A virtual network end-to-end mapping method based on service request

Abstract

The present invention discloses a virtual network end-to-end mapping method based on service requests, including: modeling and describing the relevant network topology, objects and constraints involved in the virtual network mapping problem; customizing the elements and service requests involved in the multi-domain scenario; calculating the service request priority index and scheduling it according to the service request expression method; sorting the virtual nodes in the service request and sending them to the corresponding network domain in sequence; screening a group of candidate physical nodes from each network domain for virtual node mapping; classifying the service request nodes, adopting different mapping strategies for different types of nodes, and finally completing the mapping of the virtual network to the physical network. According to the present invention, the occurrence of node resource fragmentation is effectively reduced, the algorithm running time is reduced, thereby increasing the availability of node resources, and improving the overall service reception rate and overall revenue.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-domain network virtualization, and in particular to a virtual network end-to-end mapping method based on service request. Background Art

[0002] With the continuous expansion of the Internet, the rapid growth of the number of users and the emergence of various emerging applications, the traditional Internet is gradually unable to meet and adapt to the changes in current network service needs, and has gradually become bloated and rigid. In recent years, an effective way to solve the problem of network rigidity is network function virtualization. As one of the effective means of network virtualization technology, virtual network service orchestration can provide unified and abstract management for various heterogeneous network resources by taking advantage of the decoupling of underlying network physical resources and user service requests, and maximize network performance without changing the real physical architecture.

[0003] The virtual network orchestration process can be briefly divided into three steps. First, users at the upper layer of the network put forward personalized network service requirements, which are usually composed of a group of virtual nodes and virtual links. Then, after the service provider collects the relevant business requirements of the users, it starts the orchestration process, performs technical translation, business scheduling, and corresponding mapping strategy selection, and determines the network deployment plan. Finally, the service provider needs to apply for physical network resources from different infrastructure providers so that different nodes can be deployed in different network domains. When the above steps are successful, the service provider will provide the activated network services to users. Therefore, the key to the virtual network orchestration problem lies in allocating appropriate resources to virtual nodes and links, and how to create more value by using shared physical resources.

[0004] At present, virtual network service orchestration methods are mainly divided into two categories: orchestration methods for single-domain services and orchestration methods for multi-domain services. The orchestration methods for single-domain services ignore the deployment requirements for multi-domain services in actual application scenarios. Since each domain belongs to a different infrastructure provider, it is necessary to maintain the privacy of each domain when processing virtual network services involving multiple network domains, and detailed information such as internal resources cannot be shared. At present, some multi-domain service orchestration methods only use sub-service division to deploy one by one, without considering the global location attributes, thereby reducing the service reception rate of the entire orchestration process and causing losses to the service provider. Therefore, it is very necessary to propose a virtual network end-to-end mapping method based on service requests.

[0005] Technical term: ARIMA (Autoregressive Integrated Moving Average Model): The ARIMA model consists of three parts: autoregressive model (AR), difference process (I), and moving average model (MA). The basic idea is to use the historical information of the data itself to predict the future. The label value at a point in time is affected by both the label value in the past period of time and the accidental events in the past period of time. That is, the ARIMA model assumes that the label value fluctuates around the general trend of time, where the trend is affected by historical labels, the fluctuation is affected by accidental events in a period of time, and the general trend itself is not necessarily stable.

[0006] Dijkstra algorithm: Dijkstra algorithm is a shortest path algorithm from one vertex to the rest of the vertices, solving the shortest path problem in a weighted graph. Its main feature is that it starts from the starting point and adopts a greedy algorithm strategy, traversing the adjacent nodes of the vertex closest to the starting point and not visited each time until it reaches the end point. Summary of the invention

[0007] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a virtual network end-to-end mapping method based on service requests, which effectively reduces the occurrence of node resource fragmentation and reduces the algorithm running time, thereby increasing the availability of node resources and improving the overall service reception rate and overall revenue. In order to achieve the above-mentioned purpose and other advantages according to the present invention, a virtual network end-to-end mapping method based on service requests is provided, comprising:

