Method, apparatus, device, medium and program product for reconstructing cross-domain network slices

Abstract

The application relates to the technical field of communication, in particular to a cross-domain network slice reconstruction method and device, equipment, medium and program product. The method comprises the following steps: calculating the importance value of each node in the current network slice demand according to a pre-constructed network slice demand model, and generating a node mapping sequence between nodes; combining the network slice demand model and the node mapping sequence, searching for nodes in the physical node set that simultaneously satisfy the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain an initial candidate physical node set; performing node mapping, and when the node mapping fails, activating a slice reconstruction mechanism to generate a candidate migration node set; and obtaining a complete migration physical node set of the current network slice demand by using the candidate migration node set, so as to realize slice reconstruction. Therefore, the problems in the related art that the success rate of the network slice is low, and along with the increase in the number of slices, the calculation efficiency of the network slice deployment is low, and the time consumption is long and the like are solved.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method, apparatus, device, medium, and program product for reconstructing cross-domain network slices. Background Technology

[0002] The deployment and maintenance of network slices involves the scheduling of network virtual resources, and successful deployment is a prerequisite for network slice service providers to generate revenue. However, during the deployment or maintenance process, network slice deployment may sometimes fail due to changes in business needs or network resource congestion. In addition, if the scheduling of network virtual resources remains unchanged, subsequent network slice deployments may also fail due to insufficient resources.

[0003] In related technologies, to cope with changes in the needs of network slicing services, in addition to adjusting the network slicing deployment strategy, slice reconstruction can also be performed after network slice mapping failure. It is understandable that in network slice collaborative deployment scenarios, network slice deployment failure often occurs during the node mapping stage of slice deployment. Therefore, slice reconstruction needs to clear the physical nodes causing resource congestion and migrate the virtual nodes hosted on the congested nodes back to physical nodes with lower loads to facilitate the hosting of subsequent virtual nodes.

[0004] However, the related technologies only consider the virtual node migration cost of the slice and cannot achieve load balancing of the physical underlying network, resulting in a low success rate of network slicing. In addition, as the number of slices increases, the computational efficiency of network slice deployment becomes low and the time consumption is long, and the robustness of the slice deployment system is poor, which urgently needs to be improved. Summary of the Invention

[0005] This application provides a method, apparatus, device, medium, and program product for reconstructing cross-domain network slices, in order to solve the problems in related technologies that only consider the virtual node migration cost of slices, which cannot achieve load balancing of the physical underlying network, resulting in a low success rate of network slices. In addition, as the number of slices increases, the computational efficiency of network slice deployment becomes low, the time consumption is long, and the robustness of the slice deployment system is poor.

[0006] A first aspect of this application provides a method for reconstructing cross-domain network slices. The method includes the following steps: obtaining all network slice requirements in a cross-domain network, and calculating the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement based on a pre-constructed network slice requirement model, so as to obtain a node mapping sequence between the virtual node and the physical node based on the virtual importance value and the physical importance value; combining the network slice requirement model and the node mapping sequence, searching for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain an initial candidate physical node set corresponding to the current network slice requirement based on the at least one initial candidate physical node; and performing the reconstruction of the virtual node and the physical node based on the initial candidate physical node set. The process involves initial node mapping between candidate physical nodes. If node mapping fails, the congested and overloaded physical node sets in the current network slice requirement are obtained. A slice reconstruction mechanism is then used to remap the nodes, resulting in a candidate migration physical node set and a candidate migration virtual node set. The slice reconstruction mechanism includes generating a reconstructed physical node set for the current network slice requirement using the congested and overloaded physical node sets, determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set, obtaining a complete migration-in physical node set for the current network slice requirement using the candidate migration physical node set and the candidate migration virtual node set, and redeploying the current network slice requirement based on the complete migration-in physical node set, thereby reconstructing the current network slice requirement.

[0007] Optionally, in one embodiment of this application, the step of combining the network slicing requirement model and the node mapping sequence to search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on the at least one initial candidate physical node, includes: combining the network slicing requirement model and the node mapping sequence to filter out the candidate physical node set that satisfies the virtual node mapping condition from the physical node set; combining the virtual node and the cross-domain network domain constraint condition to obtain the at least one initial candidate physical node using the candidate physical node set, and using the at least one initial candidate physical node as the root node to generate the initial candidate physical node set.

[0008] Optionally, in one embodiment of this application, determining the candidate migration physical node set and candidate migration virtual node set for the current network slice requirement based on the reconstructed physical node set includes: calculating the resource congestion level of each reconstructed physical node in the reconstructed physical node set using node resource differences and physical link utilization; sorting all reconstructed physical nodes in the reconstructed physical node set according to the resource congestion level to obtain a candidate migration physical node set where the reconstructed physical nodes meet a preset order; determining whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet preset migration conditions; if the mapped virtual nodes meet the preset migration conditions, then using the mapped virtual nodes to form an initial candidate migration virtual node set to obtain the candidate migration virtual node set, and obtaining a candidate migration virtual link set and a candidate migration physical link set based on preset link requirements.

[0009] Optionally, in one embodiment of this application, obtaining the complete set of incoming physical nodes for the current network slice requirement using the candidate set of incoming physical nodes and the candidate set of incoming virtual nodes includes: obtaining the bearer physical nodes mapped to the candidate incoming virtual nodes using the candidate incoming virtual nodes in the candidate set of incoming virtual nodes; determining the set of neighboring physical nodes corresponding to the candidate incoming virtual nodes based on the bearer physical nodes and the candidate set of incoming physical nodes; constructing a disruption operator based on the set of neighboring physical nodes, and calculating the disrupted node set and the initial set of neighboring candidate physical nodes corresponding to the set of neighboring physical nodes using the disruption operator; constructing a repair operator based on the disrupted node set, and repairing the initial set of neighboring candidate physical nodes using the repair operator to obtain the set of neighboring candidate physical nodes; calculating the migration burden of each neighboring candidate physical node in the set of neighboring candidate physical nodes, so as to generate the complete set of incoming physical nodes based on the migration burden, wherein the migration burden may include, but is not limited to, consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption.

[0010] Optionally, in one embodiment of this application, the calculation formula for the complete set of migrating physical nodes may be, but is not limited to, the following:

[0011] ,

[0012] in, It was through The set of destroyed nodes after round-neighbor iterative search. It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The set of repaired neighborhood physical nodes after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

[0013] A second aspect of this application provides a cross-domain network slice reconstruction apparatus, the apparatus comprising: an acquisition module, configured to acquire all network slice requirements in a cross-domain network, and calculate the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement according to a pre-built network slice requirement model, so as to obtain a node mapping sequence between the virtual node and the physical node based on the virtual importance value and the physical importance value; a generation module, configured to combine the network slice requirement model and the node mapping sequence to search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain an initial candidate physical node set corresponding to the current network slice requirement based on the at least one initial candidate physical node; and a determination module, configured to perform virtual node reconstruction based on the initial candidate physical node set. The system performs node mapping between the proposed node and the initial candidate physical node. If node mapping fails, it obtains the congested physical node set and the overloaded physical node set from the current network slice requirement. A slice reconstruction mechanism is then used to remap the nodes to obtain a candidate migration physical node set and a candidate migration virtual node set. The slice reconstruction mechanism includes generating a reconstructed physical node set for the current network slice requirement using the congested physical node set and the overloaded physical node set, and determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set. A reconstruction module is used to obtain a complete migration-in physical node set for the current network slice requirement using the candidate migration physical node set and the candidate migration virtual node set, and then redeploys the current network slice requirement based on the complete migration-in physical node set, thereby reconstructing the current network slice requirement.

[0014] Optionally, in one embodiment of this application, the generation module includes: a filtering unit, configured to combine the network slicing requirement model and the node mapping sequence to filter out a set of candidate physical nodes that meet the virtual node mapping conditions from the set of physical nodes; and a first generation unit, configured to combine the virtual nodes and the cross-domain network domain constraints to obtain the at least one initial candidate physical node using the set of candidate physical nodes, and to generate the initial candidate physical node set using the at least one initial candidate physical node as the root node.

