes and the number of the unopened ports on the current evaluation path are respectively represented, and beta represents the proportion of the energy consumption of one unopened forwarding node and the energy consumption of one unopened port. And in step C4, the path with the lowest priority weight is selected to reduce the energy consumption overhead of the physical network.
It should be noted that: the physical links of the physical network of the present invention can also be divided into two categories: physical backbone links and physical access links, i.e. Ls＝(Lsr,Lsh). Wherein L issrSet of backbone links, L, representing connections between all physical routing nodesshRepresents the set of access links connected between all physical routing nodes and the physical host node. Likewise, the virtual links of a virtual network can also be divided into two categories: virtual backbone links and virtual access links, i.e. Lv＝(Lvr,Lvh) Wherein L isvrRepresents a set of virtual links between virtual routing nodes, and LvhRepresents a set of virtual links between a virtual routing node and a virtual host node.
Performing virtual network mapping M to Gv→GsTemporal link mapping MlWhen M is in contact withlRepresenting a slave set LvTo the set PsIn which P issIs represented by all physical links LsA set of constituent loop-free paths. The single device also contains two child maps: mlrAnd MlhWherein M islr:Lvr→PsrSet L representing links between slave virtual routing nodesvrSet P of loop-free paths to all physical routing linkssrIs singly provided, and Mlh:Lvh→PshRepresenting data from all virtual routing nodes and virtualSet L of virtual links between pseudo-homed nodesvhTo all loop-free paths P consisting of physical links between all physical routing nodes and physical host nodesshIs provided separately. In the mapping Mlr:Lvr→PsrFor any virtual link lvr∈Lvr，Mlr(lvr)∈PsrAll physical links on a path need to satisfy lvrThe bandwidth requirement of (c). The same applies to Mlh:Lvh→Psh。
The invention also provides an electric energy overhead model of the physical link, and in the mapping process of the virtual link, as the physical routing nodes of the physical network are distributed in a plurality of regions, the electric energy overhead model is used as a communication facility for being responsible for the physical routing nodes distributed among different regions, in order to overcome the problems of long distance and poor communication signal quality, repeater equipment is required to be installed between backbone links at intervals so as to amplify the communication signals. For the inventionRepresented on the backbone long link lst∈LsrDue to bearer virtual link luv∈LvrThe power consumption of (2). According to the results of the related studies, it is shown that,and a virtual link luvAnd the communication distances of the physical routing nodes s and t are in a proportional relation, that is:
Wherein, PrRepresenting the energy consumption of the repeater per unit distance, BuvRepresenting a virtual link luvOf the flow rate ofstRepresents a physical link lstThe total flow rate.
Based on the above analysis, the total link power overhead consumed by the physical network is:
By combining the formula (5) and the formula (6) in the electric energy overhead model of the physical node, it can be obtained that the total electric energy overhead of the physical network is:
Δ E ═ Δ EN + Δ ES + Δ el, equation (9)
And the main evaluation indexes of the virtual network mapping comprise the long-term average operation income of the physical network and the long-term average electric energy expenditure of the physical network.
The long term average operating revenue of a physical network may be represented by the following equation:
Wherein, <math>
<mrow>
<msup>
<mi>R</mi>
<mi>i</mi>
</msup>
<mrow>
<mo>(</mo>
<msub>
<mi>G</mi>
<mi>v</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<munder>
<mi>Σ</mi>
<mrow>
<mi>h</mi>
<mo>∈</mo>
<msub>
<mi>N</mi>
<mi>vh</mi>
</msub>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>C</mi>
<mi>h</mi>
</msub>
<mo>+</mo>
<msub>
<mi>M</mi>
<mi>h</mi>
</msub>
<mo>+</mo>
<msub>
<mi>S</mi>
<mi>h</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>W</mi>
<mi>R</mi>
</msub>
<mo>·</mo>
<munder>
<mi>Σ</mi>
<mrow>
<mi>r</mi>
<mo>∈</mo>
<msub>
<mi>N</mi>
<mi>vr</mi>
</msub>
</mrow>
</munder>
<msub>
<mi>R</mi>
<mi>r</mi>
</msub>
<mo>+</mo>
<munder>
<mi>Σ</mi>
<mrow>
<msub>
<mi>l</mi>
<mi>v</mi>
</msub>
<mo>∈</mo>
<msub>
<mi>L</mi>
<mi>v</mi>
</msub>
</mrow>
</munder>
<mi>B</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>l</mi>
<mi>v</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
<mo>·</mo>
<msub>
<mi>t</mi>
<mi>d</mi>
</msub>
<mo>,</mo>
</mrow>
</math> indicating the revenue generated by the infrastructure provider mapping the ith virtual network request.