[0008] Receive user service requests;

[0009] Abstracting the service request to form a virtual network mapping;

[0010] According to the attributes of the service request, the service request is sent to a designated deployment network domain; any of the network domains has a corresponding set of candidate physical nodes;

[0011] It is determined that the virtual network node of the service request is a cross-domain node, then:

[0012] Obtaining the shortest path from the candidate physical node to the corresponding edge node from the cache, and selecting at least one node with the shortest path for deployment;

[0013] It is determined that the virtual network node of the service request is an intra-domain node, then:

[0014] For the intra-domain nodes that have neighboring nodes and cross-domain nodes among the adjacent nodes, obtain the shortest path between the intra-domain nodes and the edge nodes;

[0015] For the intra-domain node that has no neighboring domain node and cross-domain node among the adjacent nodes, the shortest path between the intra-domain node and its neighboring nodes is calculated by cache search or Dijkstra algorithm;

[0016] Among them, the cross-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is greater than a preset first threshold; the intra-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is less than a preset second threshold.

[0017] Preferably, it also includes:

[0018] It is determined that the virtual link of the service request is an intra-domain link, then:

[0019] Obtaining the shortest path between the node in the domain and the edge node from the cache, and selecting a link according to the shortest path;

[0020] It is determined that the virtual link of the service request is an inter-domain link, then:

[0021] The shortest path from the end node to the edge node is obtained, and a link is selected according to the shortest path.

[0022] Preferably, the deployment of the physical node is performed based on a service revenue constraint and / or a service reception rate constraint; the service revenue constraint refers to a constraint that the value of the first function is within a first specified interval; the service reception rate constraint refers to a constraint that the value of the second function is within a second specified interval;

[0023] The first function is:

[0024]

[0025] in, For the overall network business revenue, is a network service request. Is a virtual node CPU requirements, Is a virtual link The bandwidth requirement, A collection of virtual nodes. is a set of virtual links;

[0026] The second function is:

[0027] ,

[0028] in, is the overall process business acceptance rate, Indicates the number of virtual network service requests that have been successfully received. represents the total number of virtual network service requests, and T represents the service request duration.

[0029] Preferably, the plurality of service requests are based on a service request priority index Sequential scheduling;

[0030] satisfy:

[0031] ,

[0032] in, The revenue value generated by the virtual network service request in each time unit during its life cycle, is the QoS value requested by the service, This metric is used to measure the orchestration time of business requests. Related virtual nodes and links that need to be orchestrated, It is the service request time ratio;

[0033] Preferably, the step of sending the service request to a designated deployment network domain according to the attribute of the service request includes:

[0034] Calculate the global attribute value of the virtual node , according to the global attribute value The size sends the virtual nodes to the corresponding designated deployment network domain in turn;

[0035] satisfy:

[0036] ,

[0037] in, is the topological attribute value of the virtual node, for The adjacent nodes of Defined as the importance of adjacent nodes, it indicates the ratio of bandwidth capacity to total bandwidth capacity and is used to measure the relative importance of neighboring nodes.

[0038] Preferably, the candidate physical node set is obtained based on the following steps:

[0039] Get the fluctuation range Φ of CPU demand;

[0040] For a designated physical node, if it is determined that the first condition, the second condition, and the third condition are satisfied, the designated physical node is added to the candidate physical node set;

[0041] Update the designated physical node and return to the previous step until the number of elements in the candidate physical node set is equal to a preset defined number;

[0042] The first condition is that the physical node deployment domain is equal to the virtual node deployment domain requirement;

[0043] The second condition is that the adjacent bandwidth capacity around the physical node is better than the virtual node;

[0044] The third condition is that the number of physical nodes in the candidate node set of each virtual node in the service request is less than a preset defined number.