[0015] Optionally, in one embodiment of this application, the determining module includes: a first calculation unit, configured to calculate the resource congestion level of each reconstructed physical node in the reconstructed physical node set using node resource differences and physical link utilization; a sorting unit, configured to sort all reconstructed physical nodes in the reconstructed physical node set according to the resource congestion level to obtain a set of candidate migration physical nodes that meet a preset order; a judging unit, configured to judge whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet preset migration conditions; and a determining unit, configured to, when the mapped virtual nodes meet the preset migration conditions, use the mapped virtual nodes to form an initial set of candidate migration virtual nodes to obtain the set of candidate migration virtual nodes, and obtain a set of candidate migration virtual links and a set of candidate migration physical links based on preset link requirements.

[0016] Optionally, in one embodiment of this application, the reconstruction module includes: a second generation unit, configured to obtain the bearer physical node mapped to the candidate migration virtual node using the candidate migration virtual node set; a third generation unit, configured to determine the neighboring physical node set corresponding to the candidate migration virtual node based on the bearer physical node and the candidate migration physical node set; a first construction unit, configured to construct a destruction operator based on the neighboring physical node set, and calculate the destruction node set and the initial neighboring candidate physical node set corresponding to the neighboring physical node set through the destruction operator; a second construction unit, configured to construct a repair operator based on the destruction node set, and repair the initial neighboring candidate physical node set through the repair operator to obtain the neighboring candidate physical node set; and a second calculation unit, configured to calculate the migration burden of each neighboring candidate physical node in the neighboring candidate physical node set, so as to generate the complete migration-in physical node set based on the migration burden, wherein the migration burden may include, but is not limited to, consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption.

[0017] Optionally, in one embodiment of this application, the calculation formula for the complete set of migrating physical nodes may be, but is not limited to, the following:

[0018] ,

[0019] in, It was through The set of destroyed nodes after round-neighbor iterative search. It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The set of repaired neighborhood physical nodes after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

[0020] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the cross-domain network slice reconstruction method as described in the above embodiments.

[0021] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the cross-domain network slice reconstruction method described above.

[0022] A fifth aspect of this application provides a computer program product, including a computer program that, when executed, implements the cross-domain network slice reconstruction method described above.

[0023] This application embodiment evaluates the importance of nodes required for network slicing, obtains a node mapping sequence between nodes, and triggers a network slice reconstruction mechanism when node mapping fails. This determines the candidate migration node set for the current network slice requirement, and then obtains the complete set of physical nodes to be migrated for the current network slice requirement, thus achieving network slice reconstruction. This not only balances network load and reduces migration costs in cross-domain network slice deployment scenarios, but also effectively improves the efficiency of neighborhood search and slice reconstruction. Therefore, it solves the problems in related technologies that only consider the virtual node migration cost of slices, failing to achieve load balancing of the physical underlying network, resulting in a low success rate for network slicing. Furthermore, as the number of slices increases, the computational efficiency of network slice deployment decreases, the time consumption increases, and the robustness of the slice deployment system becomes poor.

[0024] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0025] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0026] Figure 1 A block diagram illustrating a cross-domain network slice reconstruction system is provided according to one embodiment of this application;

[0027] Figure 2 This is a flowchart of a cross-domain network slice reconstruction method provided according to an embodiment of this application;

[0028] Figure 3This is a flowchart of a node importance assessment mechanism based on an embodiment of the present application, namely, the AEC (Adversarial Extract Champion Method) - VIKOR (VlseKriterijumska Optimizacija I KompromisnoResenje) mechanism.

[0029] Figure 4 This is a flowchart of a breadth-first search algorithm provided according to an embodiment of this application;

[0030] Figure 5 This is a flowchart illustrating the determination of a candidate migration node set based on a reconstructed physical node set according to an embodiment of this application;

[0031] Figure 6 A flowchart illustrating the generation of a complete set of incoming physical nodes based on the ALNS (Adaptive Large Neighborhood Search) algorithm according to an embodiment of this application;

[0032] Figure 7 A flowchart illustrating the working principle of a cross-domain network slice reconstruction method according to an embodiment of this application;

[0033] Figure 8 This is a block diagram of a cross-domain network slice reconstruction apparatus provided according to an embodiment of this application;

[0034] Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0035] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0036] The following describes a method, apparatus, device, medium, and program product for reconstructing cross-domain network slices according to embodiments of this application, with reference to the accompanying drawings. Addressing the issues mentioned in the background art, which only consider the virtual node migration cost of slices, failing to achieve load balancing of the physical underlying network and resulting in a low success rate for network slices, and further contributing to the low computational efficiency and long deployment time of network slices as the number of slices increases, leading to poor robustness of the slice deployment system, this application provides a method for reconstructing cross-domain network slices. In this method, by evaluating the importance of nodes required for network slices, a node mapping sequence between nodes is obtained. When node mapping fails, a network slice reconstruction mechanism is triggered to determine the candidate migration node set for the current network slice requirement, thereby obtaining the complete set of physical nodes to be migrated for the current network slice requirement, thus achieving network slice reconstruction. This not only balances network load and reduces migration costs in cross-domain network slice deployment scenarios but also effectively improves the efficiency of neighborhood search and slice reconstruction. This solves the problems in related technologies that only consider the virtual node migration cost of slices, which cannot achieve load balancing of the physical underlying network, resulting in a low success rate of network slicing. In addition, as the number of slices increases, the computational efficiency of network slice deployment becomes low, the time consumption is long, and the robustness of the slice deployment system is poor.

[0037] Before explaining the cross-domain network slice reconstruction method proposed in the embodiments of this application, the cross-domain network slice reconstruction system involved in the embodiments of this application will be introduced first.

[0038] Specifically, Figure 1 A block diagram of a cross-domain network slice reconstruction system is provided according to one embodiment of this application.

[0039] like Figure 1 As shown in the embodiment of this application, in the cross-domain network slice deployment and reconstruction system, network slice users submit network slice service requests. A network slice requirement model is instantiated according to the network slice configuration file information. The instantiated network slice requirements are sorted in a greedy priority order to form a network slice requirement sequence. Then, the network slice requirements are reordered using the slice configuration information and slice resource information in the sequence, and slice deployment priorities are assigned. Network slices are deployed according to these priorities. If slice deployment fails, a slice reconstruction mechanism is triggered. This involves calling the node neighborhood search module for the failed slice to complete the collaborative migration and reconstruction of slice nodes and links. By comprehensively considering the slice priority evaluation strategy, the network slice requirement sequence is optimized, and the instantiation, orchestration, and deployment of slices are completed on the physical infrastructure network.

[0040] It is understood that the network slicing in this application embodiment is a logically isolated network deployed on the physical infrastructure network. The network slicing deployment process involves the allocation of slice resources and requires obtaining the abstracted physical network topology information of each network domain. Therefore, this application embodiment, based on graph theory, transforms the basic network resources and physical network topology information into an undirected weighted graph, which can be represented, but is not limited to, as follows: ,in, Represents a set of physical nodes, which are network functional units such as switches and routers in the physical infrastructure network. A single physical network node in the physical node set. use express, In cross-domain network infrastructure, this application embodiment also considers the network domain constraints of the physical network. In this application embodiment, the physical network includes two network domains, so the corresponding physical node sets also belong to two physical network domains, A and B, respectively. These are managed by different slice agents, and an adjacency matrix is ​​introduced. This represents the connection relationships between nodes and the connection status of physical links in a physical network. Each physical link is represented as... That is, connecting physical network nodes and physical network nodes physical link, , This refers to the computing resources that a physical node can provide. The total computing resources that can be provided are expressed as , Indicates the bandwidth resources that a physical link can provide; physical link The total bandwidth resources that can be provided are expressed as , This indicates the latency of the physical link. The total transmission delay is expressed as It should be noted that the physical network infrastructure in this application embodiment may include, but is not limited to, node computing resources, link bandwidth resources, link transmission latency, etc., and this application does not impose specific limitations.

