A network slice deployment method based on priority VNF backup sharing

Abstract

The application relates to a network slice deployment method based on priority VNF backup sharing, and belongs to the network technology field, and comprises the following steps: constructing a physical network topology model, an URLLC end-to-end network slice model and a network slice pre-deployment model; calculating actual reliability of VNF deployment to a physical node through a network slice actual reliability calculation model; decomposing reliability of the network slice into expected reliability of each VNF through a reliability distribution method; performing backup sharing on the VNF in the network slice through a priority VNF backup sharing method; and modeling network slice end-to-end delay, a network slice resource consumption cost model, network slice deployment and backup sharing limitation conditions.

Description

Technical Field

[0001] The present invention belongs to the field of network technology and relates to a network slicing deployment method based on priority VNF backup sharing. Background Art

[0002] Network slicing, a revolutionary architectural solution, aims to logically partition the underlying infrastructure into customized and isolated network slices. This enables the provision of heterogeneous services, serving countless bandwidth consumers, and hosting numerous vertical industries. 5G technology is now widely deployed, and for 5G services, there is an ultra-reliable low-latency communication (URLLC) network slice. With the adoption of a wider range of applications in the sixth generation (6G), future 6G networks may expand from the uRLLC network slices of the 5G era to extremely reliable low-latency communication (ERLLC or eURLLC) network slices. In both types of slices, reliability and latency remain paramount concerns. By optimizing network slicing deployment solutions, it is expected that network slice reliability will be further improved and end-to-end latency will be reduced. Furthermore, the deployment cost of network slicing is a major concern for network slicing providers. Therefore, how to improve network slice reliability, reduce end-to-end latency, and minimize network slice deployment costs during network slicing deployment becomes a challenging issue. Existing research considers VNF redundancy or migration, designing corresponding algorithms and strategies to improve network slicing reliability; and reduces end-to-end latency by reducing the number of VNF transmission hops within a network slice. However, current work has rarely comprehensively considered the combined optimization of reliability and end-to-end latency. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a network slice deployment method based on priority VNF backup sharing, deploy network slices into the physical network, optimize the reliability and end-to-end delay performance of the network slices, and minimize the network slice deployment cost.

[0004] In order to achieve the above object, the present invention provides the following technical solutions:

[0005] A network slice deployment method based on priority VNF backup sharing includes the following steps:

[0006] S1: Build the physical network topology model, URLLC end-to-end network slicing model, and network slicing pre-deployment model;

[0007] S2: Calculate the actual reliability of the VNF after it is deployed on the physical node using the network slice actual reliability calculation model;

[0008] S3: Decompose the reliability of the network slice into the expected reliability of each VNF through the reliability allocation method;

[0009] S4: Backup and share the VNFs in the network slice using the priority VNF backup sharing method;

[0010] S5: Modeling network slicing end-to-end latency, network slicing resource consumption cost model, network slicing deployment and backup sharing constraints.

[0011] Furthermore, the physical network topology model is constructed as follows: Let the undirected weighted graph G s =(N,E) represents the physical network topology, where N represents the set of physical nodes in the physical network, with m,n∈N; E represents the set of physical links in the physical network, with (m,n)∈E; let Represents the CPU resource capacity of the physical node, let Represents the storage resource capacity of the physical node, let R n Represents the reliability of the physical node, let Indicates the bandwidth resource capacity of the physical link (m,n).

[0012] Furthermore, the following assumptions exist:

[0013] The VNFs in each network slice must be deployed on different physical nodes.

[0014] The reliability of each physical node is a known constant;

[0015] The reliability of a VNF depends on the reliability of the physical node that hosts it and the reliability of the VNF itself. VNF failures are independent and not affected by other physical nodes or VNFs.

[0016] The link bandwidth resource requirements in each network slice and the CPU resource requirements and storage resource requirements of the VNF are independent random variables and have an exponential distribution;

[0017] The reliability of the VNF backup instance is consistent with that of the primary instance;

[0018] Network slice requests arrive at the network sequentially.

