A network slice deployment method combining VNF partitioning and hybrid backup sharing

Abstract

The application relates to a network slice deployment method combining VNF partition and mixed backup sharing, and belongs to the technical field of network slice deployment. The method comprises the following steps: S1: modeling a physical network model, a network slice model and a network slice pre-deployment model; S2: modeling a network slice reliability calculation model; S3: modeling a VNF mixed backup method; S4: modeling a method combining VNF partition and mixed backup sharing; S5: modeling a network slice end-to-end delay model and a network slice resource consumption cost model; and S6: modeling network slice deployment restriction conditions, VNF partition restriction conditions and mixed backup sharing restriction conditions. The application can effectively improve the reliability of VNFs with insufficient reliability in a network slice, so that the VNFs can meet the expected reliability of the network slice; can effectively reduce the end-to-end delay of the network slice, so that the end-to-end delay requirement can be met; and can significantly reduce the resource consumption of the network slice deployment while meeting the above two requirements.

Description

Technical Field

[0001] This invention belongs to the field of network slice deployment technology, and relates to a network slice deployment method that combines VNF partitioning and hybrid backup sharing. Background Technology

[0002] Compared to current mobile communication systems, 6G and future networks will not only improve basic performance requirements (such as increased data transmission rates, enhanced spectrum and efficiency, and expanded network coverage), but are also envisioned to possess unique characteristics, such as resource virtualization through network slicing and end-user virtualization through digital twins. These features will flexibly support new application scenarios, including the ultra-high reliability and ultra-low latency required for autonomous driving and the Industrial Internet of Things (IIoT). Network slicing, as a revolutionary architectural solution, aims to logically divide the underlying infrastructure into customized and isolated network slices to enable the provision of heterogeneous services, service for countless bandwidth consumers, and hosting for numerous vertical industries.

[0003] However, network slicing presents new challenges in terms of availability and service continuity. Compared to traditional hardware-based network functions, VNFs implemented through software modules are more prone to failure. The failure of a single VNF in a network slice can be caused by software errors, malicious attacks, misconfigurations, and hardware malfunctions, which can disrupt the service of the entire slice, leading to user dissatisfaction and revenue loss. Therefore, ensuring service availability in a virtualized network slicing environment is more complex than in traditional networks. Furthermore, considering service availability, network slice deployment requires significant resources, making resource-efficient reliability enhancement methods particularly important in today's resource-constrained network environments. Existing research has improved network slice reliability by considering VNF redundancy backup or migration and designing corresponding algorithms and strategies; and reduced latency and resource consumption by reducing the number of hops in VNF transmissions within a network slice. However, current work rarely comprehensively considers the integrated optimization of reliability, end-to-end latency, and network slice deployment resource consumption. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a network slice deployment method that combines VNF partitioning and hybrid backup sharing, deploys network slices into the physical network, optimizes the reliability and end-to-end latency performance of network slices, and minimizes the deployment cost of network slices.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for deploying network slices with combined VNF partitions and hybrid backup sharing includes the following steps:

[0007] S1: Modeling physical network model, network slice model, and network slice pre-deployment model;

[0008] S2: Modeling the reliability calculation model for network slices;

[0009] S3: Modeling a hybrid backup method for VNFs;

[0010] S4: Modeling a combined VNF partition and hybrid backup sharing method;

[0011] S5: Model the end-to-end latency model and the resource consumption cost model of network slices;

[0012] S6: Modeling network slice deployment restrictions, VNF partition restrictions, and hybrid backup sharing restrictions.

[0013] Furthermore, the physical network model is modeled as follows: Let the undirected weighted graph G... s = (N, E) represents the physical network topology, where N (n, m ∈ N) represents the set of physical nodes in the physical network, and E ((m, n) ∈ E) represents the set of physical links in the physical network; let Represents the CPU resource capacity of a physical node, let Let R represent the storage resource capacity of a physical node. n To represent the reliability of physical nodes, let This represents the bandwidth resource capacity of the physical link (m,n).