[0041] Furthermore, the network slicing requirements in this embodiment are virtual network topologies instantiated by the slicing service provider based on the service configuration files provided by the slicing service users, similar to the modeling of the basic physical network infrastructure topology. Therefore, this embodiment can use an undirected weighted graph to represent the virtual network topology, which can be, but is not limited to, represented as: ,in, This represents a set of virtual topology nodes. A virtual node represents a virtual network functional unit in a network slice. Each set of virtual nodes contains a single virtual node. use express, Introducing an adjacency matrix This represents the connection relationship between virtual nodes in a virtual network topology. and virtual nodes The virtual link is represented as , , This is the amount of computing resources required by the virtual node. Represents virtual nodes Required computing resources This refers to the bandwidth resources required by the virtual link. Represents virtual nodes and virtual nodes The bandwidth resources required for the virtual links between them This indicates the maximum end-to-end link transmission delay allowed by network slicing. Indicates link The maximum allowable transmission delay is This indicates the number of network topology requirements.

[0042] For example, in this embodiment of the application, the network slice demand sequence arriving at the slice service provider will execute the AEC-VIKOR mechanism to evaluate the virtual node set topology, and then perform node mapping according to the load-balanced multi-dimensional cross-domain network slice deployment algorithm. If the node mapping process fails, the slice node reconstruction mechanism will be triggered to filter the migration virtual node set, and then the node neighborhood search module will call the node neighborhood search algorithm based on the ALNS method to search for the target migration physical node set. Finally, the collaborative deployment module will call the CLBC (Completed Local Binary Count) algorithm to collaboratively map the virtual nodes and links of the network topology requirements, and complete the network slice reconstruction process.

[0043] Specifically, Figure 2 This is a flowchart of a cross-domain network slice reconstruction method provided according to an embodiment of this application.

[0044] like Figure 2 As shown, the method for reconstructing this cross-domain network slice includes the following steps:

[0045] In step S201, all network slice requirements in the cross-domain network are obtained, and the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement are calculated according to the pre-built network slice requirement model, so as to obtain the node mapping sequence between virtual nodes and physical nodes based on the virtual importance value and the physical importance value.

[0046] It is understood that the embodiments of this application can obtain network slicing requirements from different domains (which can be understood as, but is not limited to, different geographical locations, operator networks, or technology platforms, etc., and this application does not impose specific limitations). Among them, network slicing requirements can include, but are not limited to, virtual network functions, lists and their interconnection relationships, QoS (Quality of Service) indicators, etc., and this application does not impose specific limitations.

[0047] In addition, embodiments of this application can represent the obtained network slice requirements by pre-constructing a network slice requirement model.

[0048] As one possible implementation, embodiments of this application can utilize the AEC-VIKOR mechanism to evaluate all virtual nodes and physical nodes in the current network slice requirement subnet, to obtain the virtual importance value of each virtual node and the physical importance value of each physical node, and then use a breadth-first search algorithm to traverse and sort the virtual nodes in the current network slice requirement topology based on the virtual importance value and physical importance, generating a node mapping sequence between virtual nodes and physical nodes.

[0049] For example, this application embodiment comprehensively considers the topological attributes and resource attributes of nodes, and evaluates and ranks the importance of nodes during the node mapping stage. It should be noted that this application embodiment can utilize the AEC-VIKOR mechanism to evaluate node importance, and its process is as follows: Figure 3 As shown, the specific content is as follows:

[0050] Step S301: Greedily sort all network slice demand subnets based on QoS metrics.

[0051] It is understood that the distribution of queues in the network controller of this application embodiment is determined by the service mode and QoS indicators, assuming that the set of arriving network slice requests is... Calculate the QoS index of its resource requirements, which can be expressed, but is not limited to, as:

[0052] ,

[0053] in, This indicates the total computing resources required by the nodes for network slicing. This indicates the total bandwidth required for the network slicing link. The weighted parameter represents the relative weight of link bandwidth and node resources. This indicates connecting two virtual nodes. and Link ,if and If there is a link requirement, then ,on the contrary, .

[0054] Therefore, the embodiments of this application can be based on The numerical values ​​are greedily sorted from largest to smallest to enhance the priority and QoS requirements of slices and improve the slice success rate.

[0055] Step S302: Generate the normalized decision matrix.

[0056] In this embodiment, the network slice requirement slicing process is executed one by one according to the QoS greedy sorting result, and the following settings are made: For the first v The resource view of each network slice requirement is used to obtain a virtual normalized decision matrix by utilizing the importance attribute values ​​of virtual nodes. It can be, but is not limited to, expressed as:

[0057] ,

[0058] ,

[0059] in, N This indicates the number of nodes in the network topology. L This indicates the number of evaluation metrics for a node. Indicates the first i virtual nodes Relative evaluation index standards j The value, This represents the standard weighting coefficient, which ranges from [0,1], and the sum of all weighting coefficients is 1. Wherein, the standard... j Including the resource importance and topological importance of nodes, setting L Individual node attributes may include, but are not limited to, the node's resource importance value. Degree centrality value of a node The density centrality value of nodes Betweenness centrality B C eigenvector centrality E C Primitive analysis, etc. G A This application does not impose specific restrictions.

[0060] Step S303: Calculate the expected value and defect value of the node.

[0061] In the embodiments of this application, virtual nodes are calculated respectively. Expected value and shortcomings The formulas for calculating the expected value and defect value of virtual nodes can be, but are not limited to, expressed as:

[0062] ,

[0063] in, and They are The positive and negative ideal solutions.

[0064] Step S304: Use the intercept method to fit and calculate the compromise coefficient value.

[0065] In this embodiment of the application, it can be made ,but Thus, the compromise coefficient is derived. The value of can be calculated using, but is not limited to, the formula as follows:

[0066] ,

[0067] Step S305: Calculate the importance value of the node.

[0068] It is understood that, in the embodiments of this application, each virtual node can be calculated based on the obtained expected value, defect value, and compromise coefficient value. Corresponding importance value It can be, but is not limited to, expressed as:

[0069]

[0070] in, This is the compromise coefficient.

[0071] Step S306: Sort the nodes by importance to obtain the final node sequence.

[0072] In this embodiment of the application, the decision interval can be divided according to the calculated compromise coefficient value, the importance value of each decision interval can be listed and sorted from small to large, cluster feature analysis can be performed, and the node with the highest probability in each decision interval can be selected as the feature node of each interval, and the final node ranking of the virtual nodes can be listed.

[0073] Step S307: Repeat the above steps to generate the final node sequence of physical nodes.

[0074] It is understood that, according to the topology information of the acquired physical network topology, the embodiments of this application can, for each physical node in the basic physical network topology, Repeat steps S302-306 to calculate each physical node in the physical network topology. Importance value It can be, but is not limited to, expressed as:

[0075] ,

[0076] Furthermore, after obtaining the virtual importance value and the physical importance value, the embodiments of this application can use the breadth-first search algorithm to arrange the virtual network topology nodes of the network demand model to obtain the node mapping sequence in the node mapping stage.

[0077] The flowchart of the breadth-first search algorithm in this application is as follows: Figure 4 As shown, the specific description is: Find the sorted sequence with the minimum importance value. virtual nodes As the root node of the breadth-first search tree, the root node is placed at the beginning of the queue; the breadth-first search algorithm is then used to traverse the tree. Topology: Search each layer of the current network slice requirement, invoke the AEC-VIKOR mechanism, and determine the slice based on importance value. Sort the nodes in ascending order by size, return the sorting result to the virtual topology node set, retrieve each node from the queue, determine whether the node is an unchecked child node, if so, add the child node to the queue, until all layers have been executed, and output the final queue result.

[0078] In step S102, combining the network slicing requirement model and the node mapping sequence, at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition is searched, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on at least one initial candidate physical node.

[0079] As one possible implementation method, this application embodiment searches for a set of candidate physical nodes in the physical node set that meet the virtual node mapping conditions of the current network slice requirements based on the network configuration information of the current network slice requirements in the network slice requirement model and the node mapping sequence. If the candidate physical nodes in the set of candidate physical nodes meet the cross-domain network domain constraint conditions, then the candidate physical nodes are used as the initial candidate physical nodes in the initial candidate physical node set, thereby obtaining the initial candidate physical node set.