Accordingly, the long-term average power cost of a physical network may be defined as:
Wherein E isi(Gv) The total power cost generated by the infrastructure provider for mapping the ith virtual network is shown in equation 9.
Referring to fig. 3, the present invention provides a power-aware cross-domain virtual network mapping system for implementing the power-aware cross-domain virtual network mapping method, including a node mapping unit 10 that maps virtual nodes onto physical nodes and a link mapping unit 90 that maps virtual links onto physical links, where:
the node mapping unit 10 is composed of a routing node mapping subunit 20 and a host node mapping subunit 30, and is respectively configured to map the virtual routing node to the corresponding physical routing node, and map the virtual host node to the corresponding physical host node; wherein:
the routing node mapping subunit 20 includes: a list construction module 21, configured to construct a physical routing node list, where each physical routing node in the physical routing node list can meet a node requirement of the virtual routing node; a priority calculation module 22, configured to calculate, for each physical node, an electric energy cost generated for mapping the virtual routing node, and obtain, by combining the degree of each physical node and the bandwidth resource, an NR priority of each physical routing node; a routing node mapping module 23, configured to map the virtual routing node to the physical routing node with the highest NR priority;
the host node mapping subunit 30 includes: a host node set searching module 31, configured to search a physical host node set connected to the physical routing node to which all virtual routing nodes are mapped; an average distribution value calculation module 32 of resource utilization rate, configured to calculate an average distribution value of resource utilization rate for each physical host node in the physical host node set; a host node mapping module 33, configured to map the virtual host node to the physical host node with the average distribution value with the minimum resource utilization.
The link mapping unit 90 includes: the distance calculation module 91 is configured to calculate a distance between every two physical routing nodes in advance according to a freouard algorithm; a link set searching module 92, configured to search, from between two physical host nodes, a set of all physical links having a distance length within [ LEN, MAXLEN ], where LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is a maximum number of hops that can be accepted by a physical loop-free path to which a virtual link is mapped; a good and bad weighted value calculating module 93, configured to calculate the good and bad weighted values of all physical links in the set by using a method of weighting the number of unopened nodes and ports; a link mapping module 94 for mapping the virtual link to the path with the smallest good or bad weight value.
In this embodiment, the node mapping unit 10 further includes a marking module 40, configured to mark the physical routing node with the highest NR priority to which the virtual routing node is successfully mapped, or mark the physical hosting node with the average distribution value with the smallest resource utilization to which the virtual hosting node is successfully mapped.
Compared with the latest virtual network mapping algorithm, namely the D-VinE-SP algorithm, the method provided by the invention can be used for remarkably reducing the long-term average energy consumption overhead (formula 11) and the time overhead of the physical network while maintaining the long-term average operation income (formula 10) of the physical network. The experimental setup, including physical network topology setup, virtual network topology setup and other parameter setup, is first described below, and then the experimental results are given and analyzed.
The invention uses the GT-ITM tool to generate physical and virtual network topologies in terms of physical network topology set up to contain 500 physical routing nodes. These routing nodes are evenly distributed in a two-dimensional plane of size (25 × 25). The two-dimensional plane can be equally divided into 5 regions, representing 5 electric energy markets respectively. For these electric energy markets, the invention collects real electricity price data of 5 RTOs in 9 months in 2011. For each electric energy market, the electricity price changes once per hour, and the invention regards each hour as a time window. Based on the typical setup of Juniper routers, the number of the largest virtual routing nodes supported by each physical routing node is evenly distributed in the interval [3,128 ]. Each access routing node may access unequal numbers of home nodes, 32, 64, or 128, depending on the connectivity capabilities of the current industry router. In consideration of the heterogeneity of the physical network, available CPU power, available memory power and available hard disk storage power are uniformly distributed within the interval of [50,100] for each host node. All physical routes and the host node are initially in a sleep state. Regarding the setting of the underlying links, in the physical network, the probability of every two physical routing nodes being connected is 0.5. And the intra-domain link bandwidth capacity between the physical routing nodes and the host node is uniformly distributed in a [50,100] interval, and the backbone link bandwidth between the physical routing nodes is 100 times of the intra-domain bandwidth capacity.
