CN114938335B - Dynamic scheduling method and device for network slice service function - Google Patents
Dynamic scheduling method and device for network slice service function Download PDFInfo
- Publication number
- CN114938335B CN114938335B CN202210448065.7A CN202210448065A CN114938335B CN 114938335 B CN114938335 B CN 114938335B CN 202210448065 A CN202210448065 A CN 202210448065A CN 114938335 B CN114938335 B CN 114938335B
- Authority
- CN
- China
- Prior art keywords
- slice
- network
- resource
- virtual
- flow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 230000006870 function Effects 0.000 claims abstract description 125
- 238000013507 mapping Methods 0.000 claims abstract description 28
- 238000013528 artificial neural network Methods 0.000 claims description 37
- 238000006243 chemical reaction Methods 0.000 claims description 21
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000011217 control strategy Methods 0.000 claims description 10
- 230000009471 action Effects 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 7
- 230000002123 temporal effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000003062 neural network model Methods 0.000 description 4
- 230000002787 reinforcement Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 239000012141 concentrate Substances 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0893—Assignment of logical groups to network elements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/22—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks comprising specially adapted graphical user interfaces [GUI]
Abstract
The invention provides a dynamic scheduling method and a device for network slice service functions, wherein the basic idea is as follows: in the network slice operation, when the network flow changes, the virtual function resource capacity requirement of a period of time in the future is predicted according to the flow flowing through the virtual network function node on the slice, and the mapping relation of the slice is migrated according to the predicted requirement, so that the processing capacity of the slice is matched with the service requirement in real time. The invention can automatically train from historical flow data and historical node resource allowance without relying on manual experience, further quickly and efficiently generate the dynamic scheduling strategy of the slice service function, ensure that the slice processing capability can dynamically follow the flow fluctuation and the node cold and hot non-uniformity in the network, improve the time precision and the space precision of the slice service function scheduling, and realize the efficient service flexible supply of network resources.
Description
Technical Field
The invention belongs to the technical field of novel network architecture, and particularly relates to a dynamic scheduling method and device for network slice service functions.
Background
Various services in the current information network are mixed, the efficiency is greatly reduced, and as the information network becomes a daily infrastructure, the service diversity requirements are increased rapidly, and meanwhile, higher requirements are provided for network operation efficiency, isolation, automation, capability opening and the like. At this time, we have failed to meet the differentiated network requirements through one network. The slicing technology builds an end-to-end, customized and isolated logic network on demand on the same physical network, provides the combination of different functions, performances, costs and connection relations, supports independent operation and maintenance, can meet the performance requirements of different services, can realize the maximization of network resource utilization, saves the network construction cost, improves the profitability of operators, and finally achieves the balance of network service and cost benefit.
However, in the actual slice service provision process, the tolerance for service interruption is not high. Since the time of current virtual function instantiation is at least "minutes," passive network slice remapping strategies also require at least tens of seconds of operating time. How to dynamically adjust the physical resource scale occupied by the network slice according to the real-time resource demand of the user, and flexibly stretch the processing capacity of the network slice, so as to realize the on-demand matching of the network resource with the service demand is a great challenge.
Disclosure of Invention
Aiming at the problems that the prediction precision of the resource scheduling requirement is limited and the resource scheduling cannot be matched with the service requirement in real time, in order to meet the continuous and accurate matching requirement of the network slicing and the diversified service requirement, the invention discloses a dynamic network slicing service function scheduling method and device.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
the invention provides a dynamic scheduling method for network slice service functions, which comprises the following steps:
step 1: predicting future virtual network function resource requirements according to flow data flowing through virtual network functions in the network slices, and generating a slice resource capacity requirement view;
step 2: and dynamically adjusting the mapping condition of the slice virtual function through the calculation of an intelligent algorithm according to the resource capacity demand view corresponding to the network slice in real time and the resource allowance information of the service function node.
Further, the step 1 includes:
step 1.1: receiving historical change conditions of network slice flow and virtual topology connection conditions of the network slice, extracting spatial features and time features of historical flow loads of each node in the network through a neural network, and predicting future flow loads of the slice;
step 1.2: and realizing the conversion from the flow load to the resource demand by combining the neural network with the conversion relation between the future flow load of the slice and the resource capacity demand of the slice, and obtaining the future slice resource capacity demand view.
Further, the step 2 includes:
step 2.1: collecting and sorting service function node resource allowance information;
step 2.2: acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of an intelligent algorithm to an input layer of a neural network, and taking data of an output layer of the neural network as output actions after calculation of the neural network, namely, a function instance distribution adjustment strategy;
step 2.3: receiving a control strategy of an intelligent algorithm through a northbound interface, and converting the control strategy into a virtual function instance distribution adjustment strategy;
step 2.4: and converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network, and changing the mapping relation of the virtual functions.