[0045] Preferably, it includes a receiving module for obtaining a user's service request;

[0046] A virtual node mapping module, used to select an optimal physical node from the physical network according to the resources required by the virtual node, and map the virtual node to the selected physical node;

[0047] The virtual link mapping module is used to map the virtual link of the virtual network to the physical link selected in the physical network according to the shortest path accessed in the cache of the virtual network map or the shortest path from the end node to the edge node.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows: by comprehensively considering the average business revenue and computational complexity, the occurrence of node resource fragmentation is effectively reduced, the algorithm running time is reduced, thereby increasing the availability of node resources and improving the overall business reception rate and overall revenue. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Modeling a virtual network end-to-end mapping based on a service request according to the virtual network end-to-end mapping method based on a service request according to the present invention;

[0050] Figure 2 Customizing a table diagram for network elements of the virtual network end-to-end mapping method based on service requests according to the present invention;

[0051] Figure 3 It is an anti-fragmentation algorithm diagram in the candidate node screening method based on time series of the virtual network end-to-end mapping method based on business request according to the present invention;

[0052] Figure 4 The figure is a schematic diagram of a path selection structure of a virtual network end-to-end mapping method based on a service request according to the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Reference Figure 1 ,A virtual network end-to-end mapping method based on a service request, comprises: receiving a user service request;

[0055] Abstracting the service request to form a virtual network mapping;

[0056] According to the attribute of the service request, the service request is sent to a designated deployment network domain; any of the network domains has a corresponding set of candidate physical nodes.

[0057] In the process of mapping the virtual network to the actual network, the communication network of the network operator is used as the physical network mapped to the virtual network, that is, the physical network can be described as a weighted undirected graph ,in Represents a collection of physical nodes. Represents a set of physical links; each physical node Has CPU capacity attribute With domain location attribute , each physical link Has corresponding bandwidth capacity attributes Transmission delay attribute , is the set of all loop-free paths in the underlying network, represented by Composition, its .

[0058] The problem is modeled by using the user's business request abstraction as a virtual network mapped by the virtual network. The virtual network consists of a weighted undirected graph Indicates that is a set of virtual nodes, is a set of virtual links, is the waiting time for virtual network service requests. The life cycle of a business request; each virtual node Has CPU requirement attribute The corresponding deployment domain , each virtual link has a corresponding bandwidth requirement and delay constraints .

[0059] The goal to be optimized in the model is to improve the revenue created in the business process and the virtual network service reception rate in the overall orchestration process under the premise of satisfying resource constraints and mapping constraints.

[0060] The description of business benefits is as follows:

[0061] ,

[0062] in is business income, is a network service request. Is a virtual node CPU requirements, Is a virtual link bandwidth requirements.

[0063] The service reception rate is described as follows:

[0064] ,

[0065] in is the overall process business acceptance rate, Indicates the number of virtual network service requests that have been successfully received. represents the total number of virtual network service requests, and T represents the service request duration.

[0066] The constraints of the model are:

[0067] ,

[0068] ,

[0069] ,

[0070] ,

[0071] ,

[0072] in Is a physical node CPU capacity, Is a virtual node CPU requirements, is a collection of physical nodes, is a collection of virtual nodes; is the bandwidth capacity of the physical link, Is a virtual link The bandwidth requirement, is the set of loop-free paths of physical links, It is a set of virtual links in a mapped virtual network; is the transmission delay of the physical link, is the transmission delay requirement of the virtual link; Represents a physical node and virtual nodes It is a one-to-one mapping relationship; is the physical node deployment domain, It is the virtual node deployment domain requirement. Indicates that the virtual node to be mapped has a corresponding physical node.

[0073] Therefore, a virtual network end-to-end mapping method based on service requests specifically includes two steps:

[0074] Map virtual nodes to physical nodes that meet their virtual node requirements;

[0075] Map virtual links to physical links that meet their virtual link requirements.

[0076] Mapping a virtual node to a physical node that meets its virtual node requirements specifically includes the following steps: sending the service request to a designated deployment network domain according to the attributes of the service request; any of the network domains has a corresponding set of candidate physical nodes;

[0077] It is determined that the virtual network node of the service request is a cross-domain node, then:

[0078] Obtain the shortest path from the candidate physical node to the corresponding edge node from the cache; compare these paths proportionally according to the connection with the adjacent domain, and select a better node deployment; the shortest path of the corresponding edge node, such as Figure 4 ,Physical network topology: Node 3 in domain 1 to nodes 9 and 10 in domain 2. Nodes 9 and 10 are the corresponding edge nodes of node 3; the path ratios are 30 / 100 and 36 / 99.