In the aspect of virtual network topology setting, the number of virtual routing nodes requested by each virtual network is uniformly distributed between [10 and 50], and the probability that backbone links are connected between every two virtual routing nodes is 0.5. Furthermore, each access routing node may access an unequal 8, 16 or 24 host node. The QoS constraints (i.e., CPU power requirements, memory power requirements, and hard disk storage power requirements) for all nodes are evenly distributed across 0, 50. The bandwidth requirements of the links in the domain between the virtual routing nodes and the host node are uniformly distributed between [0 and 50], and the bandwidth requirement of the backbone links between the two virtual routing nodes is 100 times of the bandwidth requirement of the links in the domain. In each virtual network mapping instance, 2000 virtual network requests are included, the arrival of each virtual network request follows a poisson distribution with an average of 3 virtual network requests per time window, and the duration of each virtual network request in the physical network follows a negative exponential distribution with an average of 10 time windows. The invention runs 10 instances of virtual network mapping in each set of evaluations and then calculates the average of 10 instances as the final result.
In other parameter setting aspects, the present invention sets the position parameter LP (LP is the ratio of W to the size of the side of the two-dimensional plane) to 1/3. According to the research related to the industry and the academic community, the invention uses P in the formula 1fix、PlcAnd PportSet to 375W, 315W and 3W, respectively, P in equation 3bAnd PlSet to 165W and 1.5W/CPU units, respectively, P in equation 7rSet to 100W/distance unit, E in equation 6srAnd EshEnergy consumption set to one router and one host, respectively, at full load in 1 hour, W in equation 10RSet to 100. In the CA-VNE algorithm, the formula NR ═ α NRE(i)+(1-α)NRR(i) The weight α in (a) is set to 0.5 by default, and MAXLEN in step C2 is set to 8.
The virtual network mapping performance of the CA-VNE is illustrated by analyzing experimental data below.
As can be seen from fig. 4, the CA-VNE algorithm achieves a slightly higher long-term average operational gain of the physical network compared to the D-ViNE-SP algorithm. The reason is the following two points: 1) in the node mapping stage, the CA-VNE algorithm considers the link constraint condition in advance, thereby reducing the difficulty of link mapping and improving the success probability of virtual network mapping; 2) based on the technology of virtual network integration, namely, the virtual network request is integrated to the opened physical routing node and the physical host node as much as possible, so that physical network resources can be saved, and more resource space is reserved for future virtual network requests.
It can be seen from fig. 5 that the CA-VNE algorithm greatly reduces the physical network long term average power overhead compared to the D-ViNE-SP algorithm. For example, at the 720 th time window, the power cost of the CA-VNE is reduced by up to 40% compared to the D-VinE-SP. The reason is that the CA-VNE algorithm not only considers electricity prices but also performs the integration of the virtual network in the node and link mapping phase.
As can be seen from fig. 6, the CA-VNE algorithm reduces the on-nodes of the physical network by 30% compared to the D-ViNE-SP algorithm. The two aspects enable the CA-VNE algorithm to greatly reduce the electric energy expenditure of a physical network.
In addition, as shown in fig. 4, 5, and 6, the present invention compares the CA-VNE algorithm (a is dynamically adjusted according to physical network load) with other variant algorithms (i.e., a is set to constants 0.1, 0.5, and 0.9, etc.). In the CA-VNE algorithm, the invention sets alpha as the ratio of the currently available resources of the physical network to the total resources of the physical network. The strategy has the advantage that the two goals of maximizing the benefit and minimizing the electric energy cost can be effectively balanced in a self-adaptive mode according to the load condition of the physical network. It can be seen from fig. 4 and 5 that as α increases, the physical network long-term average power overhead decreases, but the physical network long-term average operating revenue also decreases. The alpha value self-adaptively adjusted according to the network load can effectively reduce the long-term average electric energy expenditure of the physical network when the physical network resource load is light, and effectively improve the long-term average operation income of the physical network when the physical network resource load is high. Thus, the CA-VNE algorithm achieves almost the best performance between the two criteria mentioned above, compared to other variants where a is set to a constant value.