[0014] Furthermore, the network slicing model is modeled as follows: Let I represent the set of network slices; let... Let represent the expected reliability of network slice i∈I, and let Let i ∈ I represent the time delay requirement of a network slice; let the undirected weighted graph... Represents a network slice, where V i and L i Let VNF ​​and virtual link set represent respectively; VNFv∈V on network slice i i The CPU resource requirements make VNFv∈V on network slice i i The storage resource requirements make Indicates a virtual link (u,v)∈L i The bandwidth resource requirements.

[0015] Furthermore, the network slice pre-deployment model is modeled as follows: the network slice deployment is modeled as virtual network embedding, which is divided into virtual link embedding and virtual node embedding; firstly, the shortest physical path that meets the network slice virtual link resource requirements is found in the physical network; then, for all physical nodes in the determined shortest path, they are sorted according to their reliability, and the physical node with high reliability is selected to deploy the main VNF instance.

[0016] Furthermore, the network slice reliability calculation model is modeled as follows:

[0017] Let the binary variable Represents the VNF v∈V in network slice i∈I. i Deployed on physical nodes n∈N, This indicates that it is not deployed on physical nodes n∈N; let the binary variable... Indicates a virtual link (u,v)∈L i Mapped to the physical link (n,m)∈E, This indicates that (n,m)∈E is not mapped to a physical link;

[0018] VNF reliability is represented as the product of its software reliability and the hardware reliability of its physical node, where R is the reliability of the virtual network. i,v To represent the reliability of VNF, let P i,v The actual reliability of each VNF is represented by the model as follows:

[0019]

[0020] Among them, R n Indicates the reliability of physical nodes;

[0021] Let P i The actual reliability of each network slice is represented by the following model:

[0022]

[0023] when At this time, network slicing does not meet the expected reliability;

[0024] Reliability is improved through backup sharing by decomposing the expected reliability of network slices into the expected reliability of each VNF to simplify reliability calculations; let The expected reliability of each primary VNF is represented by the following model:

[0025]

[0026] The reliability allocation method assigns the reliability requirements of a network slice to each VNF.

[0027] Furthermore, the VNF hybrid backup method is modeled as follows: For network slices that do not meet reliability requirements, the reliability of the network slice is improved by backing up VNFs; backup VNFs are divided into on-site backups and off-site backups; on-site backup VNFs are deployed on the same physical node as the primary VNF, while off-site backup VNFs are instantiated on a different physical node than the one carrying the primary VNF.

[0028] When on-site backup and off-site backup occur simultaneously, it is called a hybrid backup method; let integer variables... Indicates whether VNFv in network slice i is backed up; if so, then and It also indicates the number of backups; otherwise... Let binary variables Indicate whether VNFv is backed up on physical node n; let P i hybrid The actual reliability of the VNF when using a hybrid backup scheme is modeled as follows:

[0029]

[0030] Furthermore, the joint VNF ​​partitioning and hybrid backup sharing method is modeled as follows:

[0031] When VNFv∈V i When the actual reliability does not meet the expected reliability, the first step is to find a backup VNF that can be shared, and to meet the requirements of different latency constraints by controlling the maximum number of shareable backups and the number of hops after backup sharing.

[0032] The primary VNF is divided into multiple lower-capacity sub-VNFs that run in parallel; the sub-VNFs execute the same software applications as the primary VNF, and the reliability of each VNF remains unchanged after partitioning; let the positive integer variable l i,v ∈N + Indicates VNFv∈V i Whether it is divided into multiple sub-VNFs, l i,v =1 indicates that VNFv has not been partitioned, l i,v >1 indicates that VNFv is partitioned, and l i,v Indicates the number of partitions in VNFv;

[0033] Let binary variable y i,v,n Indicates whether VNFv in network slice i is backed up on physical node n, y i,v,n =0 indicates that VNFv is not backed up on node n, when y i,v,n =1 and Indicates on-site backup, when y i,v,n =1 and Indicates off-site backup; The actual reliability of a VNF under a VNF partitioning and hybrid backup scheme is modeled as follows:

[0034]

[0035] in,

[0036] Furthermore, the end-to-end latency model for network slicing is modeled as follows:

[0037] Each VNF in the network slice is modeled as an M / M / 1 queue, and the entire network slice is regarded as a concatenation of multiple M / M / 1 queues;

[0038] make The average response time of network slice i is represented by the model:

[0039]

[0040] Where, μ i,v =c i,v χ / α is the processing rate of VNFv in network slice i, c i,v χ represents the allocated CPU resources, α represents the CPU frequency, and α represents a fixed constant representing the number of CPU cycles required for each process.