Another aspect of the present invention provides a dynamic scheduling device for network slice service functions, including:
the slice resource capacity demand prediction unit is used for predicting the future virtual network function resource demand according to the flow data flowing through the virtual network function in the network slice, and generating a slice resource capacity demand view;
and the slice service function mapping unit is used for dynamically adjusting the mapping condition of the slice virtual function through the calculation of an intelligent algorithm according to the resource capacity demand view corresponding to the network slice in real time and the resource allowance information of the service function node.
Further, the slice resource capacity demand prediction unit includes:
the slice flow prediction module is used for receiving the historical change condition of the network slice flow and the virtual topology connection condition of the network slice, extracting the spatial characteristics and the temporal characteristics of the historical flow load of each node in the network through the neural network, and predicting the future flow load of the slice;
and the slice service function estimation module is used for realizing the conversion from the flow load to the resource demand by combining the neural network with the future flow load of the slice and the conversion relation between the slice flow and the resource capacity demand, and obtaining the future slice resource capacity demand view.
Further, the slice service function mapping unit includes:
the network information management module is used for collecting and arranging the resource allowance information of the service function node;
the intelligent algorithm module is used for acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of an intelligent algorithm to an input layer of the neural network, and taking data of an output layer of the neural network as output actions, namely a function instance distribution adjustment strategy after calculation of the neural network;
the strategy conversion module is used for receiving a control strategy of the intelligent algorithm through the northbound interface and converting the control strategy into a virtual function instance distribution adjustment strategy;
and the configuration module is used for converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network and changing the mapping relation of the virtual functions.
Compared with the prior art, the invention has the beneficial effects that:
the invention can automatically train from historical flow data and historical node resource allowance without relying on manual experience, further quickly and efficiently generate the dynamic scheduling strategy of the slice service function, ensure that the slice processing capability can dynamically follow the flow fluctuation and the node cold and hot non-uniformity in the network, improve the time precision and the space precision of the slice service function scheduling, and realize the efficient service flexible supply of network resources.
Drawings
FIG. 1 is a flowchart of a dynamic scheduling method for network slice service functions according to an embodiment of the present invention;
FIG. 2 is an explanatory diagram of the working principle of a dynamic scheduling device for network slice service function according to an embodiment of the present invention;
FIG. 3 is a functional explanatory diagram of a slice resource capacity demand prediction unit according to an embodiment of the present invention;
fig. 4 is an explanatory diagram of a working principle of a slice flow prediction module according to an embodiment of the present invention;
fig. 5 is an explanatory diagram of a working principle of a slice service function estimation module according to an embodiment of the present invention;
fig. 6 is an explanatory diagram of a working principle of a slice service function mapping unit according to an embodiment of the present invention.
Detailed Description
The invention is further illustrated by the following description of specific embodiments in conjunction with the accompanying drawings:
the overall design of the method is based on the idea that network slices decouple a physical network from a user network from a macroscopic view, and when network traffic changes, as shown in fig. 1, the mapping relationship of the slices is migrated to match the traffic demands in real time through predictive analysis of network traffic and elastic arrangement of virtual functions so as to support the slice scheduling demands in a time-varying network environment. The method specifically comprises the following steps:
step 1: and predicting future virtual network function resource demands according to the flow data flowing through the virtual network functions in the network slices, and generating a slice resource capacity demand view. Specifically, the prediction analysis is carried out on the service flow of the slice, a prediction model of network service on resource demand is established, and the scheduling demand of the slice resource is obtained, so that the network slice is matched with the severely fluctuating flow as required.
Step 2: and dynamically adjusting the mapping condition of the slice virtual function through the calculation of an intelligent algorithm according to the resource capacity demand view corresponding to the network slice in real time and the resource allowance information of the service function node.
Further, after step 2, the method may further include: and constructing pre-sensing and real-time adaptation between the state of the bottom network and the dynamic service requirement, realizing rapid migration of the slice, and ensuring continuous supply of the slice service.
Further, the step 1 includes:
step 1.1: and receiving the historical change condition of the network slice flow and the virtual topological connection condition of the network slice, extracting the spatial characteristics and the temporal characteristics of the historical flow load of each node in the network through the neural network, and predicting the future flow load of the slice. The flow prediction strategy based on space-time correlation is designed, the spatial characteristics and the temporal characteristics of the historical flow load of each node in the network are extracted through the neural network, and the future flow load of the node is predicted without manual intervention and prior information. The neural network model used includes, but is not limited to, a graph roll-up network (Graph Convolutional Network, GCN), a gated loop unit (Gated Recurrent Unit, GRU), etc.