[0080] In this embodiment, based on the current virtual node type, physical nodes that meet the virtual node mapping conditions (e.g., node computing resource requirements, which are not specifically limited in this application) can be selected from the candidate physical node set for mapping. Specifically, in this embodiment, if the current virtual node is the root node of the breadth-first search tree, the virtual node is mapped to the physical node with the smallest load; if the current virtual node is not the root node, the virtual node is mapped to a node in the set of the intersection neighbors of the virtual parent node carrying the virtual node and the set of physical nodes directly connected to the virtual node. The node load calculation for selecting the mapped node selects the physical node with the smallest load in the candidate physical node set.

[0081] Furthermore, embodiments of this application can utilize an initial set of candidate physical nodes to perform node mapping between virtual nodes and initial candidate physical nodes. If node mapping is successful, the KSP (K-Shortest Paths) algorithm is used to collaboratively map virtual links, where the link constraints consider link bandwidth as link latency, and the path selection strategy considers physical link load balancing.

[0082] Optionally, in one embodiment of this application, by combining the network slicing requirement model and the node mapping sequence, at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition is searched to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on at least one initial candidate physical node, including: combining the network slicing requirement model and the node mapping sequence to screen out the candidate physical node set that satisfies the virtual node mapping condition from the physical node set; combining the virtual node and the cross-domain network domain constraint condition to obtain at least one initial candidate physical node using the candidate physical node set, and using at least one initial candidate physical node as the root node to generate the initial candidate physical node set.

[0083] In some embodiments, this application combines a network slicing requirement model and a node mapping sequence to search for an initial candidate set of physical nodes that simultaneously satisfies the virtual node mapping condition and the cross-domain network constraint condition.

[0084] For example, in this embodiment of the application, based on the network configuration information of the current network slice requirement in the network slicing requirement model, the physical node set... Select virtual nodes that meet the current network configuration information. The set of candidate physical nodes for mapping conditions .

[0085] In this embodiment of the application, for each virtual node The initial candidate physical nodes must first satisfy cross-domain network constraints, which can be, but are not limited to, expressed as:

[0086] ,

[0087] in, It is a binary variable, when the physical node Satisfy virtual nodes The value is 1 when the geographic network domain constraint condition is met, and 0 otherwise.

[0088] If virtual node If it is the root node of the breadth-first search tree, then map it to the node with the lowest importance value that can be mapped. On a physical node, the computational resource constraints that a node can be mapped to can be, but are not limited to, represented as:

[0089] ,

[0090] in, It is a binary variable. A virtual node can only be mapped to one physical node, but a physical node can support multiple virtual nodes if the remaining resources meet the resource requirements.

[0091] If virtual node If it is not the root node, find the virtual parent node that carries the virtual node in the breadth-first search tree of the virtual node, as well as all nodes that have been mapped to the current virtual node. Directly connected virtual nodes, each virtual parent node is defined as Find each physical node it maps to, and find the set of all neighboring physical nodes directly connected to these physical nodes; this is the candidate physical node set. It is the set of physical nodes that satisfy the node mapping resource constraints in the intersection of these neighbor node sets.

[0092] Furthermore, in this embodiment of the application, the set of candidate physical nodes is... Each physical node in According to load value Sort the nodes in ascending order to obtain the initial candidate physical node set. The formula for calculating node load in this application embodiment can be, but is not limited to, expressed as:

[0093] ,

[0094] in, It is a node The remaining computing resources of the current physical resources reflect the resource occupancy of the nodes in the current state.

[0095] Furthermore, in this embodiment of the application, for the initial candidate physical node set Each initial candidate physical node in First, obtain the virtual node. and virtual parent node Virtual links between Find the virtual parent node Mapped physical nodes At physical nodes Physical nodes mapped to virtual parent nodes Find the virtual link that satisfies the requirements The set of candidate physical links for the requirements, whose link constraints, including bandwidth resource conditions and latency conditions, may be, but are not limited to:

[0096] ,

[0097] in, It is a binary variable, when the virtual link Mapping to physical link The value is 1 if the condition is met, and 0 otherwise. It is the end-to-end latency requirement.

[0098] In step S103, node mapping between virtual nodes and initial candidate physical nodes is performed based on the initial candidate physical node set. If node mapping fails, the congested physical node set and overloaded physical node set in the current network slice requirement are obtained. The node mapping is re-performed using a slice reconstruction mechanism to obtain a candidate migration physical node set and a candidate migration virtual node set. The slice reconstruction mechanism includes generating a reconstructed physical node set for the current network slice requirement using the congested physical node set and the overloaded physical node set, and determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set.

[0099] As can be seen from the above analysis, the embodiments of this application can perform node mapping between virtual nodes and initial candidate physical nodes based on the initial candidate physical node set. If the node mapping between the virtual node and the initial candidate physical node fails, the slice reconstruction mechanism will be triggered. Node migration search is a key step in the slice reconstruction mechanism. Determining the candidate migration physical node set and the candidate migration virtual node set is the first step in node migration. The embodiments of this application comprehensively consider the node migration cost and the resource congestion degree of physical network nodes to determine the candidate migration physical node set and the candidate migration virtual node set.

[0100] It is understood that the embodiments of this application trigger a slice reconstruction mechanism to reassess and update the status of the congested physical node set and the overloaded physical node set, and search for the congested physical node set causing resource congestion in the current physical infrastructure network. The virtual nodes of the network slice requirement topology have been traversed and sorted using a breadth-first search algorithm. If a node mapping failure occurs at the root virtual node of the current network slice requirement, then the set of congested physical nodes will be affected. This is the set of physical nodes in the current physical infrastructure network that do not meet the computational resource requirements of this virtual node. If the node mapping failure occurs at a non-root virtual node in the current network slice requirement, then the set of congested physical nodes is considered congested. For the virtual parent node that hosts the virtual node and the various physical nodes that are directly connected to and mapped to it. The set of physical nodes consisting of the intersection of adjacent nodes.

[0101] Furthermore, embodiments of this application can search for sets of overloaded physical nodes with resource overload in the current physical infrastructure network. Calculate the node load of each physical node carrying virtual nodes in the current physical infrastructure network. If the load exceeds the node workload threshold... If the physical node is found to be overloaded, it will be included in the set of overloaded physical nodes. In this embodiment, the formula for calculating the node workload threshold can be, but is not limited to, expressed as:

[0102] ,

[0103] in, Physical nodes The current remaining computing resources.

[0104] Furthermore, in this embodiment of the application, the congested physical node set and the overloaded physical node set are combined into a reconstructed physical node set. .

[0105] Optionally, in one embodiment of this application, determining the candidate migration physical node set and candidate migration virtual node set for the current network slice requirement based on the reconstructed physical node set includes: calculating the resource congestion level of each reconstructed physical node in the reconstructed physical node set using node resource differences and physical link utilization; sorting all reconstructed physical nodes in the reconstructed physical node set according to their resource congestion levels to obtain a candidate migration physical node set whose reconstructed physical nodes meet a preset order; determining whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet preset migration conditions; if the mapped virtual nodes meet the preset migration conditions, then using the mapped virtual nodes to form an initial candidate migration virtual node set to obtain a candidate migration virtual node set, and obtaining a candidate migration virtual link set and a candidate migration physical link set based on preset link requirements.

[0106] In some embodiments, the process of determining the candidate set of physical nodes and the candidate set of virtual nodes for migration required for the current network slice based on the reconstructed physical node set is as follows: Figure 5 As shown, these mapped virtual nodes can be directly added to the candidate migration physical node set. This process includes evaluating the performance and resource matching of each mapped virtual node to ensure that all selected mapped virtual nodes can effectively support the required network service quality and resource requirements. Specifically:

[0107] Step S501: Calculate the resource congestion level of each reconstructed physical node in the reconstructed physical node set.

[0108] In this embodiment, the resource congestion level of each reconstructed physical node in the reconstructed physical node set is calculated. There are two main factors to consider, specifically:

[0109] (1) Node resource difference can be understood as the virtual nodes currently deployed in this application embodiment. The difference between the required node computing resources and the allocated computing resources of the target physical node and the total node resources that the physical node can provide. The larger the node resource difference, the more difficult it is for the remaining resources of this physical node to meet the resource requirements of the virtual node. The formula for calculating the node resource difference in this embodiment can be, but is not limited to, expressed as:

[0110] ,

[0111] (2) Physical link workload can be understood as the workload of the physical link connected to the current physical node when evaluating the congestion level of the physical node in this application embodiment.