[0041] After performing VNF partitioning, assume that the processing capacity of each sub-VNF is μ. i,v / l i,v Model all child VNFs of a given VNF as an M / M / m queue, where the minimum value of m is 1 and the maximum value is 1. The average response time of network slice i is modeled as follows:

[0042]

[0043] in:

[0044]

[0045] The amount of data transmitted in a network slice is represented by Dt. i =f i ·fd i ·du i , where fd i and du i Let represent the average data size and lifespan of network slice i, respectively; The transmission latency of a network slice is represented by the following model:

[0046]

[0047] Let d iThe total latency of a network slice is represented by the following model:

[0048]

[0049] Furthermore, the network slice resource consumption cost model is modeled as follows:

[0050] Let binary variables Indicates a virtual link (u,v)∈L i Whether to back up on the physical link (m,n)∈E; let The deployment resource consumption of network slice i after combined VNF sharding and hybrid backup is modeled as follows:

[0051]

[0052] When the actual reliability of VNFv does not meet the expected reliability, we first look for unshareable backup VNFs, and by constraining the maximum number of shareable backups and the number of link hops after backup sharing, we can meet the requirements of different latency constraints.

[0053] Let the binary variable q i,u,v Indicates v∈V i With u∈V i Should backups be shared? The resource consumption of network slice i after joint VNF ​​sharding and hybrid backup sharing is modeled as follows:

[0054]

[0055] As the number of VNFs increases, sharing backup VNFs reduces the resource consumption of building backup VNFs, thereby reducing the resource consumption of network slice deployment.

[0056] Furthermore, the network slice deployment constraints, VNF partition constraints, and hybrid backup sharing constraints are modeled as follows: Let the binary variable p i,u,v Indicates v∈V i With u∈V i Are they the same type?

[0057] (1) Modeling VNF backup instance sharing constraints:

[0058]

[0059] (2) Resource limitations for modeling shared backup instances:

[0060]

[0061] (3) Modeling sharing quantity constraints:

[0062]

[0063] (4) Modeling VNF deployment constraints:

[0064]

[0065] (5) Modeling CPU resource constraints:

[0066]

[0067] (6) Modeling storage resource constraints:

[0068]

[0069] (7) Modeling bandwidth resource constraints:

[0070]

[0071] (8) Modeling flow conservation constraints:

[0072]

[0073] (9) Modeling end-to-end delay constraints:

[0074]

[0075] (10) Modeling reliability constraints:

[0076]

[0077] (11) Modeling constraints and integrity limitations:

[0078]

[0079] The beneficial effects of this invention are as follows: This invention can effectively improve the reliability of VNFs with insufficient reliability in network slices, so that they meet the expected reliability of network slices; it can effectively reduce the end-to-end latency of network slices, so that they meet the end-to-end latency requirements; while meeting the above two requirements, it can significantly reduce the resource consumption of network slice deployment.

[0080] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0081] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0082] Figure 1 This is a diagram of the combined VNF partitioning and hybrid backup sharing model used in this invention;

[0083] Figure 2 This is a schematic diagram illustrating the process of deploying network slices that combine VNF partitioning and hybrid backup sharing according to the present invention. Detailed Implementation

[0084] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0085] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0086] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0087] Please see Figures 1-2 , Figure 1 This is a diagram of the combined VNF partitioning and hybrid backup sharing model used in this invention, as shown below. Figure 1 As shown, this model deploys VNFs on physical machines in the underlying physical network and adopts VNF partitioning and hybrid backup sharing. By combining VNF partitioning and hybrid backup sharing methods for network slice deployment, the network slice deployment cost can be minimized while meeting the network slice reliability requirements and end-to-end latency requirements. Figure 2 This is a flowchart illustrating the network slice deployment method of the present invention, which combines VNF partitioning and hybrid backup sharing. Figure 2 As shown, the method specifically includes the following steps:

[0088] Step 1: Physical network modeling;

[0089] Physical network modeling, specifically: Let the undirected weighted graph G s = (N, E) represents the physical network topology, where N (n, m ∈ N) represents the set of physical nodes in the physical network, and E ((m, n) ∈ E) represents the set of physical links in the physical network; let Represents the CPU resource capacity of a physical node, let Let R represent the storage resource capacity of a physical node. n To represent the reliability of physical nodes, let This represents the bandwidth resource capacity of the physical link (m,n).