The invention can be applied to the backbone network or data center network environment supporting the network virtualization technology, and provides higher economic benefit for infrastructure providers by reducing the energy consumption expense of the physical network. The CA-VNE algorithm takes topological structure and resource capacity conditions requested by a physical network and a virtual network as input, and takes a better electric energy perception virtual network mapping scheme as output.
Referring to fig. 7, fig. 7 is a schematic diagram of a cross-domain virtual network map according to an embodiment of the present invention, and fig. 7(a) shows a virtual network map, in which circles represent routing nodes and rectangles represent host nodes. FIG. 7(b) shows a cross-regional physical network, e.g., it is possible to assign RADeployed in los Angeles, RBDeployed in Chicago, RCDeployed in New York, RDDeployed in new jersey.
In this embodiment, when mapping a cross-domain virtual network, the prior art virtual network node mapping scheme is { { Ra→RA,Rb→RC},{Ha1→HA1,Ha2→HA2,Hb1→HC1}, the link mapping scheme is a { (cushion)(Ra,Rb)→(RA,RB,RC)},{(Ra,Ha1)→(RA,HA1),(RA,Ha2)→(RA,HA2),(RB,Hb1)→(RC,HC1)}}. When the electric energy perception cross-domain virtual network mapping method or system is adopted, the obtained node mapping scheme is Rb→RDThe node mapping scheme is more current than scheme Rb→RCIs superior. Because, on the one hand, RCThe node is in an inactive state, and large energy consumption is needed for starting the node. In another aspect, RAAnd RCR on the link betweenBThe node is also in inactive state and also needs to be opened for forwarding the data packet. In contrast, Rb→RDThis solution effectively avoids both of the above problems.
The electric energy perception cross-domain virtual network mapping method and the electric energy perception cross-domain virtual network mapping system can keep the long-term operation income of the physical network in the cross-domain virtual network mapping process and reduce the electric energy expenditure of the physical network.
While the preferred embodiments of the present invention have been illustrated in the accompanying drawings, those skilled in the art will appreciate that various modifications can be made to the present invention without departing from the scope and spirit of the invention. For instance, features illustrated or described as part of one embodiment, can be used with another embodiment to yield a still further embodiment. The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the scope of the present invention, which is defined in the appended claims.
Claims (10)
1. A power-aware cross-domain virtual network mapping method comprises mapping a virtual node to a physical node and mapping a virtual link to a physical link, and is characterized in that:
when the virtual node is mapped to the physical node, respectively mapping a virtual routing node and a virtual host node in the virtual node to a corresponding physical routing node and a corresponding physical host node in the physical node; wherein:
when mapping the virtual routing node to the corresponding physical routing node, the method comprises the following steps:
a1, constructing a physical routing node list, wherein each physical routing node in the physical routing node list can meet the node requirement of the virtual routing node;
a2, calculating the electric energy cost generated by mapping the virtual routing nodes for each physical routing node, and obtaining the NR priority of each physical routing node according to the degree and the bandwidth resource of each physical routing node;
a3, mapping the virtual routing node to the physical routing node with the highest NR priority;
when mapping the virtual host node to the corresponding physical host node, the method comprises the following steps:
b1, searching a physical host node set connected with the physical routing node mapped by all the virtual routing nodes;
b2, calculating the average distribution value of the resource utilization rate of each physical host node in the physical host node set;
b3, mapping the virtual host node to the physical host node with the average distribution value of the minimum resource utilization rate.