[0112] At this point, the embodiments of this application can calculate the resource congestion level of each reconstructed physical node in the reconstructed physical node set, and the calculation formula can be, but is not limited to, expressed as:

[0113] ,

[0114] in, It is related to physical nodes The total number of connected physical links. It is the physical link load.

[0115] Step S502: Generate a set of candidate physical nodes for migration that satisfy a certain order.

[0116] In this embodiment, each reconstructed physical node in the reconstructed physical node set is sorted from highest to lowest resource congestion level, and the order of the reconstructed physical nodes in the reconstructed physical node set is updated to obtain a candidate migration physical node set that satisfies a certain order. This certain order can be set by those skilled in the art according to actual circumstances, and this application does not impose specific limitations.

[0117] Step S503: Determine whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet certain migration conditions.

[0118] It is understood that the virtual nodes carried by each candidate migration physical node in the candidate migration physical node set are retrieved in this application embodiment. The candidate migration virtual nodes in the candidate migration virtual node set should meet certain migration conditions, which may include migration conditions and migration cost conditions. The specific conditions can be set by those skilled in the art according to the actual situation, and this application does not impose specific restrictions.

[0119] In this embodiment, the migration condition can be understood as the computing resource requirements that can be met after migration to satisfy the deployment needs of the currently failed virtual node. For the initial candidate virtual node to migrate... The formula for calculating the migration condition can be, but is not limited to, expressed as:

[0120] ,

[0121] When the computing resource requirement of a virtual node is greater than the difference between the required node computing resources and the allocated computing resources of the target physical node and the total node resources that the physical node can provide, it means that migrating the current virtual node can at least guarantee the successful deployment of the virtual node that failed to be deployed, so as to obtain an initial set of candidate virtual nodes to be migrated that meet the migration conditions.

[0122] Migration costs can be understood as the expense of resource scheduling, which may include, but is not limited to, the computing resources spent deploying virtual nodes and the bandwidth resources occupied by virtual links directly connected to the virtual nodes. This application does not impose specific limitations. In this application's embodiments, the virtual nodes... The formula for calculating migration costs can be, but is not limited to, expressed as:

[0123] ,

[0124] Step S504: Determine the candidate set of virtual nodes for migration.

[0125] Furthermore, in this embodiment, the initial candidate virtual node set for migration is reordered according to the migration cost from lowest to highest to obtain a candidate virtual node set for migration. Then, based on certain link requirements, the direct links of all nodes are selected as a candidate virtual link set for migration, thereby generating a candidate physical link set for migration. The specific link requirements are set by those skilled in the art according to actual conditions, and this application does not impose any particular limitations.

[0126] In step S104, the complete set of physical nodes to migrate for the current network slice requirement is obtained by using the candidate set of physical nodes to migrate and the candidate set of virtual nodes to migrate, so as to redeploy the current network slice requirement according to the complete set of physical nodes to migrate, thereby realizing the reconstruction of the current network slice requirement.

[0127] As one possible implementation, after obtaining the candidate set of physical nodes and the candidate set of virtual nodes for migration, this application embodiment can obtain a complete set of incoming physical nodes after a certain number of iterative searches by executing an ALNS-based algorithm, thereby completing the reconstruction of the current network slicing requirements. The "certain number" can be set by those skilled in the art according to actual circumstances, and this application does not impose specific limitations.

[0128] Optionally, in one embodiment of this application, obtaining a complete set of incoming physical nodes for the current network slice requirement using a set of candidate migration physical nodes and a set of candidate migration virtual nodes includes: obtaining the bearer physical nodes mapped to the candidate migration virtual nodes using the candidate migration virtual nodes in the set of candidate migration virtual nodes; determining the neighboring physical node set corresponding to the candidate migration virtual nodes based on the bearer physical nodes and the set of candidate migration physical nodes; constructing a disruption operator based on the neighboring physical node set, and calculating the disruption node set and the initial neighboring candidate physical node set corresponding to the neighboring physical node set using the disruption operator; constructing a repair operator based on the disruption node set, and repairing the initial neighboring candidate physical node set using the repair operator to obtain the neighboring candidate physical node set; calculating the migration burden of each neighboring candidate physical node in the neighboring candidate physical node set, so as to generate a complete set of incoming physical nodes based on the migration burden, wherein the migration burden may include, but is not limited to, the consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption. The calculation formula for the complete set of incoming physical nodes may be, but is not limited to, as follows:

[0129] ,

[0130] in, It was through The set of destroyed nodes after round-neighbor iterative search. It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The set of repaired neighborhood physical nodes after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

[0131] As one possible implementation, the process of generating a complete set of migrating physical nodes based on the ALNS algorithm in this application embodiment is as follows: Figure 6 As shown, the process can be as follows:

[0132] Step S601: Obtain the physical node that carries the load.

[0133] This can be understood as meaning that, for each candidate migration virtual node set in the embodiments of this application, This allows us to determine the current candidate virtual migration nodes. Mapped to the physical node .

[0134] Step S602: Determine the set of neighboring physical nodes.

[0135] This can be understood as the embodiment of the application searching for the set of neighboring physical nodes that carry a physical node with an adjacent hop. Each physical node in the neighborhood physical node set must satisfy the node mapping condition, that is, satisfy the computing resource requirements of the virtual node and the cross-domain network constraint condition, which can be expressed as, but is not limited to:

[0136] ,

[0137] Step S603: Construct the destruction operator to obtain the destruction node set and the initial neighborhood candidate physical node set.

[0138] It can be understood that the embodiments of this application can construct the destruction operator of the ALNS algorithm. The destruction operator includes three destruction principles: the first principle is to use a greedy strategy to remove the set of neighboring physical nodes. The physical node with the highest load is the number of nodes to be removed. The second violation principle is to consider the importance of physical nodes and remove neighboring node sets. The node with the highest physical importance in the middle, the number of which is removed is The importance of physical nodes considers the node's degree centrality, proximity centrality, and link resource robustness. The calculation formula can be, but is not limited to, the following:

[0139] ,

[0140] The formulas for calculating the degree centrality, proximity centrality, and link resource robustness of a node can be, but are not limited to, as follows:

[0141] ,

[0142] ,

[0143] ,

[0144] in, This represents the degree centrality value of a node. Representing the proximity centrality of nodes, This indicates the robustness of a node's link resources. It is a node and The number of hops for the shortest path between nodes, where the robustness of a node's link resources is related to the physical nodes. The integration of connected physical link resources can reflect the physical nodes. The status of adjacent link resources.

[0145] The third violation principle is to randomly remove the set of neighboring physical nodes. The number of physical nodes in the system is This is to avoid the search algorithm getting stuck in a local optimum.

[0146] Furthermore, in this embodiment of the application, after using the above three destruction operators, the removed nodes and the filtered nodes are sorted to obtain a set of destroyed nodes. and the initial neighborhood candidate physical node set .

[0147] Step S604: Construct a repair operator to obtain a set of candidate physical nodes in the neighborhood.

[0148] This can be understood as the embodiment of this application concerning the destruction of the node set. Constructing a repair operator The first repair principle considers node load, sorting nodes by their physical load from smallest to largest, and selecting the node with the lowest load. Each node is "repaired," and the set of physical nodes in the repair neighborhood is added. The second repair principle considers the importance of nodes and calculates the set of damaged nodes. The physical nodes are ranked according to their importance from highest to lowest, and the nodes with the highest importance are selected. Each physical node is "repaired" and added to the set of physical nodes in the repair neighborhood. The third repair principle involves randomly restoring the set of damaged nodes. Each physical node is added to the set of physical nodes in the repair neighborhood. After one complete round of damage repair iteration, the number of neighborhood searches is obtained. The following is a set of candidate physical nodes in the neighborhood. It can be, but is not limited to, expressed as:

[0149] ,

[0150] In addition, after the operator identifier is destroyed, the embodiments of this application can selectively apply the repair operator according to the real-time status and priority policy of the network, and repair or reconfigure these nodes only on the premise of ensuring that the overall network performance will not be affected, so as to optimize the overall network configuration and performance.

[0151] Step S605: Calculate the migration burden of each neighboring candidate physical node in the neighboring candidate physical node set.