Step 1.2: and the neural network is used for combining the future flow load of the slice and the conversion relation between the flow of the slice and the resource capacity demand to realize the conversion from the flow load to the resource demand, so as to obtain a future slice resource capacity demand view, and the number of virtual network function instances of each specification which need to be generated in the next time period is defined. Specifically, because the throughput of each type of function instance in the network at any moment and the processing requirement of the network traffic should be met, the optimal number of each type of function instance required at the next moment can be estimated only by counting the accumulated scaling condition when the predicted traffic reaches each type of function instance. The method specifically comprises the steps of converting flow load into resource demand by combining a neural network with a conversion relation between the flow load in the future of slicing and virtual function demand, outputting a prediction result, and obtaining a view of the capacity demand of the resource of the future slice. The result and the increment of the number of each function instance in the current network are the function instance estimation result to be processed at the next moment.
Further, the step 2 includes:
step 2.1: collecting and arranging service function node resource allowance information (including network flow information and network topology information); the method can be specifically realized through OpenFlow.
Step 2.2: acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of an intelligent algorithm to an input layer of a neural network, and taking data of an output layer of the neural network as output actions after calculation of the neural network, namely, a function instance distribution adjustment strategy; where the reward for the intelligent algorithm is the number of dormant nodes, the neural network model used includes, but is not limited to, deep reinforcement learning (Deep Reinforcement Learning, DRL). The step concentrates the data flow in the slicing network to fewer network devices and links as much as possible through the strategy of artificial intelligence, and improves the resource efficiency of the network nodes in space.
Step 2.3: receiving a control strategy of an intelligent algorithm through a northbound interface, and converting the control strategy into a virtual function instance distribution adjustment strategy.
Step 2.4: and converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network, and changing the mapping relation of the virtual functions.
In summary, the invention supports the pre-deployment of the network slices aiming at the continuously changing communication demands in the novel network, and can adjust the mapping relation between the slices and the physical network according to the demands so as to realize the resource high efficiency of the slice processing capacity.
On the basis of the embodiment, the invention designs the dynamic scheduling device for the network slice service function on the basis of the proposed dynamic scheduling method for the network slice service function, the working principle of the device is shown in figure 2, and the input of the device consists of three parts: 1) Slicing historical service flow information; 2) Structural information of the network; 2) Resource change information of the network node. In the process of slicing to provide services for users, the device combines the historical flow condition and the slicing network condition to calculate a slicing resource capacity demand view for the users at a certain moment in the future, and the number of virtual function instances of various specifications which need to be generated in the next time period is defined. And then, combining the network resource allowance condition of the slice to carry out elastic arrangement on the virtual functions in the slice, dynamically adjusting the mapping relation between the virtual functions and the physical network on the slice, and completing migration of the virtual functions. The whole process completes the on-demand efficient matching of network resources under the condition of ensuring the continuous supply of the service.
For facilitating further understanding of the present invention, the following describes in detail the dynamic scheduling apparatus for network slice service function according to the present invention with reference to fig. 2, and the apparatus is divided into 2 large units (slice resource capacity demand prediction unit, slice service function mapping unit), 6 large modules (slice flow prediction module, slice service function estimation module, network information management module, intelligent algorithm module, policy conversion module, configuration module), specifically as follows:
and the slice resource capacity demand prediction unit is used for predicting the future virtual network function resource demand according to the flow data flowing through the virtual network function in the network slice, and generating a slice resource capacity demand view. Specifically, a slice resource capacity demand prediction unit is loaded on a virtual network layer and a control plane where a network slice is positioned; taking fig. 3 as an example, the slice resource capacity demand prediction unit performs prediction analysis on the service flow of the slice, establishes a prediction model of network service on resource demand through the slice flow prediction module and the slice service function estimation module, and obtains the scheduling demand of the slice resource, so that the network slice matches with the severely-fluctuating flow as required.
Further, the slice resource capacity demand prediction unit includes:
the slice flow prediction module is used for receiving the historical change condition of the network slice flow and the virtual topology connection condition of the network slice, extracting the spatial characteristics and the temporal characteristics of the historical flow load of each node in the network through the neural network, and predicting the future flow load of the slice. Specifically, as shown in fig. 4, the module designs a flow prediction strategy based on space-time correlation, extracts spatial features and temporal features of historical flow load of each node in the network through a neural network, predicts future flow load of the node, and does not need manual intervention and prior information. The neural network model used includes, but is not limited to, a graph roll-up network (Graph Convolutional Network, GCN), a gated loop unit (Gated Recurrent Unit, GRU), etc.