2. The power-aware cross-domain virtual network mapping method according to claim 1, wherein the step a2 specifically includes:
when mapping virtual routing node r, using a formulaAnd <math>
<mrow>
<msubsup>
<mi>ΔPN</mi>
<mi>i</mi>
<mi>r</mi>
</msubsup>
<mo>=</mo>
<mfenced open='{' close='}'>
<mtable>
<mtr>
<mtd>
<msub>
<mi>P</mi>
<mi>fix</mi>
</msub>
<mo>+</mo>
<mi>ΔL</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>lc</mi>
</msub>
<mo>+</mo>
<mi>ΔP</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>port</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>if</mi>
<msub>
<mi>PS</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mi>ΔL</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>lc</mi>
</msub>
<mo>+</mo>
<mi>ΔP</mi>
<mo>·</mo>
<msub>
<mi>P</mi>
<mi>port</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>otherwise</mi>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
</math> calculating the electric energy cost generated by mapping the virtual routing node for each physical routing node i; wherein Pri(t) represents the electricity price of the area where the physical routing node i is located in the current time window t, PfixRepresenting fixed energy consumption, P, within chassis, etc., of a physical routing nodelcIndicating energy consumption of line cards, PportRepresenting the energy consumption of each port, L and P respectively representing the number of line cards and ports, PSi0 represents that the physical routing node is in an inactive state, and the other represents that the physical routing node is in an active state;
sorting all physical routing nodes in the physical routing node list in ascending order according to the corresponding electric energy cost, and using NRE(i) Representing the sequence number of the physical routing node i in the ascending order;
calculating all physical routing nodes in the physical routing node list by using a Noderank node measurement methodThe Noderank value of the point is used for representing the degree and bandwidth resources of each physical node, and the physical routing nodes are sorted from large to small according to the Noderank value, and NR is usedR(i) Representing the sequence number of the physical routing node i in the sequence;
using the formula NR ═ α NRE(i)+(1-α)NRR(i) And obtaining the NR priority of each physical routing node, wherein alpha satisfies the condition that alpha is more than or equal to 0 and less than or equal to 1.
3. The power-aware cross-domain virtual network mapping method according to claim 1, wherein when mapping the virtual routing nodes onto the corresponding physical routing nodes, the method further comprises step a 4: if the mapping is successful in step a3, marking the physical routing node with the highest NR priority to which the virtual routing node is mapped, and excluding the marked physical routing node from the physical routing node list.
4. The power-aware cross-domain virtual network mapping method of claim 1, wherein in step B2, a formula is used <math>
<mrow>
<msub>
<mi>δ</mi>
<mi>j</mi>
</msub>
<mo>=</mo>
<msqrt>
<mfrac>
<mrow>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>C</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>M</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>D</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>B</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<msub>
<mover>
<mi>u</mi>
<mo>‾</mo>
</mover>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
<mn>4</mn>
</mfrac>
</msqrt>
</mrow>
</math> Calculating an average distribution value of resource utilization rate of each physical host node j in the physical host node set; wherein, Cj、MjAnd SjRespectively representing available CPU resources, available memory resources and available hard disk storage resources after the virtual host node is pre-mapped to j; b isjRepresenting the link l to which the virtual host node is connectedhPre-mapping to j connected link ljThe next available bandwidth resource; u. ofjRepresents the average resource utilization, u, of the physical host node jj＝(Cj+Mj+Sj+Bj)/4。
5. The power-aware cross-domain virtual network mapping method according to claim 1, further comprising step B4, when mapping the virtual host nodes onto the corresponding physical host nodes: if the mapping is successful in step B3, marking the average distribution value with the minimum resource utilization to which the virtual host node is mapped on the physical host node, where the physical host node set in step B1 does not include the marked physical host node.
6. The power-aware cross-domain virtual network mapping method according to claim 1, comprising the steps of, when mapping the virtual links onto the physical links:
c1, calculating the distance between every two physical routing nodes in advance according to a Flouard algorithm;
c2, searching a set of all physical links with a distance length within [ LEN, MAXLEN ] capable of meeting the bandwidth requirement of the virtual link from two physical host nodes, wherein LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is the maximum number of hops that can be accepted by a physical loop-free path to which the virtual link is mapped;
c3, calculating the weighted value of all physical links in the set by using the weighting method of the number of unopened nodes and ports;
c4, mapping the virtual link to the path with the least good and bad weighted value.
7. The power-aware cross-domain virtual network mapping method according to claim 6, wherein in the step C3, the formula L ═ β N is usedf+(1-β)NpCalculating the quality weighted values of all physical links in the set; wherein N isfAnd NpThe number of the unopened forwarding nodes and the number of the unopened ports on the current evaluation path are respectively represented, and beta represents the proportion of the energy consumption of one unopened forwarding node and the energy consumption of one unopened port.
