5G network resource management and control method based on service function chain mapping
By combining the service function chain mapping with service type division resource slices and computing path reliability product, the problem of service level guarantee in wide-area deterministic network is solved, and customized reliable services and efficient data transmission are realized.
Patent Information
- Application Number
- CN202211238144.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-10-10
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-10-10
AI Technical Summary
The existing service function chain orchestration and deployment solutions cannot effectively guarantee the service level of different services in wide-area deterministic networks, and cannot perceive business dynamic changes and network topology changes, resulting in the inability to guarantee data transmission quality.
Through a service function chain mapping method, resource slices are divided according to the service type, multiple shorter paths are selected, path reliability product is calculated, SFC is mapped to paths with higher reliability, and the reliability and delay are trade-offs are made between them, and choose paths with lower latency for deployment, so as to improve network resource utilization and service acceptance rate.
It realizes customized and reliable services for different services, reduces end-to-end service delays, improves network resource utilization and service acceptance rate, and ensures the quality of data transmission.
Smart Images

Figure CN115714724B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of network data transmission technology, and in particular to a 5G network resource management and control method based on service function chain mapping. Background Art
[0002] Network Function Virtualization (NFV) is an important technology for the current development of 5G networks. It aims to use virtualization and other technologies to decouple common network functions such as load balancers and firewalls from traditional physical proprietary devices and deploy them on communication equipment in a software-based form. Virtual network functions can not only rely on general hardware to implement specific network functions, but also can be dynamically created and flexibly expanded based on the shared physical resource pool provided by network operators, reducing operating costs and improving the efficiency of network management and services. Unlike NFV, which is mainly responsible for the virtualization of various network elements, Software Defined Network (SDN) is mainly responsible for the virtualization of the network itself. Combining NFV with SDN and applying it to 5G networks, operators can achieve flexible scheduling and dynamic allocation of network resources. As the types of services become more diverse, end-to-end network slicing technology has been proposed. In a software-based network where the control plane and the service plane are separated, operators can provide customized services for different types of services by dividing network resource slices. Each network slice serving a service involves network equipment, access network, transmission network, and core network. They are logically isolated from each other, and exclusive performance such as service quality, network bandwidth, and security performance needs to be fully guaranteed. In the network architecture based on NFV / SDN, it is only necessary to place and link the network functions in order according to the user request information and the corresponding data traffic forwarding diagram in the network to create end-to-end network services for users. These service nodes arranged according to business logic form a service function chain (SFC).
[0003] Currently, there has been a lot of work and research on service function chain technology, mainly focusing on three aspects: service function chain orchestration, deployment and dynamic optimization. SFC orchestration refers to combining network functions in an appropriate manner according to user request information to build a service function chain. SFC deployment refers to selecting appropriate nodes in the physical network for VNF instantiation. Dynamic optimization is to ensure the stability and reliability of services in dynamic networks through VNF expansion and contraction, dynamic allocation of virtual resources, etc.
[0004] However, the current research on resource management technology based on service function chain is not mature and complete, and further research and design are needed. 5G deterministic network capability is the direction of sustainable development of 5G network. Among them, wide area deterministic network is the latest development stage of 5G deterministic network, which mainly includes four types of application scenarios: wide area broadband access, wide area large-scale connection, wide area time-sensitive connection and wide area high-precision positioning. There are a large number of business requirements with different service levels in these scenarios, but the wide area network structure is complex and has certain time-varying characteristics. The existing service function chain orchestration and deployment scheme does not take into account the business characteristics of different application scenarios of wide area deterministic network, and it is difficult to guarantee the service level of different services. Although the SFC orchestration method based on delay perception reduces the end-to-end delay, it cannot perceive the dynamic changes of business, network element damage or dynamic changes of network topology, and cannot detect the reliability of communication network nodes in processing data and the changing link quality, so it cannot guarantee the quality of data transmission.