[0019] Furthermore, the URLLC end-to-end network slicing model is constructed as follows:

[0020] Let S = {s i |i=1,2,...,I} represents the set of network slices, let Represents network slice s i The expected reliability of Denote the end-to-end delay requirement of the network slice; let the undirected weighted graph Gv =(V,L) represents the URLLC end-to-end network slice, where V represents the set of VNFs in the network slice, u,v∈V, L represents the set of virtual links, (u,v)∈L; let Denote the CPU resource requirement of v∈V, let Denote the storage resource requirement of v∈V, let Ty v Denote the VNF category of v∈V, let R v Denotes the self-reliability of v∈V; let represents the bandwidth resource requirement of the virtual link (u,v)∈L.

[0021] Furthermore, the network slice deployment is constructed as virtual network embedding, which is divided into virtual link embedding and virtual node embedding. First, the shortest physical path that meets the virtual link resource requirements of the network slice 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.

[0022] Furthermore, the actual reliability calculation model of the network slice in step S2 specifically includes: let p v represents the actual reliability of VNFv∈V after pre-deployment, which is modeled as:

[0023]

[0024] in, Represents the reliability of the physical node carrying VNFv∈V;

[0025] There are two basic VNF combinations: series and parallel. The reliability model of a complex network slice is:

[0026]

[0027] in Indicates the actual reliability of the network slice, Represents network slice s i Reliability of all serially connected VNFs in Represents network slice s i Reliability of all VNFs connected in parallel;

[0028] make Denotes the set of backup instances of VNFv∈V, let VNFv∈V contains the total set of primary and backup instances, where Represents the primary instance, Modeled as:

[0029]

[0030] Among them, V series represents a set of VNFs connected in series, Modeled as:

[0031]

[0032] Among them, V paralled Represents a set of VNFs connected in parallel.

[0033] Furthermore, the reliability allocation method in step S3 is as follows:

[0034] make Indicates the expected reliability of the slice The expected reliability of VNFv∈V under the requirement is modeled as:

[0035]

[0036] The reliability allocation method allocates the reliability requirements of the network slice to each VNF.

[0037] Furthermore, the priority VNF backup sharing method in step S4 includes:

[0038] Modeling the betweenness centrality of physical nodes n :

[0039]

[0040] Among them, σ st represents the number of shortest paths from physical node s to node t, σ st (n) represents the number of physical nodes n located on the shortest path from node s to node t;

[0041] Based on the betweenness centrality of the modeled physical nodes, all physical nodes in the physical network are prioritized. When a backup VNF instance is created and needs to be deployed, a node with a high priority is selected to deploy the backup VNF instance based on the physical node priority.

[0042] Furthermore, the end-to-end delay of the network slice specifically includes:

[0043] VNFs generate node delays during operation, including processing delays and queuing delays, which limit the service flow arrival rate of the network slice to be less than the VNF computing capacity. This can be modeled as follows:

[0044] λ≤comp v

[0045] Where λ represents the service flow arrival rate of the network slice, comp v represents the computing capability of the VNF node, let Represents network slice si The processing delay is modeled as:

[0046]

[0047] where dt v Indicates the amount of data transmitted, comp v represents the computing capability of VNFv∈V;

[0048] make Represents network slice s i The transmission delay of is modeled as:

[0049]

[0050] The end-to-end delay is modeled as:

[0051]

[0052] Furthermore, the network slicing resource consumption cost model specifically includes: setting the binary variable Indicates whether the primary or backup instance of VNFv∈V is mapped to the physical node n∈N. Let the binary variable Indicates whether the virtual link is mapped to the physical link (m,n)∈E. The total node resource consumption is modeled as:

[0053]

[0054] The total link bandwidth resource consumption is modeled as:

[0055]

[0056] in, Indicates a virtual link Is it mapped to the physical link (m,n)∈E? Indicates a virtual link Bandwidth resource requirements;

[0057] Therefore, network slices i Total resource consumption Re total Modeled as:

[0058] Re total =μ1C com +μ2B com

[0059] Among them, μ1 and μ2 are the weight coefficients of the two resource consumptions respectively.