8. A power-aware cross-domain virtual network mapping system includes a node mapping unit that maps virtual nodes onto physical nodes and a link mapping unit that maps virtual links onto physical links, characterized in that:
the node mapping unit is composed of a routing node mapping subunit and a host node mapping subunit, and is respectively used for mapping the virtual routing node to the corresponding physical routing node and mapping the virtual host node to the corresponding physical host node; wherein:
the routing node mapping subunit includes:
the list construction module is used for constructing a physical routing node list, and each physical routing node in the physical routing node list can meet the node requirement of the virtual routing node;
the priority calculation module is used for calculating the electric energy cost generated by mapping the virtual routing nodes for each physical node and obtaining the NR priority of each physical routing node by combining the degree of each physical node and the bandwidth resource;
a routing node mapping module for mapping the virtual routing node to the physical routing node with the highest NR priority;
the host node mapping subunit includes:
a host node set searching module, configured to search a physical host node set connected to the physical routing nodes to which all the virtual routing nodes are mapped;
the resource utilization rate average distribution value calculation module is used for calculating the resource utilization rate average distribution value of each physical host node in the physical host node set;
and the host node mapping module is used for mapping the virtual host node to the physical host node with the average distribution value of the minimum resource utilization rate.
9. The power-aware cross-domain virtual network mapping system of claim 8, wherein the node mapping unit further comprises a marking module, configured to mark the physical routing node with the highest NR priority to which the virtual routing node is successfully mapped, or mark the physical hosting node with an average distribution value with the smallest resource utilization to which the virtual hosting node is successfully mapped.
10. The power-aware cross-domain virtual network mapping system of claim 8, wherein the link mapping unit comprises:
the distance calculation module is used for calculating the distance between every two physical routing nodes in advance according to a Flouard algorithm;
a link set searching module, configured to search, from between two physical host nodes, a set of all physical links having a distance length within [ LEN, MAXLEN ], where LEN is a matrix of the number of shortest paths between any two physical nodes, and MAXLEN is a maximum number of hops that can be accepted by a physical loop-free path to which a virtual link is mapped;
the quality weighted value calculation module is used for calculating the quality weighted values of all the physical links in the set by using a method of weighting the number of unopened nodes and ports;
and the link mapping module is used for mapping the virtual link to the path with the minimum quality weighting value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410589835.5A CN104320343A (en) | 2014-10-28 | 2014-10-28 | Electric energy perception cross-domain virtual network mapping method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410589835.5A CN104320343A (en) | 2014-10-28 | 2014-10-28 | Electric energy perception cross-domain virtual network mapping method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104320343A true CN104320343A (en) | 2015-01-28 |
Family
ID=52375517
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410589835.5A Pending CN104320343A (en) | 2014-10-28 | 2014-10-28 | Electric energy perception cross-domain virtual network mapping method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104320343A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104869044A (en) * | 2015-06-19 | 2015-08-26 | 中国人民解放军信息工程大学 | Mapping method and mapping device for virtual network |
| CN107566194A (en) * | 2017-10-20 | 2018-01-09 | 重庆邮电大学 | A kind of method for realizing the mapping of cross-domain virtual network network |
| CN107637024A (en) * | 2015-07-30 | 2018-01-26 | 华为技术有限公司 | The virtual network insertion that connectivity perceives |
| CN110995619A (en) * | 2019-10-17 | 2020-04-10 | 北京邮电大学 | Service quality aware virtual network mapping method and device |
| CN117707693A (en) * | 2023-12-11 | 2024-03-15 | 之江实验室 | Heterogeneous intelligent computing platform virtualization management system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102932479A (en) * | 2012-11-16 | 2013-02-13 | 北京邮电大学 | Virtual network mapping method for realizing topology awareness based on historical data |