[0079] It is determined that the virtual network node of the service request is an intra-domain node, then:

[0080] For the intra-domain node having neighboring domain nodes and cross-domain nodes among the adjacent nodes, obtaining the shortest path between the intra-domain node and the edge node so that the location of the physical node is closer to the acceptable domain node;

[0081] For the intra-domain node that has no neighboring domain node and cross-domain node among the adjacent nodes, the shortest path between the intra-domain node and its neighboring nodes is calculated by cache search or Dijkstra algorithm to select a better node;

[0082] Among them, the cross-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is greater than a preset first threshold; the intra-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is less than a preset second threshold.

[0083] In this application, virtual nodes are mapped to physical nodes that meet their virtual node requirements. Virtual node mapping is performed in a multi-domain scenario. The global attribute values ​​of all virtual nodes in the business request are first calculated, and the virtual nodes are sent to the corresponding designated deployment network domain in sequence according to the size of the global attributes.

[0084] The calculation of the virtual node global attribute value is as follows:

[0085] ,

[0086] in is the topological attribute value of the virtual node, Defined as the importance of adjacent nodes, it indicates the ratio of bandwidth capacity to total bandwidth capacity and is used to measure the relative importance of neighboring nodes. and The detailed definition is as follows:

[0087] ,

[0088] ,

[0089] ,

[0090] ,

[0091] in Indicates the strength attribute of a virtual node. For a virtual node , whose strength attribute represents the weighted sum of the strengths of its adjacent links, Represents neighbor nodes

[0092] Link bandwidth capacity The sum of is the node load rate. is the maximum computing resource of the node, The remaining computing resources of the node. The CPU computing power before mapping the request business, The CPU computing power required to request the service. Indicates a virtual link bandwidth requirements.

[0093] In an embodiment of the present invention, during the virtual node mapping process, multiple physical nodes are screened based on the candidate node screening method of the time series in each network domain to form a candidate node set, so that one virtual node in each network domain corresponds to multiple physical nodes. Figure 3As shown, when screening candidate nodes, the anti-fragmentation algorithm is used, based on the record of previous CPU demand conditions, the differential autoregressive moving average model (ARIMA) is used to predict the fluctuation range of the CPU demand value, and the physical nodes with sufficient available CPU resources are selected as candidate physical nodes; then, the physical nodes with available CPU resources closest to the CPU demand value of the virtual node to be currently scheduled are selected as candidate nodes. During screening, the predicted resource demand value of the next business node is reserved in the available resources of the physical node, and the physical node with the available resource value closest to the resource demand value is selected, so that the physical node can have resources for deploying virtual nodes in other businesses.

[0094] The specific steps for screening candidate physical nodes are as follows:

[0095] According to the current virtual network service, each virtual node , filter out physical nodes with insufficient CPU resources, and add physical nodes that meet the requirements to the historical demand data set . Use the ARIMA model to predict the next possible expected value of the CPU demand . Based on historical demand data sets Calculate the standard deviation of CPU demand , and then calculate the fluctuation range of CPU demand:

[0096] ;

[0097] For all virtual nodes in the virtual service request and physical nodes not deployed in the physical network, when ,if The surrounding adjacent bandwidth capacity is better than , and the candidate node set of each virtual node in the business request If the number of physical nodes in the table is less than the defined number, then add the physical node to the table. In; if If the number of physical nodes in the calculation is greater than the defined number, then the corresponding physical nodes are calculated. The surrounding resource status value:

[0098] ,

[0099] and select The physical node with the largest value.

[0100] For all virtual nodes in the virtual service request and physical nodes not deployed in the physical network, when ,if The surrounding adjacent bandwidth capacity is no better than , and the candidate node set of each virtual node in the business request If the number of physical nodes in the network is less than the defined number, the node with sufficient adjacent bandwidth capacity and the closest available CPU resources is selected. Node.