[0060] Furthermore, the network slice deployment and backup sharing constraints specifically include: let the binary variable Indicates whether v∈V shares a backup instance with u∈V. Let the binary variable S(u,v) denote whether u∈V and v∈V are of the same type of VNF;

[0061] (1) Modeling VNF backup instance sharing constraints:

[0062]

[0063] (2) Modeling shared backup instance resource constraints:

[0064]

[0065] (3) Modeling sharing quantity restrictions:

[0066]

[0067] (4) Modeling VNF deployment constraints:

[0068]

[0069] (5) Modeling CPU resource constraints:

[0070]

[0071] (6) Modeling storage resource constraints:

[0072]

[0073] (7) Modeling bandwidth resource constraints:

[0074]

[0075] (8) Modeling flow conservation constraints:

[0076]

[0077] (9) Modeling end-to-end delay constraints:

[0078]

[0079] (10) Modeling constraint integrity restrictions:

[0080]

[0081]

[0082] The beneficial effects of the present invention are: the present 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 delay of network slices, so that they meet the end-to-end delay requirements; while meeting the above two requirements, it can save network slice deployment costs.

[0083] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0085] Figure 1 This is a diagram of the network slice deployment model based on priority VNF backup sharing adopted by the present invention;

[0086] Figure 2 This is a flow chart of the network slice deployment method based on priority VNF backup sharing of the present invention. DETAILED DESCRIPTION

[0087] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways 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 illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0088] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0089] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0090] See also Figures 1 and 2 , Figure 1 This is a network slice deployment model diagram based on priority VNF backup sharing adopted by the present invention, as shown in Figure 1 As shown in Figure 1, the model contains i network slice requests and a physical network. Network slice deployment through a priority-based VNF backup sharing method can minimize the network slice deployment cost while meeting the network slice reliability requirements and end-to-end delay requirements. Figure 2 This is a flow chart of a network slice deployment method based on priority VNF backup sharing according to the present invention. Figure 2 As shown, the method specifically includes the following steps:

[0091] Step 1: Physical network topology modeling;

[0092] Physical network topology modeling, specifically: set the physical network topology as a BA scale-free network, abstracted as an undirected weighted graph, represented by G s =(N,E), where N is the node set, representing the set of physical nodes in the physical network, with m,n∈N, and E is the edge set, representing the set of physical links in the physical network, with (m,n)∈E. The CPU resource capacity, storage resource capacity, and reliability of physical node n∈N are expressed as: and R n ; The bandwidth resource capacity of the physical link (m,n)∈E is expressed as

[0093] Step 2: URLLC end-to-end network slice modeling;

[0094] URLLC end-to-end network slicing modeling, specifically: For URLLC end-to-end network slicing, it is usually modeled as a linear chain structure, because the uRLLC network has extremely high requirements for latency and reliability, and the linear chain structure can provide an orderly packet processing path, allowing packets to pass through various network functions in a fixed order, thereby minimizing packet processing delay and transmission delay.

[0095] To meet the model requirements, the following reasonable assumptions need to be made:

[0096] 1. The VNFs in each network slice must be deployed on different physical nodes;

[0097] 2. The reliability of each physical node is a known constant;

[0098] 3. The reliability of a VNF depends on the reliability of the physical node that hosts it and the reliability of the VNF itself. Therefore, a VNF failure is independent and will not be affected by other physical nodes or VNFs.

[0099] 4. The link bandwidth resource requirements in each network slice and the CPU resource requirements and storage resource requirements of the VNF are independent random variables and have an exponential distribution.

[0100] 5. The reliability of the VNF backup instance is consistent with the reliability of the primary instance.

[0101] 6. Network slicing requests arrive at the network in sequence.