[0005] Based on the above-mentioned technical problems existing in the prior art, the present invention provides a 5G network resource management and control method based on service function chain mapping. Summary of the invention
[0006] The present invention proposes a 5G network resource management and control method based on service function chain mapping. In order to ensure the service level of different services, provide users with customized reliable services, reduce end-to-end service delay, and ensure the quality of data transmission, different types of resource slices are divided according to different service types, and multiple shorter paths are selected on the slices. By sensing the normal working probability and load conditions of the current network nodes and links, the reliability product of the deployment path is calculated, and the SFC is mapped to a path with higher reliability. When the reliability of the selected path is low, a path with lower delay is selected for deployment. The end-to-end delay and service reliability are comprehensively considered, the service delay is reduced on the basis of ensuring reliability, services are provided for different types of services, and the network resource utilization rate and service acceptance rate are improved.
[0007] The present invention adopts the following technical solutions:
[0008] A 5G network resource management and control method based on service function chain mapping, comprising:
[0009] Step 1, obtaining service function chain request and user demand information;
[0010] Step 2: Classify the service function chain requests according to user demand information, and correspond different types of service function chain requests to different types of network slices;
[0011] Step 3: In the corresponding network slice, search for all connected paths from the source node to the destination node, sort the paths in ascending order according to the end-to-end delay, and select the top 10 paths using the KSP algorithm;
[0012] Step 4: Calculate the reliability of each node and link in each path according to the node load parameters, node normal working probability, link load parameters and link connectivity in each path, multiply the reliability of each path and the path through which the path passes, and obtain the reliability product corresponding to each path. The path with the largest reliability product is used as the deployment path of the service function chain.
[0013] Step 5: Calculate the reliability product of the deployment path and compare the reliability product with the reliability threshold. If the reliability product is greater than or equal to the reliability threshold, instantiate the VNF on the corresponding 5G network node, map the selected virtual link to the actual physical network, form a service function chain, and update the load of each node and link according to the resource occupancy of the deployed service function chain. If the reliability product is less than the reliability threshold, select the path with the shortest end-to-end delay as the deployment path of the service function chain, instantiate the VNF in the network, and build the service function chain.
[0014] Furthermore, in step 5, the reliability threshold is determined by the node load threshold, the node normal operation probability threshold, the link load threshold and the link connectivity probability threshold.
[0015] Furthermore, in step 5, the 5G network nodes include base station nodes, router nodes, gateway nodes, and data center nodes.
[0016] Furthermore, step 2 or step 3 also includes a step of generating a virtual service function sequence according to user demand information.
[0017] Furthermore, in step 1, the user demand information includes the bandwidth required for deployment and the delay requirement for the deployment service function chain.
[0018] Furthermore, the types of network slices in step 2 and / or step 3 include wide-area broadband access slices, wide-area large-scale connection slices, wide-area time-sensitive connection slices, and wide-area high-precision positioning slices.
[0019] Furthermore, in step 4, the reliability product is determined by the probability of normal operation of the network node, the computing resource load of the network node, the connectivity of the network link, and the bandwidth occupancy of the network link.
[0020] Furthermore, in step 4, the network node reliability is related to the node normal working probability and the node load parameter, and the network node reliability is defined as:
[0021]
[0022] Y i (q) is the probability that network node i works normally, It is the ratio of the computing resources used on node i and the node computing resources required by the SFC request to be deployed in the network to all the computing resources of the node.
[0023] Furthermore, in step 4, the link reliability is related to the current link connectivity probability and load condition, and the link reliability is defined as:
[0024]
[0025] Among them, S i,j (q) is link l i,j The connectivity probability, in [e -0.005(q-1) ,e -0.005q ] is evenly distributed on the network, q is the total number of SFCs in the network slice, It is the ratio of the occupied bandwidth on the link and the link bandwidth required in the SFC request to the total bandwidth of the link.