[0101] In this embodiment, a cache area is established in the model of the present application, and the cache area is used to store the shortest links between nodes and the shortest paths between deployed nodes and edge nodes that have been successfully deployed in the previous cycle. When the shortest path in the cache area does not meet the resource requirements in the newly arrived virtual network service request, the latest shortest path with sufficient resources is updated to the cache.

[0102] The virtual nodes in the virtual network include cross-domain nodes and intra-domain nodes, where the cross-domain nodes are nodes whose number of virtual links across the domain is greater than or equal to the number of virtual links connected only within the domain. When mapping the cross-domain nodes, the shortest path from the candidate node to the corresponding edge node is obtained from the cache; according to the connection with the adjacent domain, the paths are compared proportionally and a better physical node deployment is selected.

[0103] When mapping intra-domain nodes, if there are neighboring domain nodes and cross-domain nodes among the adjacent nodes, then the shortest path between the intra-domain nodes and the edge nodes is obtained to make the location of the physical node closer to the acceptable neighboring nodes; if there are only intra-domain nodes among the neighboring nodes, then the shortest path between the intra-domain node and its neighboring nodes is calculated through cache search or Dijkstra algorithm to select a better physical node.

[0104] In an embodiment of the present invention, a virtual link is mapped to a physical link that meets its virtual link requirements, and the virtual link of the service request is determined to be an intra-domain link.

[0105] Then: obtain the shortest path between the node in the domain and the edge node from the cache, and select a link according to the shortest path;

[0106] It is determined that the virtual link of the service request is an inter-domain link, then:

[0107] Get the shortest path from the end node to the edge node, and select a link based on the shortest path. The shortest path is a ratio value: for example, 30 / 100, where the left side of " / " is the delay and the right side is the bandwidth capacity. The smaller the value, the shorter the path.

[0108] The specific steps include:

[0109] If it is determined that the virtual link is an intra-domain link, then:

[0110] Obtaining from a cache the shortest path calculated by the Dijkstra algorithm, and selecting a link according to the shortest path;

[0111] It is determined that the virtual link is an inter-domain link, then:

[0112] Obtain the shortest path from the end node to the edge node, select a link according to the shortest path, and map the virtual link in the same way as the mapping of the virtual node. If there is no shortest path in the cache area that meets the requirements, the inter-domain link adopts the inter-domain path, which is composed of the shortest path from the end node to the edge node. The intra-domain link uses the Dijkstra algorithm to calculate the path, and finally updates the shortest path to the cache area.

[0113] Map all virtual nodes and virtual links according to the above steps until all virtual nodes and virtual links are successfully mapped. At this point, the virtual network The mapping is completed.

[0114] In this implementation, when multiple service requests are received in the model of this application, the priority index of the service request is calculated according to the service request expression method, and the mapping order of the service request is scheduled according to the priority index. The specific priority index is:

[0115] ,

[0116] in, The revenue value generated by the virtual network service request in each time unit during its life cycle, is the QoS value requested by the service, This metric is used to measure the orchestration time of business requests. Related virtual nodes and links that need to be orchestrated, It is the service request time ratio.

[0117] in , and The detailed definition is as follows:

[0118] ,

[0119] ,

[0120] ,

[0121] ,

[0122] in, is a variable with a value between 0 and 1, which is used to adjust the consideration ratio of CPU, bandwidth and delay in QoS calculation, and , yes The sum of the CPU requirements of all virtual nodes in the and They are The delay constraints and bandwidth requirements of all links in the network; is the value of the virtual node in the virtual network, is the value of the virtual link in the virtual network, yes The number of orchestration domains involved in The revenue value generated for the business request, The life cycle of a business request; Waiting time for business requests, The maximum waiting time.