[0102] According to Assumption 6, the network slice request set is expressed as S = {s i |i=1,2,...,I}, the slices arrive at the network in sequence, and the network slice s i The expected reliability and end-to-end delay requirements of ∈S are expressed as and The URLLC end-to-end network slice is abstracted as an undirected weighted graph, represented as G v =(V,L), similar to the physical network topology, V is the node set, which is represented by the VNF set in the network slice, with u,v∈V, L is the edge set, which is represented by the virtual link set, with (u,v)∈L; the CPU resource requirement, storage resource requirement, category and self-reliability of VNFv∈V are expressed as Ty v and R v ; The bandwidth resource requirement of the virtual link (u,v)∈L is expressed as

[0103] Step 3: Network slice pre-deployment modeling;

[0104] Network slicing pre-deployment modeling specifically involves: Network slice deployment is performed through virtual network embedding, using a non-coordinated approach, divided into virtual link embedding and virtual node embedding. To ensure end-to-end latency within the network slice, the number of hops between VNFs in the network slice must be minimized. Therefore, virtual network embedding is performed first. A K-shortest path algorithm is used within the physical network to find the shortest physical path that meets the virtual link resource requirements within the network slice. All physical nodes within the determined shortest path are then ranked by reliability, and the most reliable physical nodes are selected for optimal deployment of the primary VNF instance.

[0105] Step 4: Calculation and modeling of actual reliability of network slices;

[0106] The actual reliability calculation model of network slicing is as follows: According to Assumption 3, the actual reliability of VNF after being deployed to the physical node is determined by the reliability of the physical node hosting it and the reliability of the VNF itself. The actual reliability of VNFv∈V after pre-deployment is p v Expressed as

[0107]

[0108] in, Represents the reliability of the physical node carrying VNFv∈V.

[0109] There are two basic combinations of VNFs in URLLC end-to-end network slices: series and parallel. The reliability of a complex network slice is expressed as Modeled as:

[0110]

[0111] in Represents network slice s i Reliability of all serially connected VNFs in Represents network slice s i The reliability of all VNFs connected in parallel in .

[0112] The set of backup instances of VNFv∈V is expressed as The total set of VNFv∈V including the primary instance and backup instance is expressed as in Indicates the primary instance. According to assumption 3, It can be modeled as:

[0113]

[0114] Among them, V series represents the set of VNFs connected in series. According to Assumption 5, It can be modeled as:

[0115]

[0116] Among them, V paralled Represents a set of VNFs connected in parallel.

[0117] Step 5: Modeling of reliability allocation method;

[0118] The reliability allocation method is modeled as follows: Through the reliability allocation method, the reliability of the network slice is decomposed into the expected reliability of each VNF. Indicates the expected reliability R of the slice si The expected reliability of VNFv∈V under the requirement is modeled as:

[0119]

[0120] The reliability allocation method allocates the reliability requirements of the network slice to each VNF.

[0121] Step 6: Modeling the priority VNF backup sharing method;

[0122] Priority VNF backup sharing method modeling, specifically: There are generally two fault recovery schemes, proactive fault recovery (PFR) and reactive fault recovery (RFR). The fault recovery process consists of three main stages: (1) starting backup VNF and image migration; (2) flow reconfiguration; (3) state synchronization. The execution of each stage will generate considerable delay, which not only leads to network performance degradation, but also violates the service level agreement due to long service interruption time. Through fault prediction, PFR can reduce delay by starting some of the above stages in advance. The present invention performs backup sharing of VNFs in the network slice during the deployment stage, so it belongs to PFR. In addition, the betweenness centrality b of the modeled physical node n :

[0123]

[0124] Among them, σ st represents the number of shortest paths from physical node s to node t, σ st (n) represents the number of physical nodes n located on the shortest path from node s to node t.

[0125] Based on the betweenness centrality of the modeled physical nodes, all physical nodes in the physical network are prioritized. When a backup VNF instance is created and needs to be deployed, a node with a high priority is selected to deploy the backup VNF instance based on the physical node priority.