[0152] In this embodiment of the application, for the neighborhood candidate physical node set Each neighborhood candidate physical node The migration burden of each node is calculated according to the migration evaluation burden formula. The migration burden may include, but is not limited to, the consumption of computing resources, increased link latency, temporary memory requirements, and bandwidth consumption; this application does not impose specific limitations on these factors. Furthermore, the migration evaluation burden formula may be expressed, but is not limited to, as follows:

[0153] ,

[0154] in, It's the migration cost. This represents the node workload of the candidate physical nodes. It is to migrate virtual nodes The distance between the current physical node and the candidate physical node (represented by hop count). These are the weighting coefficients for the destruction and repair operators.

[0155] Step S606: Select the target set of physical nodes for migration based on the migration evaluation burden formula, update the operator weights for each iteration, and obtain the complete set of physical nodes for migration.

[0156] This can be understood as the embodiments of this application assigning the calculated migration burden to the neighborhood candidate physical node set. The physical nodes in the array are sorted from smallest to largest. For each node... If it is in the first half of the sorting sequence, then it is included in the target migration physical node set. And assign positive weights to the corresponding destruction operator and repair operator. Conversely, assign negative weights to the corresponding destruction and repair operators. The update formula for the operator can be, but is not limited to, expressed as:

[0157] ,

[0158] ,

[0159] in, These are the initial weights. It is the weight update coefficient, which controls how fast the weights are updated. The weights are assigned after one iteration of the operator. This is the number of times the operator has been used.

[0160] Repeat steps S603-S605 to continue searching for candidate migration virtual node sets. Each virtual node is carried in the middle. physical nodes that carry Adjacent two-hop physical sets During the execution of the ALNS algorithm, the roulette wheel algorithm is used to select the destruction operator and the repair operator according to the updated operator weights, and the iteration is continued until the complete set of the migrating physical nodes is obtained. It can be, but is not limited to, expressed as:

[0161] ,

[0162] Furthermore, it should be noted that when traversing the remaining virtual nodes of the current network slice requirement in the network slice requirement model in this application embodiment, the node link collaborative mapping and reconstruction deployment of cross-domain network slices can be completed. If any virtual node or virtual link cannot be mapped during the current network slice requirement collaborative mapping stage, the current network slice requirement cannot be successfully deployed.

[0163] The working principle of the cross-domain network slice reconstruction method proposed in this application will be introduced below with reference to a specific embodiment.

[0164] in, Figure 7 This is a flowchart illustrating the working principle of a cross-domain network slice reconstruction method provided according to an embodiment of this application.

[0165] Step S701: Obtain network slicing requirements.

[0166] Step S702: Perform cross-domain network collaborative slicing.

[0167] Step S703: Execute the AEC-VIKOR mechanism to evaluate the node.

[0168] Step S704: Perform virtual node mapping.

[0169] Step S705: Determine whether the node has been successfully mapped.

[0170] Step S706: Execute the node migration mechanism for slice reconstruction.

[0171] Step S707: Perform node reconstruction based on adaptive large neighborhood search.

[0172] Step S708: Cooperatively map virtual links.

[0173] Step S709: Complete slice mapping.

[0174] This application embodiment can be understood as follows: the AEC-VIKOR mechanism is used to traverse and sort the virtual nodes in the current network slice requirement topology using a breadth-first search algorithm to obtain a node mapping sequence.

[0175] Furthermore, in this embodiment of the application, the physical node set is searched based on the network configuration information of the current network slicing requirements in the network slicing requirement model. Virtual nodes that meet current network slicing requirements Candidate physical node set Specifically, in this embodiment of the application, the candidate physical node set is selected based on the current virtual node type. Select physical nodes that meet the computational resource constraints for mapping: (1) If the current virtual node is the root node of the breadth-first search tree, then map the virtual node to the node with the minimum importance value. (2) If it is not the root node, the virtual node is mapped to the node of the cross neighbor node set of the physical node set that carries the virtual parent node and the directly connected node. The node load calculation of the mapped node selects the physical node with the smallest node load in the candidate physical node set.

[0176] Furthermore, in this embodiment of the application, if node mapping fails, a slice reconstruction mechanism is triggered, which comprehensively considers the cost of node migration and the resource congestion of physical network nodes to determine the candidate set of physical nodes to be migrated and the candidate set of virtual nodes to be migrated.

[0177] (1) Search for the set of overloaded physical nodes in the current physical infrastructure network that are overloaded with resources. Calculate the node load of each physical node carrying virtual nodes in the current physical infrastructure network. If the load exceeds the node workload threshold... If the physical node is not found, it will be added to the overloaded physical node set. The congested physical node set and the overloaded physical node set will then be combined into a reconstructed physical node set. .

[0178] Sort the physical nodes in the reconstructed physical node set in descending order of resource congestion, and update the reconstructed physical node set. The order of physical nodes in the system.

[0179] (2) Determine the set of candidate migration virtual nodes The process retrieves the virtual nodes hosted by each reconstructed physical node in the reconstructed physical node set. Candidate virtual nodes for migration should meet certain migration conditions, which may include, but are not limited to, migration conditions and migration cost conditions. The migration conditions can be understood as the computational resource requirements that can be met after migration to deploy the currently failed virtual node. When the computational resource requirements of a virtual node are greater than the difference between the required node's computational resources and the allocated computational resources of the target physical node and the total node resources that the physical node can provide, it indicates that migrating the current virtual node can at least guarantee the successful deployment of the failed virtual node. This yields a set of candidate virtual nodes that meet the cost migration conditions, denoted as... .

[0180] Furthermore, embodiments of this application execute the ALNS algorithm to obtain... Complete set of incoming physical nodes after rounds of iterative search The set of target incoming physical nodes is selected based on the migration evaluation burden formula. Update the operator weights in each iteration to obtain a complete set of incoming physical nodes. .

[0181] Traverse the remaining virtual nodes in the current network slice requirement model to complete the node and link collaborative mapping and reconstruction deployment for cross-domain network slices. During the node and link collaborative mapping phase of the current network slice requirement, if any virtual node or virtual link cannot be mapped, the current network slice requirement cannot be successfully deployed; if the node mapping is successful, the KSP algorithm is used to collaboratively map the virtual links.

[0182] The cross-domain network slice reconstruction method proposed in this application evaluates the importance of nodes required for network slices, obtains a node mapping sequence between nodes, and triggers a network slice reconstruction mechanism when node mapping fails. This determines the candidate migration node set for the current network slice requirement, thereby obtaining the complete set of physical nodes to be migrated for the current network slice requirement, thus achieving network slice reconstruction. This not only balances network load and reduces migration costs in cross-domain network slice deployment scenarios but also effectively improves the efficiency of neighborhood search and slice reconstruction. Therefore, it solves the problems in related technologies that only consider the virtual node migration cost of slices, failing to achieve load balancing of the physical underlying network, resulting in a low success rate for network slices. Furthermore, as the number of slices increases, the computational efficiency of network slice deployment decreases, the time consumption increases, and the robustness of the slice deployment system deteriorates.

[0183] Next, the reconstruction apparatus for cross-domain network slices according to embodiments of this application is described with reference to the accompanying drawings.

[0184] Figure 8This is a block diagram of a cross-domain network slice reconstruction apparatus provided according to an embodiment of this application.

[0185] like Figure 8 As shown, the cross-domain network slice reconstruction device 10 includes: an acquisition module 100, a generation module 200, a determination module 300, and a reconstruction module 400.

[0186] The acquisition module 100 is used to acquire all network slice requirements in the cross-domain network, and calculate the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement according to the pre-built network slice requirement model, so as to obtain the node mapping sequence between virtual nodes and physical nodes based on the virtual importance value and the physical importance value.

[0187] The generation module 200 is used to combine the network slicing requirement model and the node mapping sequence to search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on at least one initial candidate physical node.

[0188] The determination module 300 is used to perform node mapping between virtual nodes and initial candidate physical nodes based on the initial candidate physical node set. In the case of node mapping failure, it obtains the congested physical node set and the overloaded physical node set in the current network slice requirement, and uses the slice reconstruction mechanism to re-map the nodes to obtain the candidate migration physical node set and the candidate migration virtual node set. The slice reconstruction mechanism includes generating the reconstructed physical node set of the current network slice requirement using the congested physical node set and the overloaded physical node set, and determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set.