[0026] Furthermore, in step 5, the deployment path reliability product is determined by the node and link reliabilities on the deployment path, and is expressed as:
[0027]
[0028] in, Represents virtual node k u Whether to map to network node n i superior, Indicates virtual link e m,n Whether to map to the actual link l i,j In the above, N is the set of physical nodes, F is the set of virtual nodes corresponding to the SFC request q, L is the set of physical links, and E is the set of virtual links corresponding to the SFC request q.
[0029] Compared with the prior art, the present invention has the following advantages:
[0030] 1. The 5G network resource management and control method based on service function chain mapping described in the present invention divides network slices into different types, and can deploy service function chains in different network slices according to user request information, thereby ensuring different types of service requirements and providing users with customized and reliable services.
[0031] 2. The 5G network resource management and control method based on service function chain mapping described in the present invention comprehensively considers the reliability and end-to-end delay of the deployed service function chain, first uses the KSP algorithm to select multiple paths with shorter delays in the corresponding slices, detects the working status and load of network nodes and links on different paths, and calculates the reliability product of different paths. If the maximum reliability product is greater than the threshold value, the path with the highest reliability is selected as the deployment path, otherwise, the path with the shortest delay is selected as the deployment path. A trade-off is made between reliability and delay, which not only ensures the reliability of the service but also reduces the end-to-end delay.
[0032] 3. The 5G network resource management and control method based on service function chain mapping described in the present invention can map SFC to paths with lower node and link resource utilization, slowing down the saturation speed of node and link resources, and being able to accept more SFC requests, thereby improving network resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 It is a flow chart of a 5G network resource control method based on service function chain mapping in an embodiment of the present invention;
[0034] Figure 2 Schematic diagram comparing the reliability of deployment paths of the method described in the embodiment of the present invention and other algorithms;
[0035] Figure 3 is a schematic diagram comparing node resource utilization of the method described in the embodiment of the present invention and other algorithms;
[0036] Figure 4 is a schematic diagram comparing link resource utilization of the method described in an embodiment of the present invention and other algorithms;
[0037] Figure 5 3 is a schematic diagram comparing the service acceptance rates of the method described in the embodiment of the present invention and other algorithms. DETAILED DESCRIPTION
[0038] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.
[0039] Example
[0040] like Figure 1 As shown, the 5G network resource management method based on service function chain mapping includes:
[0041] S1: accepts the service function chain request and analyzes the user demand information in the service function chain request. The user demand information includes the bandwidth required for deployment and the delay requirement for deploying the service function chain.
[0042] S2: Classify requests according to user demand information. Network slice types include wide-area broadband access slices, wide-area large-scale connection slices, wide-area time-sensitive connection slices, and wide-area high-precision positioning slices. If the bandwidth required for deploying the service function chain in the user demand information is greater than 30MHz and the latency requirement is less than 160ms, the service chain request corresponds to the wide-area time-sensitive connection slice; if the bandwidth required for deploying the service function chain in the user demand information is greater than 30MHz and the latency requirement is higher than 160ms, the service chain request corresponds to the wide-area broadband access slice; if the bandwidth required for deploying the service function chain in the user demand information is less than 30MHz and the latency requirement is lower than 160ms, the service chain request corresponds to the wide-area high-precision positioning slice; if the bandwidth required for deploying the service function chain in the user demand information is less than 30MHz and the latency requirement is higher than 160ms, the service chain request corresponds to the wide-area large-scale connection slice;
[0043] S3: Generate a virtual network function sequence based on the user request information of the deployed service function chain. Specifically, list the corresponding service function modules according to the different characteristics of the business requirements and the different types of network functions used. The resource types and service functions on different nodes are different. Generate a virtual service function sequence corresponding to the requirements, so as to facilitate the subsequent mapping of the virtual service functions to the nodes containing the service functions in sequence, and realize the instantiation of the virtual network functions.
[0044] S4: In the selected network slice, use the KSP algorithm to select multiple actual paths with low latency, including the network nodes and network links used;
[0045] S5: Calculate the reliability product of each network node and link on each path. The reliability product is determined by the probability of normal operation of the network node, the computing resource load of the network node, the connectivity of the network link, and the bandwidth occupancy of the network link.