| CN103457752A (en) * | 2012-05-30 | 2013-12-18 | 中国科学院声学研究所 | Virtual network mapping method |
| CN103475504A (en) * | 2013-08-23 | 2013-12-25 | 北京邮电大学 | Virtual network remapping method based on topology awareness |
-
2014
- 2014-10-28 CN CN201410589835.5A patent/CN104320343A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103457752A (en) * | 2012-05-30 | 2013-12-18 | 中国科学院声学研究所 | Virtual network mapping method |
| CN102932479A (en) * | 2012-11-16 | 2013-02-13 | 北京邮电大学 | Virtual network mapping method for realizing topology awareness based on historical data |
| CN103475504A (en) * | 2013-08-23 | 2013-12-25 | 北京邮电大学 | Virtual network remapping method based on topology awareness |
Non-Patent Citations (1)
| Title |
|---|
| SEN SU: "Energy-Aware Virtual Network Embbedding", 《IEEE/ACM TRANSCTIONS ON NETWORKING》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104869044A (en) * | 2015-06-19 | 2015-08-26 | 中国人民解放军信息工程大学 | Mapping method and mapping device for virtual network |
| CN104869044B (en) * | 2015-06-19 | 2019-01-29 | 中国人民解放军信息工程大学 | A kind of virtual net mapping method and device |
| CN107637024A (en) * | 2015-07-30 | 2018-01-26 | 华为技术有限公司 | The virtual network insertion that connectivity perceives |
| CN107637024B (en) * | 2015-07-30 | 2020-04-14 | 华为技术有限公司 | Related method and system for connectivity-aware virtual network embedding |
| CN107566194A (en) * | 2017-10-20 | 2018-01-09 | 重庆邮电大学 | A kind of method for realizing the mapping of cross-domain virtual network network |
| CN107566194B (en) * | 2017-10-20 | 2020-07-31 | 重庆邮电大学 | Method for realizing cross-domain virtual network mapping |
| CN110995619A (en) * | 2019-10-17 | 2020-04-10 | 北京邮电大学 | Service quality aware virtual network mapping method and device |
| CN110995619B (en) * | 2019-10-17 | 2021-09-28 | 北京邮电大学 | Service quality aware virtual network mapping method and device |
| CN117707693A (en) * | 2023-12-11 | 2024-03-15 | 之江实验室 | Heterogeneous intelligent computing platform virtualization management system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107566194B (en) | Method for realizing cross-domain virtual network mapping | |
| CN104320343A (en) | Electric energy perception cross-domain virtual network mapping method and system | |
| CN101119308B (en) | Routing device and method of wireless mobile self-organizing network of dynamic assurance service quality | |
| Xiao et al. | Simultaneous routing and resource allocation via dual decomposition | |
| Hou et al. | Rate allocation and network lifetime problems for wireless sensor networks | |
| Ko et al. | CG-E2S2: Consistency-guaranteed and energy-efficient sleep scheduling algorithm with data aggregation for IoT | |
| Huang et al. | Software-defined QoS provisioning for fog computing advanced wireless sensor networks | |
| CN103746852B (en) | Service routing configuration method and network management equipment | |
| Dandapat et al. | Smart association control in wireless mobile environment using max-flow | |
| CN103873364A (en) | Inter-domain multi-path rooting implementation method | |
| CN110061881B (en) | Energy consumption perception virtual network mapping algorithm based on Internet of things | |
| CN106162680A (en) | Dilatation parameter determination method and device | |
| Campos et al. | A fast algorithm for computing minimum routing cost spanning trees | |
| Zhu et al. | A modified ACO algorithm for virtual network embedding based on graph decomposition | |
| CN101616431A (en) | The implementation method of network in wireless mesh network and node | |
| Behrouz Vaziri et al. | Brad-OF: an enhanced energy-aware method for parent selection and congestion avoidance in RPL protocol | |
| Sharma et al. | Fuzzy based energy efficient clustering for designing WSN-based smart parking systems | |
| Hu et al. | Joint load balancing and offloading optimization in multiple parked vehicle‐assisted edge computing | |
| Xie et al. | A throughput-aware joint vehicle route and access network selection approach based on SMDP | |
| Khan et al. | Congestion avoidance in wireless sensor network using software defined network | |
| CN102946443B (en) | Multitask scheduling method for realizing large-scale data transmission | |
| CN103618674A (en) | A united packet scheduling and channel allocation routing method based on an adaptive service model | |
| Ducrocq et al. | Balancing energy consumption in clustered wireless sensor networks | |
| Zhang et al. | Effect of transfer costs on traffic dynamics of multimodal transportation networks | |
| Mao et al. | A controller-based roadside unit plane architecture for software-defined internet of vehicles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150128 |
|
| RJ01 | Rejection of invention patent application after publication |