[0189] The reconstruction module 400 is used to obtain the complete set of incoming physical nodes for the current network slice requirements by using the candidate set of migration physical nodes and the candidate set of migration virtual nodes, so as to redeploy the current network slice requirements according to the complete set of incoming physical nodes and realize the reconstruction of the current network slice requirements.

[0190] Optionally, in one embodiment of this application, the generation module 200 includes a filtering unit and a first generation unit.

[0191] The filtering unit is used to combine the network slicing requirement model and the node mapping sequence to filter out the set of candidate physical nodes that meet the virtual node mapping conditions from the physical node set.

[0192] The first generation unit is used to combine virtual nodes and cross-domain network domain constraints to obtain at least one initial candidate physical node using the set of candidate physical nodes, and to generate an initial candidate physical node set using the at least one initial candidate physical node as the root node.

[0193] Optionally, in one embodiment of this application, the determining module 300 includes: a first calculation unit, a sorting unit, a judgment unit, and a determining unit.

[0194] The first calculation unit is used to calculate the resource congestion level of each reconstructed physical node in the reconstructed physical node set by using the node resource difference and physical link utilization.

[0195] The sorting unit is used to sort all the reconstructed physical nodes in the reconstructed physical node set according to the degree of resource congestion, so as to obtain a set of candidate migration physical nodes that meet the preset order of the reconstructed physical nodes.

[0196] The judgment unit is used to determine whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet the preset migration conditions.

[0197] The determining unit is used to construct an initial candidate migration virtual node set using the mapped virtual nodes when the mapped virtual nodes meet the preset migration conditions, so as to obtain a candidate migration virtual node set, and obtain a candidate migration virtual link set and a candidate migration physical link set based on the preset link requirements.

[0198] Optionally, in one embodiment of this application, the reconstruction module 400 includes: a second generation unit, a third generation unit, a first construction unit, a second construction unit, and a second calculation unit.

[0199] The second generation unit is used to obtain the physical node to which the candidate migration virtual node is mapped using the candidate migration virtual node set.

[0200] The third generation unit is used to determine the set of neighboring physical nodes corresponding to the candidate migration virtual nodes based on the carrying physical nodes and the candidate migration physical node set.

[0201] The first construction unit is used to construct a destruction operator based on the neighborhood physical node set, and to calculate the destruction node set and the initial neighborhood candidate physical node set corresponding to the neighborhood physical node set through the destruction operator.

[0202] The second building unit is used to construct a repair operator based on the damaged node set, and repair the initial neighborhood candidate physical node set through the repair operator to obtain the neighborhood candidate physical node set.

[0203] The second computing unit is used to calculate the migration burden of each neighboring candidate physical node in the neighboring candidate physical node set, so as to generate a complete set of migrating-in physical nodes based on the migration burden. The migration burden may include, but is not limited to, the consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption.

[0204] Optionally, in one embodiment of this application, the calculation formula for the complete set of migrating physical nodes may be, but is not limited to, the following:

[0205] ,

[0206] in, It was through The set of destroyed nodes after round-neighbor iterative search. It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The set of repaired neighborhood physical nodes after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

[0207] It should be noted that the foregoing explanation of the cross-domain network slice reconstruction method embodiment also applies to the cross-domain network slice reconstruction apparatus of this embodiment, and will not be repeated here.

[0208] The cross-domain network slice reconstruction apparatus proposed in this application evaluates the importance of nodes required for network slices, obtains a node mapping sequence between nodes, and triggers a network slice reconstruction mechanism when node mapping fails. This determines the candidate migration node set for the current network slice requirement, and ultimately obtains the complete set of physical nodes to be migrated for the current network slice requirement, thus achieving network slice reconstruction. This not only balances network load and reduces migration costs in cross-domain network slice deployment scenarios but also effectively improves the efficiency of neighborhood search and slice reconstruction. Therefore, it solves the problems in related technologies that only consider the virtual node migration cost of slices, failing to achieve load balancing of the physical underlying network, resulting in a low success rate for network slices. Furthermore, as the number of slices increases, the computational efficiency of network slice deployment decreases, the time consumption increases, and the robustness of the slice deployment system deteriorates.

[0209] Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. The electronic device may include:

[0210] The memory 901, the processor 902, and the computer program stored on the memory 901 and capable of running on the processor 902.

[0211] When processor 902 executes the program, it implements the cross-domain network slice reconstruction method provided in the above embodiments.

[0212] Furthermore, electronic devices also include:

[0213] Communication interface 903 is used for communication between memory 901 and processor 902.

[0214] The memory 901 is used to store computer programs that can run on the processor 902.

[0215] The memory 901 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0216] If the memory 901, processor 902, and communication interface 903 are implemented independently, then the communication interface 903, memory 901, and processor 902 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0217] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0218] The processor 902 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0219] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for reconstructing cross-domain network slices.

[0220] This application also provides a computer program product, including a computer program that, when executed, implements the above-described method for reconstructing cross-domain network slices.

[0221] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0222] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0223] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0224] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0225] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0226] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0227] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0228] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for reconstructing cross-domain network slices, characterized in that, Includes the following steps: Obtain all network slice requirements in the cross-domain network, and calculate the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement based on the pre-built network slice requirement model, so as to obtain the node mapping sequence between the virtual node and the physical node based on the virtual importance value and the physical importance value; Combining the network slicing requirement model and the node mapping sequence, search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on the at least one initial candidate physical node. Based on the initial candidate physical node set, node mapping is performed between the virtual node and the initial candidate physical node. If node mapping fails, the congested physical node set and the overloaded physical node set in the current network slice requirement are obtained, and the node mapping is re-performed using a slice reconstruction mechanism to obtain a candidate migration physical node set and a candidate migration virtual node set. The slice reconstruction mechanism includes generating a reconstructed physical node set for the current network slice requirement using the congested physical node set and the overloaded physical node set, and determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set. The complete set of physical nodes for migration to the current network slice requirement is obtained by using the set of candidate migration physical nodes and the set of candidate migration virtual nodes, so as to redeploy the current network slice requirement according to the complete set of migration physical nodes and realize the reconstruction of the current network slice requirement. The step of obtaining the complete set of incoming physical nodes for the current network slice requirement using the candidate set of migration physical nodes and the candidate set of migration virtual nodes includes: The physical node to which the candidate migration virtual node is mapped is obtained using the candidate migration virtual node set. Based on the carrying physical node and the candidate migration physical node set, determine the neighborhood physical node set corresponding to the candidate migration virtual node; A destruction operator is constructed based on the neighborhood physical node set, and the destruction node set and the initial neighborhood candidate physical node set corresponding to the neighborhood physical node set are calculated through the destruction operator; The destruction operator includes: The first principle of destruction is to adopt a greedy strategy to filter and remove the number of nodes with the highest load from the set of neighboring physical nodes. The second destruction principle considers the importance of the node and filters out the most important nodes from the set of neighboring physical nodes, removing the second number of such nodes. The third violation principle is to randomly remove a third number of nodes from the set of neighboring physical nodes. After performing calculations using the first destruction principle, the second destruction principle, and the third destruction principle, the removed nodes and the filtered nodes are sorted out to obtain the corresponding set of destroyed nodes and the initial set of candidate physical nodes in the neighborhood. The formula for calculating the importance of a node is: ， ， ， ， in, This represents the degree centrality value of a node. Representing the proximity centrality of nodes, This indicates the robustness of a node's link resources. It is a node and The number of hops for the shortest path between nodes, where the robustness of a node's link resources is related to the physical nodes. The integration of connected physical link resources can reflect the physical nodes. The status of adjacent link resources; A repair operator is constructed based on the damaged node set, and the initial neighborhood candidate physical node set is repaired by the repair operator to obtain the neighborhood candidate physical node set; The repair operator includes: The first repair principle is to consider the load of the node, select the fourth number of nodes with the smallest load from the set of damaged nodes, and add the node with the smallest load to the set of physical nodes in the repair neighborhood. The second repair principle is to consider the importance of the nodes, select the fifth most important node from the set of damaged nodes, and add the most important node to the set of physical nodes in the repair neighborhood. The third repair principle is to randomly select a fifth number of random nodes from the set of damaged nodes and add the random nodes to the set of physical nodes in the repair neighborhood; After repairing and iterating using the first repair principle, the second repair principle, and the third repair principle, the neighborhood candidate physical node set is obtained; The migration burden of each neighboring candidate physical node in the neighboring candidate physical node set is calculated to generate the complete set of incoming physical nodes based on the migration burden, wherein the migration burden includes consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption.