[0046] The reliability of a network node is related to the probability of normal operation of the node and the node load parameters. The reliability of a network node is defined as:
[0047]
[0048] Y i (q) is the probability that network node i works normally, is the ratio of the computing resources used on node i and the node computing resources required by the SFC request to be deployed in the network to all the computing resources of the node;
[0049] Link reliability is related to the current link connectivity probability and load conditions. Link reliability is defined as:
[0050]
[0051] S i,j (q) is link l i,j The connectivity probability, in [e -0.005(q-1) ,e -0.005q ] is evenly distributed on the network, where q is the total number of SFCs in the network slice, It is the ratio of the occupied bandwidth on the link and the link bandwidth required in the SFC request to the total bandwidth of the link;
[0052] The deployment path reliability product is determined by the node and link reliabilities on the deployment path and is expressed as:
[0053]
[0054] in, Represents virtual node k u Whether to map to network node n i superior, Indicates virtual link e m,n Whether to map to the actual link l i,j In the above, N is the set of physical nodes, F is the set of virtual nodes corresponding to the SFC request q, L is the set of physical links, and E is the set of virtual links corresponding to the SFC request q;
[0055] S6: Select the path with the largest reliability product among multiple paths as the deployment path of the service function chain;
[0056] S7: Judge the reliability product of the deployment path selected in the previous step. If the reliability product is greater than the threshold value, instantiate the VNF on the corresponding 5G network node, select the network link, form a service function chain, and update the load of each node and link; otherwise, select the path with the shortest delay as the deployment path of the service function chain, instantiate the VNF in the network, and build the service function chain.
[0057] like Figure 2-5 As shown, in order to verify the superior technical effect of the method described in this embodiment, the method described in this embodiment, the SPO algorithm, and the SDSN algorithm are compared. The Reliability Awareness Map algorithm is a service function chain mapping algorithm based on reliability awareness proposed in this embodiment; the SPO algorithm is a mapping algorithm based on the shortest path, which directly calculates the shortest path between user endpoints as the basic path for data forwarding, and deploys the VNF on the physical node of the path; the SDSN algorithm is a mapping algorithm based on minimum delay, which aims to minimize the end-to-end delay, takes into account the node utilization and deployment probability, and combines the shortest path algorithm to select the mapping path of the SFC in the 5G network;
[0058] The method described in this embodiment selects the path with the highest reliability from multiple shorter paths as the deployment path, and has high reliability. However, the SPO algorithm maps the SFC to the network and only considers whether the end-to-end path is the shortest, without considering the node and link load conditions, and has lower reliability than the method proposed in this embodiment. The SDSN algorithm deploys the SFC on the gateway node with low resource utilization, and selects the path according to the shortest path algorithm. When the number of SFC requests is small, the deployment path reliability is higher than the method described in this embodiment. When the number of SFC requests increases, the SFC is deployed on more different paths based on the method described in this embodiment, which improves the balance of the network load and makes the deployment path more reliable.
[0059] When the number of SFC requests is greater than 120, the node resource utilization rate is increased by 27.8%, and the link resource utilization rate is increased by 40.9%. The method described in this embodiment selects different types of network slices according to the SFC request, and the network resources in each slice are more fully utilized. In addition, the deployment path reliability is related to the node and link load, and the SFC can be mapped to a path with low node and link resource utilization efficiency. The node and link resource saturation speed is slowed down, and more SFC requests can be accepted, thereby improving the network resource utilization rate;
[0060] When the number of SFC requests increases, the method described in this embodiment utilizes network slicing classification and comparison of deployment path reliability to deploy SFC to more paths and distribute the load to different paths, thereby improving the SFC request service acceptance rate. When the number of SFC requests is greater than 120, the service acceptance rate of the method described in this embodiment is improved by 27.4% compared with the SPO algorithm.