[0126] Step 7: Modeling the end-to-end latency of network slices;

[0127] The end-to-end delay modeling of network slices is as follows: VNFs generate node delays during operation, including processing delays and queuing delays. To avoid VNF overload and node queuing delays dominating the end-to-end delay, the service flow arrival rate of the network slice is limited to less than the VNF computing capacity. The modeling is as follows:

[0128] λ≤comp v

[0129] Where λ represents the service flow arrival rate of the network slice, comp v Indicates the computing capacity of the VNF node. Network slice s i The processing delay is expressed as Modeled as:

[0130]

[0131] where dt v Indicates the amount of data transmitted, comp v Represents the computing capability of VNFv∈V.

[0132] Network Slices i The transmission delay is expressed as Modeled as:

[0133]

[0134] Therefore, the end-to-end delay can be modeled as:

[0135]

[0136] Step 8: Modeling network slice resource consumption costs;

[0137] Network slicing resource consumption cost modeling, specifically: defining binary variables Indicates whether the primary instance (g=0) or backup instance (g=1,2,...,K) of VNFv∈V is mapped to the physical node n∈N, a binary variable Indicates whether the virtual link (u, v) is mapped to the physical link (m, n) ∈ E. Based on two binary variables, the total node resource consumption in the network slice is modeled as:

[0138]

[0139] The total link bandwidth resource consumption is modeled as:

[0140]

[0141] in, Indicates a virtual link Is it mapped to the physical link (m,n)∈E? Indicates a virtual link bandwidth resource requirement.

[0142] Total resource consumption Re of each network slice total is modeled as:

[0143] Re total = μ1C com + μ2B com

[0144] where μ1 and μ2 are weight coefficients of two types of resource consumption, respectively. Because different types of network slices have different requirements for node resource consumption and link bandwidth consumption, weight coefficients are used for adjustment.

[0145] Step 9: Modeling of network slice deployment and backup sharing restriction conditions;

[0146] Modeling of network slice deployment and backup sharing restriction conditions is as follows: defining binary variable to represent whether v e V shares a backup instance with u e V; and binary variable S(u, v) represents whether u e V and v e V are VNFs of the same type.

[0147] (1) Modeling of VNF backup instance sharing restriction conditions:

[0148]

[0149] (2) Modeling of shared backup instance resource restriction conditions:

[0150]

[0151] (3) Modeling of sharing quantity restriction conditions:

[0152]

[0153] (4) Modeling of VNF deployment restriction conditions:

[0154]

[0155] (5) Modeling of CPU resource restriction conditions:

[0156]

[0157] (6) Modeling of storage resource restriction conditions:

[0158]

[0159] (7) Modeling of bandwidth resource restriction conditions:

[0160]

[0161] (8) Modeling flow conservation constraints:

[0162]

[0163] (9) Modeling end-to-end delay constraints:

[0164]

[0165] (10) Modeling constraint integrity restrictions:

[0166]

[0167]