[0123] like Figure 2 As shown, in this implementation, various elements involved in the multi-domain scenario are first abstracted and classified before being customized. The elements are divided into two types: objects and conditions. Objects are elements to be arranged in virtual network service requests, consisting of virtual nodes and virtual links, where virtual links are logical connections between virtual nodes; conditions can be divided into object conditions and overall conditions. Object conditions include object requirements, which include CPU requirements for virtual nodes, bandwidth requirements for virtual links, and requirements for deployment domains and object constraints, such as latency constraints of virtual links on physical paths; overall conditions are restrictions on the entire service request, which include service lifecycle and QoS requirements.

[0124] A virtual network end-to-end mapping device based on service request, comprising a receiving module for obtaining a user's service request;

[0125] An abstraction module, used to abstract the service request to form a virtual network mapping;

[0126] A service request sending module, used for sending the service request to a designated deployment network domain according to the attributes of the service request;

[0127] The virtual node mapping module determines that the virtual network node of the service request is a cross-domain node, then:

[0128] Obtaining the shortest path from the candidate physical node to the corresponding edge node from the cache, and selecting at least one node with the shortest path for deployment;

[0129] It is determined that the virtual network node of the service request is an intra-domain node, then:

[0130] For the intra-domain nodes that have neighboring nodes and cross-domain nodes among the adjacent nodes, obtain the shortest path between the intra-domain nodes and the edge nodes;

[0131] For the intra-domain node that has no neighboring domain node and cross-domain node among the adjacent nodes, the shortest path between the intra-domain node and its neighboring nodes is calculated by cache search or Dijkstra algorithm;

[0132] Among them, the cross-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is greater than a preset first threshold; the intra-domain node refers to a node in adjacent virtual links, where the ratio of cross-domain links to intra-domain links is less than a preset second threshold.

[0133] The virtual link mapping module is used to map the virtual link of the virtual network to the physical link selected in the physical network according to the shortest path accessed in the cache of the virtual network or the shortest path from the end node to the edge node. The shortest path is a ratio value: for example, 30 / 100, where the left side of " / " is the delay and the right side is the bandwidth capacity. The smaller the value, the shorter the path.

[0134] A terminal device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described in the present application when executing the computer program.

[0135] A computer-readable storage medium stores a computer program, wherein the computer program implements the method described in the present application when executed by a processor.

[0136] A computer program includes a software program for executing a virtual network end-to-end mapping method based on service requests.

[0137] The number of devices and processing scales described here are used to simplify the description of the present invention, and the application, modification and variation of the present invention will be obvious to those skilled in the art.

[0138] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and implementation modes. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A virtual network end-to-end mapping method based on service request, characterized in that: include: Receive user service requests; Abstracting the service request to form a virtual network mapping; According to the attributes of the service request, the service request is sent to a designated deployment network domain; any of the network domains has a corresponding set of candidate physical nodes; It is determined that the virtual network node of the service request is a cross-domain node, then: Obtaining the shortest path from the candidate physical node to the corresponding edge node from the cache, and selecting at least one node with the shortest path for deployment; It is determined that the virtual network node of the service request is an intra-domain node, then: For the intra-domain nodes that have neighboring nodes and cross-domain nodes among the adjacent nodes, obtain the shortest path between the intra-domain nodes and the edge nodes; For the intra-domain node that has no neighboring domain node and cross-domain node among the adjacent nodes, the shortest path between the intra-domain node and its neighboring nodes is calculated by cache search or Dijkstra algorithm; The cross-domain node refers to a node in adjacent virtual links where the ratio of cross-domain links to intra-domain links is greater than a preset first threshold; the intra-domain node refers to a node in adjacent virtual links where the ratio of cross-domain links to intra-domain links is less than a preset second threshold; The step of sending the service request to a designated deployment network domain according to the attribute of the service request includes: Calculate the global attribute value of the virtual node , according to the global attribute value The size sends the virtual nodes to the corresponding designated deployment network domain in turn; satisfy: ； in, For virtual nodes, is the topological attribute value of the virtual node, for The adjacent nodes of Defined as the importance of neighboring nodes, it represents the ratio of bandwidth capacity to total bandwidth capacity, and is used to measure the relative importance of neighboring nodes; The candidate physical node set is obtained based on the following steps: Get the fluctuation range set Φ of CPU demand; For a designated physical node, if it is determined that the first condition, the second condition, and the third condition are satisfied, the designated physical node is added to the candidate physical node set; Update the designated physical node and return to the previous step until the number of elements in the candidate physical node set is equal to a preset defined number; The first condition is that the physical node deployment domain meets the virtual node deployment domain requirements, and the CPU capacity of the physical node belongs to Φ; The second condition is that the adjacent bandwidth capacity of the physical node is better than the virtual node requirement; The third condition is that the number of physical nodes in the candidate node set is less than a preset defined number.