[0061] The present invention is not limited by the above embodiments. The above embodiments and descriptions are only for explaining the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, which are within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the appended claims.
Claims
1. A 5G network resource management and control method based on service function chain mapping, It is characterized in that include: Step 1, obtaining service function chain request and user demand information; Step 2: Classify the service function chain requests according to user demand information, and correspond different types of service function chain requests to different types of network slices; Step 3: In the corresponding network slice, search for all connected paths from the source node to the destination node, sort the paths in ascending order according to the end-to-end delay, and select the top 10 paths using the KSP algorithm; Step 4: Calculate the reliability of each node and link in each path according to the node load parameters, node normal working probability, link load parameters and link connectivity in each path, multiply the reliability of each path and the path through which the path passes, and obtain the reliability product corresponding to each path. The path with the largest reliability product is used as the deployment path of the service function chain. Step 5: Calculate the reliability product of the deployment path and compare the reliability product with the reliability threshold. If the reliability product is greater than or equal to the reliability threshold, instantiate the VNF on the corresponding 5G network node, map the selected virtual link to the actual physical network, form a service function chain, and update the load of each node and link according to the resource occupancy of the deployed service function chain. If the reliability product is less than the reliability threshold, select the path with the shortest end-to-end delay as the deployment path of the service function chain, instantiate the VNF in the network, and build the service function chain.
2. According to claim 1, the 5G network resource management and control method based on service function chain mapping, It is characterized in that Step 2 or step 3 also includes the step of generating a virtual service function sequence according to user demand information.
3. According to claim 1, the 5G network resource management and control method based on service function chain mapping, It is characterized in that In step 1, the user demand information includes the bandwidth required for deployment and the latency requirement for the deployment service function chain.
4. According to claim 1, the 5G network resource management and control method based on service function chain mapping, It is characterized in that Types of network slicing include wide-area broadband access slicing, wide-area large-scale connection slicing, wide-area time-sensitive connection slicing, and wide-area high-precision positioning slicing.
5. According to claim 1, the 5G network resource management and control method based on service function chain mapping, It is characterized in that In step 4, the reliability product is determined by the probability of normal operation of the network node, the computing resource load of the network node, the connectivity of the network link, and the bandwidth occupancy of the network link.
6. According to claim 5, the 5G network resource management and control method based on service function chain mapping, It is characterized in that In step 4, the network node reliability is related to the node normal working probability and the node load parameter. The network node reliability is defined as: Y i (q) is the probability that network node i works normally, It is the ratio of the computing resources used on node i and the node computing resources required by the SFC request to be deployed in the network to all the computing resources of the node.
7. According to claim 5, the 5G network resource management and control method based on service function chain mapping, It is characterized in that In step 4, the link reliability is related to the current link connectivity probability and load conditions. The link reliability is defined as: S i,j (q) is link l i,j The connectivity probability, in [e -0.005(q-1) ,e -0.005q ] is evenly distributed on the network, where q is the total number of SFCs in the network slice, It is the ratio of the occupied bandwidth on the link and the link bandwidth required in the SFC request to the total bandwidth of the link.
8. According to claim 5, the 5G network resource management and control method based on service function chain mapping, It is characterized in that In step 5, the deployment path reliability product is determined by the node and link reliabilities on the deployment path, expressed as: in Represents virtual node k u Whether to map to network node n i superior, Indicates virtual link e m,n Is it mapped to the actual link? i,j In the above, N is the set of physical nodes, F is the set of virtual nodes corresponding to the SFC request q, L is the set of physical links, and E is the set of virtual links corresponding to the SFC request q.
9. The 5G network resource management and control method based on service function chain mapping according to claim 1, It is characterized in that In step 5, the 5G network nodes include base station nodes, router nodes, gateway nodes, and data center nodes.
10. The 5G network resource management and control method based on service function chain mapping according to claim 1, It is characterized in that In step 5, the reliability threshold is determined by the node load threshold, the node normal operation probability threshold, the link load threshold and the link connectivity probability threshold.