[0090] Step 2: Network slicing modeling;

[0091] Network slicing modeling, specifically: Let I represent the set of network slices; let... Let represent the expected reliability of network slice i∈I, and let Let i ∈ I represent the time delay requirement of a network slice; let the undirected weighted graph... Represents a network slice, where V i and L i Let VNF ​​and virtual link set represent respectively; VNFv∈V on network slice i i The CPU resource requirements make VNFv∈V on network slice i i The storage resource requirements make Indicates a virtual link (u,v)∈L i bandwidth resource requirements;

[0092] Step 3: Network slice pre-deployment modeling;

[0093] Network slice pre-deployment modeling is as follows: Network slice deployment is performed through virtual network embedding, using a non-coordinated approach, which is divided into virtual link embedding and virtual node embedding. To ensure end-to-end latency in network slices, the hop count between VNFs within the network slice needs to be minimized. Therefore, virtual network embedding is performed first. The K-shortest path algorithm is used in the physical network to find the shortest physical path that meets the virtual link resource requirements of the network slice. Then, for all physical nodes in the determined shortest path, they are sorted according to their reliability, and the most reliable physical nodes are selected for the appropriate deployment of the primary VNF instance.

[0094] Step 4: Network slice reliability calculation and modeling;

[0095] Network slicing reliability calculation modeling, specifically: Let binary variables... Represents the VNF v∈V in network slice i∈I. i Deployed on physical nodes n∈N, This indicates that it is not deployed on physical nodes n∈N; let the binary variable... Indicates a virtual link (u,v)∈L i Mapped to the physical link (n,m)∈E, This indicates that the link (n,m) is not mapped to the physical link E.

[0096] VNF reliability is represented as the product of its software reliability and the hardware reliability of its physical node, where R is the reliability of the virtual network. i,v To represent the reliability of VNF, let P i,v The actual reliability of each VNF can be modeled as follows:

[0097]

[0098] Among them, R n This represents the reliability of the physical node. Let P... i The actual reliability of each network slice can be modeled as follows:

[0099]

[0100] when At this time, network slices do not meet the expected reliability. This scheme improves reliability through backup sharing, decomposing the expected reliability of network slices into the expected reliability of each VNF to simplify reliability calculations. Let... The expected reliability of each primary VNF can be modeled as follows:

[0101]

[0102] The reliability allocation method assigns the reliability requirements of a network slice to each VNF.

[0103] Step 5: Modeling the VNF hybrid backup method;

[0104] The VNF hybrid backup method is modeled as follows: For network slices that do not meet reliability requirements, the reliability of the network slice can be improved by backing up VNFs. Backup VNFs can be divided into two forms: on-site backup and off-site backup. On-site backup VNFs are deployed on the same physical node as the primary VNF, while off-site backup VNFs are instantiated on a different physical node than the primary VNF. When on-site backup and off-site backup occur simultaneously, it is called a hybrid backup method. Let integer variables... Indicates whether VNFv in network slice i is backed up. If yes, then... and It also indicates the number of backups; otherwise... Let binary variables This indicates whether VNFv is backed up on physical node n. Therefore, let P... i hybrid The actual reliability of a VNF when using a hybrid backup scheme can be modeled as follows:

[0105]

[0106] Step 6: Modeling the combined VNF partitioning and hybrid backup sharing method;