[0168] Those skilled in the art will understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program. The program can be stored in a computer-readable storage medium. When the program is executed, the steps of the method can be implemented. The storage medium, such as ROM / RAM, disk, optical disk, etc.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A network slice deployment method based on priority VNF backup sharing, characterized by: The following steps are involved: S1: Build the physical network topology model, URLLC end-to-end network slicing model, and network slicing pre-deployment model; S2: Calculate the actual reliability of the VNF after it is deployed on the physical node using the network slice actual reliability calculation model; S3: Decompose the reliability of the network slice into the expected reliability of each VNF through the reliability allocation method; Network slice pre-deployment modeling includes: deploying network slices through virtual network embedding, using a non-coordinated approach, including virtual link embedding and virtual node embedding; first, using the K-shortest path algorithm to find the shortest physical path in the physical network that meets the virtual link resource requirements of the network slice; then, for all physical nodes in the determined shortest physical path, sorting them according to their reliability, and selecting the most reliable physical node to deploy the primary VNF instance; Actual reliability calculation modeling of network slices, including: VNF Actual reliability after pre-deployment Expressed as: in, Indicates the hosting VNF The reliability of the physical nodes; The reliability of a complex network slice is expressed as , modeled as: in Represents network slicing Reliability of all serially connected VNFs in Represents network slicing Reliability of all VNFs connected in parallel; VNF The set of backup instances is represented as , VNF The total set of primary and backup instances is represented as ,in Represents the primary instance, Modeled as: in, Represents a set of VNFs connected in series; Modeled as: in, Represents a set of VNFs connected in parallel; Reliability allocation method modeling, including: Indicates the expected reliability of the slice VNF under requirements The expected reliability of is modeled as: The reliability allocation method allocates the reliability requirements of the network slice to each VNF; End-to-end latency modeling for network slicing involves limiting the service flow arrival rate of the network slice to less than the VNF computing capacity. The modeling is as follows: in represents the service flow arrival rate of the network slice, Indicates the computing capability of the VNF node; Network slicing The processing delay is expressed as , modeled as: in Indicates the amount of data transmitted. Indicates VNF computing power; Network slicing The transmission delay is expressed as , modeled as: Therefore, the end-to-end delay is modeled as: Network slicing resource consumption cost modeling, including: defining binary variables Indicates VNF Whether the primary or backup instance is mapped to a physical node When g = 0, it represents the primary instance, and when g = 1, 2, ..., K, it represents the backup instance; binary variable Indicates a virtual link Whether to map to a physical link Above, based on two binary variables, the total node resource consumption in the network slice is modeled as: The total link bandwidth resource consumption is modeled as: in, Indicates a virtual link Whether to map to a physical link superior, Indicates a virtual link Bandwidth resource requirements; Total resource consumption per network slice Modeled as: in, and are the weight coefficients of the two resource consumptions; S4: Back up and share the VNFs in the network slice using a priority VNF backup sharing method. The priority VNF backup sharing method in step S4 includes: Modeling betweenness centrality of physical nodes : in, Indicates that from the physical node To Node The number of shortest paths, Represents a physical node Located at the node To Node The number of shortest paths; Based on the betweenness centrality of the modeled physical nodes, all physical nodes in the physical network are prioritized. When a backup VNF instance is created and needs to be deployed, the node with the highest priority is selected based on the physical node priority to deploy the backup VNF instance. Modeling network slicing end-to-end delay, network slicing resource consumption cost model, network slicing deployment and backup sharing constraints; the network slicing deployment and backup sharing constraints specifically include: let the binary variable express Whether Shared backup instance, let binary variable express and Whether they are VNFs of the same type; (1) Modeling VNF backup instance sharing constraints: (2) Modeling shared backup instance resource constraints: (3) Modeling sharing quantity restrictions: (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 constraint integrity restrictions: 。 2. The network slice deployment method based on priority VNF backup sharing according to claim 1 is characterized in that: The physical network topology model is constructed as follows: Let the undirected weighted graph Represents the physical network topology, where Represents the set of physical nodes in the physical network, ; Represents the set of physical links in the physical network, ; make Represents the CPU resource capacity of the physical node, let Represents the storage resource capacity of the physical node, let Represents the reliability of the physical node, let Indicates a physical link bandwidth resource capacity.

3. The network slice deployment method based on priority VNF backup sharing according to claim 1 is characterized in that: The following assumptions exist: The VNFs in each network slice must be deployed on different physical nodes. The reliability of each physical node is a known constant; The reliability of a VNF depends on the reliability of the physical node that hosts it and the reliability of the VNF itself. VNF failures are independent and not affected by other physical nodes or VNFs. The link bandwidth resource requirements in each network slice and the CPU resource requirements and storage resource requirements of the VNF are independent random variables and have an exponential distribution; The reliability of the VNF backup instance is consistent with that of the primary instance; Network slice requests arrive at the network sequentially; The URLLC end-to-end network slicing model is constructed as follows: make Denote the set of network slices, let Represents network slicing The expected reliability of Denote the end-to-end delay requirement of the network slice; let the undirected weighted graph represents the URLLC end-to-end network slice, where Represents the VNF set in the network slice, , Represents a set of virtual links, ;make express The CPU resource requirements of express The storage resource requirements of express VNF category, let express The reliability of itself; Indicates a virtual link bandwidth resource requirements.

Images