And the slice service function estimation module is used for converting the flow load into the resource demand by combining the neural network with the future flow load of the slice and the conversion relation between the slice flow and the resource capacity demand, obtaining a future slice resource capacity demand view, and determining the number of virtual network function instances of each specification which need to be generated in the next time period. Specifically, as shown in fig. 5, the module further processes the predicted result of the network traffic, and since the throughput of each type of function instance in the network at any moment should meet the processing requirement of the network traffic, the module can estimate the optimal number of each function instance required at the next moment only by counting the accumulated scaling condition when the predicted traffic reaches each type of function instance. The method specifically comprises the steps of converting flow load into resource demand by combining a neural network with a conversion relation between the flow load in the future of slicing and virtual function demand, outputting a prediction result, and obtaining a view of the capacity demand of the resource of the future slice. The result and the increment of the number of each function instance in the current network are the function instance estimation result to be processed at the next moment.
Further, the slice service function mapping unit is configured to dynamically adjust the mapping condition of the slice virtual function according to the resource capacity demand view corresponding to the network slice in real time and the service function node resource allowance information through calculation of the intelligent algorithm. Specifically, the present invention loads slice service function mapping units on a control plane and an intelligent algorithm plane above a virtual network layer where a network slice is located. Taking fig. 6 as an example, after the slice resource capacity demand prediction unit obtains the slice resource capacity demand view, the slice resource capacity demand prediction unit is used as input of the intelligent algorithm module of the slice service function mapping unit together with the service function node resource allowance information obtained by the network information management module. The resource adjustment strategy calculated by the intelligent algorithm module forms a node performance index through the strategy conversion module and guides the configuration module to update the routing strategy of the physical network.
Further, the slice service function mapping unit includes:
and the network information management module is used for collecting and arranging the service function node resource allowance information (including network flow information and network topology information). Specifically, the invention loads a network information management module on a control plane. The module is mainly responsible for collecting and arranging data such as network flow information, network topology information, node resource allowance information and the like. The function implementation can be realized through OpenFlow.
The intelligent algorithm module is used for acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of the intelligent algorithm to an input layer of the neural network, and taking data of an output layer of the neural network as output actions, namely a function instance distribution adjustment strategy after calculation of the neural network. Specifically, the invention loads an intelligent algorithm module on the control plane. Continuing with the example of FIG. 6, the intelligent algorithm module exercises control over the entire process of slice service function mapping. The module acquires the slice resource capacity demand view generated by the slice resource capacity demand prediction unit and the service function node resource allowance acquired by the network information management module in real time. The module then transmits both of these information as input data for the intelligent algorithm to the input layer of its neural network. After the calculation of the neural network, the data of the neural network output layer is used as output action, namely, the function instance distribution adjustment strategy is transmitted to the strategy conversion module of the control plane. Where the reward for the intelligent algorithm is the number of dormant nodes, the neural network model used includes, but is not limited to, deep reinforcement learning (Deep Reinforcement Learning, DRL). The module concentrates the data flow in the slicing network to fewer network devices and links as much as possible through an artificial intelligence strategy, and improves the resource efficiency of the network nodes in space.
The policy conversion module is used for receiving the control policy of the intelligent algorithm module through the northbound interface and converting the control policy into a virtual function instance distribution adjustment policy. Specifically, the invention loads a policy conversion module on a control plane.
And the configuration module is used for converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network and changing the mapping relation of the virtual functions. Specifically, the present invention loads a configuration module in the control plane.
In sum, through the system operation of the slice resource capacity demand prediction unit and the slice service function mapping unit and the cooperative operation of the slice flow prediction module, the slice service function estimation module, the network information management module, the intelligent algorithm module, the strategy conversion module and the configuration module, the invention can automatically train from historical flow data and historical node resource allowance without depending on manual experience, further quickly and efficiently generate a slice service function dynamic scheduling strategy, ensure that slice processing capacity can dynamically follow the characteristics of flow fluctuation and node cold and hot unevenness in a network, improve time precision and space precision of slice service function scheduling, and realize efficient service flexible supply of network resources.
The foregoing is merely illustrative of the preferred embodiments of this invention, and it will be appreciated by those skilled in the art that changes and modifications may be made without departing from the principles of this invention, and it is intended to cover such modifications and changes as fall within the true scope of the invention.