[0107] The joint VNF ​​partitioning and hybrid backup sharing method is modeled as follows: when VNFv∈V i When actual reliability does not meet expected reliability, the first step is to find a backup VNF that can be shared. The maximum number of shareable backups and the hop count after backup sharing are controlled to meet different latency constraints. To improve resource utilization while ensuring network slice reliability with fewer backups, the primary VNF is divided into multiple parallel, lower-capacity sub-VNFs. The sub-VNFs execute the same software applications as the primary VNF, and the reliability of each VNF remains unchanged after partitioning. Let the positive integer variable l... i,v ∈N + VNF v∈V i Whether it is divided into multiple sub-VNFs, l i,v =1 indicates that VNFv has not been partitioned, l i,v >1 indicates that VNFv is partitioned, and l i,v This represents the number of partitions in VNFv. Let the binary variable y... i,v,n Indicates whether VNFv in network slice i is backed up on physical node n, y i,v,n =0 indicates that VNFv is not backed up on node n, when y i,v,n =1 and Indicates on-site backup, when y i,v,n =1 and This indicates off-site backup. Let P... i h+d The actual reliability of a VNF under a VNF partitioning and hybrid backup scheme can be modeled as follows:

[0108]

[0109] in,

[0110] Step 7: End-to-end latency modeling for network slicing;

[0111] Network slice end-to-end latency modeling specifically involves modeling each VNF in network slice i as an M / M / 1 queue, and treating the entire network slice as a concatenation of multiple M / M / 1 queues. According to Burke's theorem, the flow arrival rate f of all VNFs in the concatenation of multiple M / M / 1 queues is... i Same. Let The average response time of network slice i can be modeled as:

[0112]

[0113] Where, μ i,v =c i,v χ / α is the processing rate of VNFv in network slice i. Where c i,v χ represents the allocated CPU resources, χ represents the CPU frequency (cycles / second), and α represents a fixed constant representing the number of CPU cycles required for each process.

[0114] After performing VNF partitioning, assume that the processing capacity of each sub-VNF is μ. i,v / l i,v Model all child VNFs of a given VNF as an M / M / m queue, where the minimum value of m is 1 and the maximum value is 1. Therefore, the average response time of network slice i can be modeled as:

[0115]

[0116] in:

[0117]

[0118] The transmission latency of virtual links in network slicing is mainly related to the amount of data transmitted in the network slice, Dt. i It is related to the virtual link bandwidth. The amount of data transmitted in a network slice is represented by Dt. i =f i ·fd i ·du i , where fd i and du i Let represent the average data volume and lifespan of network slice i, respectively. The transmission latency of a network slice can be modeled as follows:

[0119]

[0120] Let d i The total latency of a network slice can be modeled as follows:

[0121]

[0122] Step 8: Model the resource consumption cost of network slicing;

[0123] Network slicing resource consumption cost modeling, specifically: Let binary variables... Indicates a virtual link (u,v)∈L i Whether to perform backup on the physical link (m,n)∈E. Therefore, let The deployment resource consumption of network slice i after combined VNF sharding and hybrid backup can be modeled as follows:

[0124]

[0125] To further reduce resource consumption during slice deployment, a backup sharing scheme is added on top of the above. When the actual reliability of a VNF does not meet the expected reliability, the system first searches for available backup VNFs that can be shared. By constraining the maximum number of shareable backup VNFs and the link hop count after backup sharing, different latency constraints can be met.

[0126] Let the binary variable q i,u,v Indicates v∈V i With u∈V i Should backups be shared? The resource consumption of deploying network slice i after joint VNF ​​sharding and hybrid backup sharing can be modeled as follows:

[0127]

[0128] As the number of VNFs increases, sharing backup VNFs can reduce the resource consumption of building backup VNFs, thereby reducing the resource consumption of network slice deployment.

[0129] Step 9: Network slice deployment, VNF partitioning, and hybrid backup sharing constraint modeling;

[0130] Modeling the constraints of network slicing deployment and backup sharing, specifically: Let the binary variable p i,u,v Indicates v∈V i With u∈V i Are they the same type?