2. A virtual network end-to-end mapping method based on service request as claimed in claim 1, characterized in that: Also includes: It is determined that the virtual link of the service request is an intra-domain link, then: Obtaining the shortest path between the node in the domain and the edge node from the cache, and selecting a link according to the shortest path; It is determined that the virtual link of the service request is an inter-domain link, then: The shortest path from the end node to the edge node is obtained, and a link is selected according to the shortest path.

3. A virtual network end-to-end mapping method based on service request as claimed in claim 1 or 2, characterized in that: The deployment of the physical node is performed based on a service revenue constraint and / or a service reception rate constraint; the service revenue constraint refers to a constraint that a value of the first function is within a first specified interval; the service reception rate constraint refers to a constraint that a value of the second function is within a second specified interval; The first function is: ； in, For the overall network business revenue, is a network service request. Is a virtual node CPU requirements, Is a virtual link The bandwidth requirement, A collection of virtual nodes. is a set of virtual links; The second function is: ； in, is the overall process business acceptance rate, Indicates the number of virtual network service requests that have been successfully received. represents the total number of virtual network service requests, and T represents the service request duration.

4. A virtual network end-to-end mapping method based on service request as claimed in claim 1, characterized in that: The plurality of service requests are based on a service request priority index Sequential scheduling; satisfy: ； in, The revenue value generated by the virtual network service request in each time unit during its life cycle, is the QoS value requested by the service, This metric is used to measure the orchestration time of business requests. Related virtual nodes and links that need to be orchestrated, It is the service request time ratio.

5. A virtual network end-to-end mapping device based on service request, characterized in that: include: A receiving module is used to obtain the user's business request; An abstraction module, used to abstract the service request to form a virtual network mapping; A service request sending module, used for sending the service request to a designated deployment network domain according to the attributes of the service request; A virtual node mapping module determines that the virtual network node of the service request is a cross-domain node, then: obtain the shortest path from the candidate physical node to the corresponding edge node from the cache, and select at least one node with the shortest path for deployment; determine that the virtual network node of the service request is an intra-domain node, then: for the intra-domain node that has neighboring domain nodes and cross-domain nodes among adjacent nodes, obtain the shortest path between the intra-domain node and the edge node; for the intra-domain node that has no neighboring domain nodes and cross-domain nodes among adjacent nodes, calculate the shortest path between the intra-domain node and its neighboring nodes through cache search or Dijkstra algorithm; wherein, the cross-domain node refers to a node in an adjacent virtual link, where the ratio of the cross-domain link to the intra-domain link is greater than a preset first threshold; the intra-domain node refers to a node in an adjacent virtual link, where the ratio of the cross-domain link to the intra-domain link is less than a preset second threshold; The candidate physical node set module is used to obtain the candidate physical node set based on the following steps: Get the fluctuation range set Φ of CPU demand; For a designated physical node, if it is determined that the first condition, the second condition, and the third condition are satisfied, the designated physical node is added to the candidate physical node set; Update the designated physical node and return to the previous step until the number of elements in the candidate physical node set is equal to a preset defined number; The first condition is that the physical node deployment domain meets the virtual node deployment domain requirements, and the CPU capacity of the physical node belongs to Φ; The second condition is that the adjacent bandwidth capacity of the physical node is better than the virtual node requirement; The third condition is that the number of physical nodes in the candidate node set is less than a preset defined number.

6. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

8. A computer program product, characterized in that When the program is executed by a processor, the method according to any one of claims 1 to 4 can be implemented.

Images