2. The method according to claim 1, characterized in that, The step of combining the network slicing requirement model and the node mapping sequence to search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on the at least one initial candidate physical node, includes: Combining the network slicing requirement model and the node mapping sequence, a set of candidate physical nodes that meet the virtual node mapping conditions is selected from the set of physical nodes; By combining the virtual node and the cross-domain network constraints, at least one initial candidate physical node is obtained using the set of candidate physical nodes, and the at least one initial candidate physical node is used as the root node to generate the initial candidate physical node set.

3. The method according to claim 1, characterized in that, The step of determining the candidate set of physical nodes and the candidate set of virtual nodes for migration based on the reconstructed physical node set for the current network slice requirement includes: The resource congestion level of each reconstructed physical node in the reconstructed physical node set is calculated using the node resource difference and physical link utilization. All reconstructed physical nodes in the reconstructed physical node set are sorted according to the resource congestion level to obtain a set of candidate migration physical nodes that satisfy the preset order of reconstructed physical nodes; Determine whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet the preset migration conditions; If the mapped virtual node meets the preset migration conditions, then the mapped virtual node is used to form an initial candidate migration virtual node set to obtain the candidate migration virtual node set, and a candidate migration virtual link set and a candidate migration physical link set are obtained based on the preset link requirements.

4. The method according to claim 1, characterized in that, The calculation formula for the complete set of migrated physical nodes is as follows: ， in, It was through The set of destroyed nodes after round-neighbor iterative search It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The repaired set of physical nodes in the neighborhood after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

5. A reconstruction apparatus for cross-domain network slices, characterized in that, include: The acquisition module is used to acquire all network slice requirements in the cross-domain network, and calculate the virtual importance value of each virtual node and the physical importance value of each physical node in the current network slice requirement according to the pre-built network slice requirement model, so as to obtain the node mapping sequence between the virtual node and the physical node based on the virtual importance value and the physical importance value. The generation module is used to combine the network slicing requirement model and the node mapping sequence to search for at least one initial candidate physical node in the physical node set that simultaneously satisfies the virtual node mapping condition and the cross-domain network domain constraint condition, so as to obtain the initial candidate physical node set corresponding to the current network slicing requirement based on the at least one initial candidate physical node. The determination module is used to perform node mapping between the virtual node and the initial candidate physical node based on the initial candidate physical node set, and in the case of node mapping failure, obtain the congested physical node set and the overloaded physical node set in the current network slice requirement, and re-perform node mapping using a slice reconstruction mechanism to obtain a candidate migration physical node set and a candidate migration virtual node set. The slice reconstruction mechanism includes generating a reconstructed physical node set for the current network slice requirement using the congested physical node set and the overloaded physical node set, and determining the candidate migration physical node set and the candidate migration virtual node set based on the reconstructed physical node set. The reconstruction module is used to obtain a complete set of incoming physical nodes for the current network slice requirement using the candidate set of migration physical nodes and the candidate set of migration virtual nodes, so as to redeploy the current network slice requirement according to the complete set of incoming physical nodes and realize the reconstruction of the current network slice requirement. The reconstruction module includes: The second generation unit is used to obtain the physical node to which the candidate migration virtual node is mapped using the candidate migration virtual node in the candidate migration virtual node set. The third generation unit is used to determine the set of neighboring physical nodes corresponding to the candidate migration virtual node based on the carrying physical node and the set of candidate migration physical nodes. The first construction unit is used to construct a destruction operator based on the neighborhood physical node set, and to calculate the destruction node set and the initial neighborhood candidate physical node set corresponding to the neighborhood physical node set through the destruction operator; The destruction operator includes: The first principle of destruction is to adopt a greedy strategy to filter and remove the number of nodes with the highest load from the set of neighboring physical nodes. The second destruction principle considers the importance of the node and filters out the most important nodes from the set of neighboring physical nodes, removing the second number of such nodes. The third violation principle is to randomly remove a third number of nodes from the set of neighboring physical nodes. After performing calculations using the first destruction principle, the second destruction principle, and the third destruction principle, the removed nodes and the filtered nodes are sorted out to obtain the corresponding set of destroyed nodes and the initial set of candidate physical nodes in the neighborhood. The formula for calculating the importance of a node is: ， ， ， ， in, This represents the degree centrality value of a node. Representing the proximity centrality of nodes, This indicates the robustness of a node's link resources. It is a node and The number of hops for the shortest path between nodes, where the robustness of a node's link resources is related to the physical nodes. The integration of connected physical link resources can reflect the physical nodes. The status of adjacent link resources; The second construction unit is used to construct a repair operator based on the damaged node set, and repair the initial neighborhood candidate physical node set through the repair operator to obtain the neighborhood candidate physical node set; The repair operator includes: The first repair principle is to consider the load of the node, select the fourth number of nodes with the smallest load from the set of damaged nodes, and add the node with the smallest load to the set of physical nodes in the repair neighborhood. The second repair principle is to consider the importance of the nodes, select the fifth most important node from the set of damaged nodes, and add the most important node to the set of physical nodes in the repair neighborhood. The third repair principle is to randomly select a fifth number of random nodes from the set of damaged nodes and add the random nodes to the set of physical nodes in the repair neighborhood; After repairing and iterating using the first repair principle, the second repair principle, and the third repair principle, the neighborhood candidate physical node set is obtained; The second computing unit is used to calculate the migration burden of each neighboring candidate physical node in the neighboring candidate physical node set, so as to generate the complete set of migrating-in physical nodes based on the migration burden, wherein the migration burden includes the consumed computing resources, increased link latency, and temporary memory requirements and bandwidth consumption.

6. The apparatus according to claim 5, characterized in that, The generation module includes: The filtering unit is used to combine the network slicing requirement model and the node mapping sequence to filter out a set of candidate physical nodes that meet the virtual node mapping conditions from the physical node set. The first generation unit is used to combine the virtual node and the cross-domain network domain constraints, use the set of candidate physical nodes to obtain the at least one initial candidate physical node, and use the at least one initial candidate physical node as the root node to generate the initial candidate physical node set.

7. The apparatus according to claim 5, characterized in that, The determining module includes: The first calculation unit is used to calculate the resource congestion level of each reconstructed physical node in the reconstructed physical node set using the node resource difference and physical link utilization. The sorting unit is used to sort all the reconstructed physical nodes in the reconstructed physical node set according to the resource congestion level, so as to obtain a set of candidate migration physical nodes that satisfy the preset order of the reconstructed physical nodes; The judgment unit is used to determine whether the mapped virtual nodes corresponding to the candidate migration physical nodes in the candidate migration physical node set meet the preset migration conditions. The determining unit is configured to, when the mapped virtual node satisfies the preset migration conditions, use the mapped virtual node to construct an initial candidate migration virtual node set, so as to obtain the candidate migration virtual node set, and obtain a candidate migration virtual link set and a candidate migration physical link set based on preset link requirements.

8. The apparatus according to claim 5, characterized in that, The calculation formula for the complete set of migrated physical nodes is as follows: ， in, It was through The set of destroyed nodes after round-neighbor iterative search It was through The set of physical nodes in the neighborhood is selected after round-neighbor iterative search. go through The repaired set of physical nodes in the neighborhood after round-neighbor iterative search. It was through The set of candidate physical nodes in the neighborhood after round-neighborhood iterative search.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the cross-domain network slice reconstruction method as described in any one of claims 1-4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the method for reconstructing cross-domain network slices as described in any one of claims 1-4.

11. A computer program product, characterized in that, Includes a computer program, which, when executed, is used to implement the method for reconstructing cross-domain network slices as described in any one of claims 1-4.

Images