[0131] (1) Modeling VNF backup instance sharing constraints:

[0132]

[0133] (2) Resource limitations for modeling shared backup instances:

[0134]

[0135] (3) Modeling sharing quantity constraints:

[0136]

[0137] (4) Modeling VNF deployment constraints:

[0138]

[0139] (5) Modeling CPU resource constraints:

[0140]

[0141] (6) Modeling storage resource constraints:

[0142]

[0143] (7) Modeling bandwidth resource constraints:

[0144]

[0145] (8) Modeling flow conservation constraints:

[0146]

[0147] (9) Modeling end-to-end delay constraints:

[0148]

[0149] (10) Modeling reliability constraints:

[0150]

[0151] (11) Modeling constraints and integrity limitations:

[0152]

[0153] In the above embodiments, the reference to "this embodiment" in the specification indicates that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least some embodiments, but not necessarily all embodiments. Multiple appearances of "this embodiment" do not necessarily all refer to the same embodiment.

[0154] In the above embodiments, although the invention has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory structures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed. The embodiments of the invention are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims.

[0155] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the methods in this embodiment.

[0156] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0157] The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to cause the terminal to perform any of the methods in this embodiment.

[0158] As will be understood by those skilled in the art, the computer-readable storage medium described in this embodiment allows for the implementation of all or part of the steps in the above method embodiments by computer program-related hardware. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0159] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication between them. The memory is used to store computer programs, the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer programs, so that the electronic terminal performs the steps of the above method.

[0160] In this embodiment, the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0161] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0162] This invention can be used in a wide range of general-purpose or special-purpose computing system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0163] This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0164] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A network slice deployment method combining VNF partitioning and hybrid backup sharing, characterized in that: Includes the following steps: S1: Modeling physical network model, network slice model, and network slice pre-deployment model; The physical network model is modeled as follows: Let an undirected weighted graph... Represents the physical network topology, where Represents the set of physical nodes in a physical network. , Let represent the set of physical links in a physical network; Represents the CPU resource capacity of a physical node, let Represents the storage resource capacity of a physical node, let To represent the reliability of physical nodes, let Indicates physical link Bandwidth resource capacity; The network slicing model is modeled as follows: Let Represents the set of network slices; let Represents network slices The expected reliability makes Represents network slices The latency requirement; let the undirected weighted graph Represents a network slice, where and Let VNF ​​and virtual link set represent respectively; Represents network slices VNF on The CPU resource requirements make Represents network slices VNF on The storage resource requirements make Indicates virtual link bandwidth resource requirements; The network slice pre-deployment model is modeled as follows: the network slice deployment is modeled as virtual network embedding, which is divided into virtual link embedding and virtual node embedding; first, the shortest physical path that meets the network slice virtual link resource requirements is found in the physical network; then, for all physical nodes in the determined shortest path, they are sorted according to their reliability, and the physical nodes with high reliability are selected to deploy the main VNF instance. S2: Modeling a network slice reliability calculation model; the network slice reliability calculation model is modeled as follows: Let the binary variable Represents network slices VNF in Deployed on physical nodes superior, Represents a VNF set. This indicates that it is not deployed on a physical node. Above; let the binary variable Indicates virtual link Mapping to physical link , This indicates that it is not mapped to a physical link. ; VNF reliability is represented as the product of its software reliability and the hardware reliability of the physical node it resides in, let To indicate the reliability of VNF, let The actual reliability of each VNF is represented by the model as follows: in, Represents network slices Expected reliability Indicates the reliability of physical nodes; make The actual reliability of each network slice is represented by the following model: when At this time, network slicing does not meet the expected reliability; Reliability is improved through backup sharing by decomposing the expected reliability of network slices into the expected reliability of each VNF to simplify reliability calculations; let The expected reliability of each primary VNF is represented by the following model: The reliability allocation method assigns the reliability requirements of network slices to each VNF; S3: Modeling a hybrid VNF backup method; The hybrid VNF backup method is modeled as follows: For network slices that do not meet reliability requirements, the reliability of the network slice is improved by backing up VNFs; Backup VNFs are divided into on-site backups and off-site backups; On-site backup VNFs are deployed on the same physical node as the primary VNF, while off-site backup VNFs are instantiated on a different physical node than the one carrying the primary VNF; When on-site backup and off-site backup occur simultaneously, it is called a hybrid backup method; let integer variables... Represents network slices VNF in Do you want to back up? If so, then... ,and It also indicates the number of backups; otherwise... Let binary variables VNF Is it at the physical node? Upload backup; Order The actual reliability of the VNF when using a hybrid backup scheme is modeled as follows: in Represents the set of physical nodes in a physical network. ; S4: Modeling a combined VNF partition and hybrid backup sharing method; the combined VNF partition and hybrid backup sharing method is modeled as follows: When VNF When the actual reliability does not meet the expected reliability, first look for unshareable backup VNFs, and control the maximum number of shareable backups and the number of hops after backup sharing to meet the requirements of different latency constraints. The primary VNF is divided into multiple lower-capacity sub-VNFs that run in parallel; the sub-VNFs execute the same software applications as the primary VNF, and the reliability of each VNF remains unchanged after partitioning; let positive integer variables... VNF Whether it is divided into multiple sub-VNFs, VNF It was not divided. VNF Divided, and VNF The number of divisions; Let binary variables Represents network slices VNF in Is it at the physical node? Back up the data. VNF At the node If no backup is performed, when and Indicates on-site backup, when and Indicates off-site backup; The actual reliability of a VNF under a VNF partitioning and hybrid backup scheme is modeled as follows: in, ; S5: Model the end-to-end latency model and the resource consumption cost model of network slices; The network slice end-to-end latency model is modeled as follows: Each VNF in the network slice is modeled as an M / M / 1 queue, and the entire network slice is regarded as a concatenation of multiple M / M / 1 queues; make Represents network slices The average response time is modeled as follows: in, It is network slicing VNF Processing speed, Indicates the allocated CPU resources. Indicates CPU frequency. This represents a fixed constant representing the number of CPU cycles required to process each process. Represents network slices The arrival rate of the flow; After performing VNF partitioning, assume that the processing capacity of each sub-VNF is... , VNF The number of partitions; modeling all child VNFs of a VNF as an M / M / m queue, where the minimum value of m is 1 and the maximum value is 1. Network slicing The average response time is modeled as follows: in: The amount of data transmitted in a network slice is expressed as follows: ,in and These represent network slices. Average data volume and lifespan of the stream; let The transmission latency of a network slice is represented by the following model: in Indicates virtual link The bandwidth resource requirements; The total latency of a network slice is represented by the following model: The network slice resource consumption cost model is modeled as follows: Let binary variables Indicates virtual link Is it on the physical link? Upload backup; Order This refers to network slices resulting from combined VNF sharding and hybrid backup. The deployment resource consumption is modeled as follows: When VNF When the actual reliability does not meet the expected reliability, we first look for unshareable backup VNFs, and then meet the requirements of different latency constraints by constraining the maximum number of shareable backups and the number of link hops after backup sharing. Let the binary variable express and Should backups be shared? This indicates a network slice resulting from combined VNF sharding and hybrid backup sharing. Deployment resource consumption is modeled as follows: Represents a set of virtual links; Indicates virtual link bandwidth resource requirements; Represents network slices VNF on CPU resource requirements; Represents network slices VNF on Storage resource requirements; Represents network slices VNF in Is it at the physical node? Back up the data. VNF At the node If no backup is performed, when and Indicates on-site backup, when and Indicates off-site backup; Indicates virtual link Mapping to physical link , This indicates that it is not mapped to a physical link. ; Represents the set of physical nodes in a physical network. ; When the number of VNFs increases, sharing backup VNFs reduces the resource consumption of building backup VNFs, thereby reducing the resource consumption of network slice deployment. S6: Model network slice deployment constraints, VNF partition constraints, and hybrid backup sharing constraints; the network slice deployment constraints, VNF partition constraints, and hybrid backup sharing constraints are modeled as follows: Let binary variables... express and Are they the same type? (1) Modeling VNF backup instance sharing constraints: (2) Resource limitations for modeling shared backup instances: (3) Constraints on the number of shared models: (4) Modeling VNF deployment constraints: (5) Modeling CPU resource constraints: (6) Modeling storage resource constraints: (7) Modeling bandwidth resource constraints: (8) Modeling flow conservation constraints: (9) Modeling end-to-end delay constraints: (10) Modeling reliability constraints: (11) Modeling constraint integrity restrictions: 。

Images