Claims (2)
1. The dynamic scheduling method for the network slice service function is characterized by comprising the following steps of:
step 1: predicting future virtual network function resource requirements according to flow data flowing through virtual network functions in the network slices, and generating a slice resource capacity requirement view;
step 2: dynamically adjusting the mapping condition of the virtual functions of the slice through the calculation of an intelligent algorithm according to the resource capacity demand view corresponding to the network slice in real time and the resource allowance information of the service function node;
the step 1 comprises the following steps:
step 1.1: receiving historical change conditions of network slice flow and virtual topology connection conditions of the network slice, extracting spatial features and time features of historical flow loads of each node in the network through a neural network, and predicting future flow loads of the slice;
step 1.2: the neural network is used for combining the future flow load of the slice and the conversion relation between the slice flow and the resource capacity demand to realize the conversion from the flow load to the resource demand, so as to obtain a future slice resource capacity demand view;
the step 2 comprises the following steps:
step 2.1: collecting and sorting service function node resource allowance information;
step 2.2: acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of an intelligent algorithm to an input layer of a neural network, and taking data of an output layer of the neural network as output actions after calculation of the neural network, namely, a function instance distribution adjustment strategy;
step 2.3: receiving a control strategy of an intelligent algorithm through a northbound interface, and converting the control strategy into a virtual function instance distribution adjustment strategy;
step 2.4: and converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network, and changing the mapping relation of the virtual functions.
2. A dynamic scheduling device for network slice service functions, comprising:
the slice resource capacity demand prediction unit is used for predicting the future virtual network function resource demand according to the flow data flowing through the virtual network function in the network slice, and generating a slice resource capacity demand view;
the slice service function mapping unit is used for dynamically adjusting the mapping condition of the slice virtual function through the calculation of an intelligent algorithm according to the resource capacity demand view corresponding to the network slice in real time and the resource allowance information of the service function node;
the slice resource capacity demand prediction unit includes:
the slice flow prediction module is used for receiving the historical change condition of the network slice flow and the virtual topology connection condition of the network slice, extracting the spatial characteristics and the temporal characteristics of the historical flow load of each node in the network through the neural network, and predicting the future flow load of the slice;
the slice service function estimation module is used for realizing the conversion from the flow load to the resource demand by combining the neural network with the future flow load of the slice and the conversion relation between the slice flow and the resource capacity demand, and obtaining a future slice resource capacity demand view;
the slice service function mapping unit includes:
the network information management module is used for collecting and arranging the resource allowance information of the service function node;
the intelligent algorithm module is used for acquiring a slice resource capacity demand view and a service function node resource allowance in real time, transmitting the two kinds of information as input data of an intelligent algorithm to an input layer of the neural network, and taking data of an output layer of the neural network as output actions, namely a function instance distribution adjustment strategy after calculation of the neural network;
the strategy conversion module is used for receiving a control strategy of the intelligent algorithm through the northbound interface and converting the control strategy into a virtual function instance distribution adjustment strategy;
and the configuration module is used for converting the virtual function instance distribution adjustment strategy into configuration information of the bottom network and changing the mapping relation of the virtual functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210448065.7A CN114938335B (en) | 2022-04-27 | 2022-04-27 | Dynamic scheduling method and device for network slice service function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210448065.7A CN114938335B (en) | 2022-04-27 | 2022-04-27 | Dynamic scheduling method and device for network slice service function |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114938335A CN114938335A (en) | 2022-08-23 |
| CN114938335B true CN114938335B (en) | 2024-01-26 |
Family
ID=82861464
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210448065.7A Active CN114938335B (en) | 2022-04-27 | 2022-04-27 | Dynamic scheduling method and device for network slice service function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114938335B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110275758A (en) * | 2019-05-09 | 2019-09-24 | 重庆邮电大学 | A kind of virtual network function intelligence moving method |
| CN112910710A (en) * | 2021-02-08 | 2021-06-04 | 清华大学 | Network flow space-time prediction method and device, computer equipment and storage medium |
| CN113542002A (en) * | 2021-05-27 | 2021-10-22 | 广州杰赛科技股份有限公司 | Slice scheduling method, device, equipment and storage medium for 5G wireless access network |
| CN113743594A (en) * | 2021-09-30 | 2021-12-03 | 中国联合网络通信集团有限公司 | Network flow prediction model establishing method and device, electronic equipment and storage medium |
| CN113923129A (en) * | 2021-09-08 | 2022-01-11 | 中国人民解放军战略支援部队信息工程大学 | VNF demand prediction method and system based on data driving |
| CN113992527A (en) * | 2021-10-26 | 2022-01-28 | 郑州轻工业大学 | Network service function chain online migration method and system |
-
2022
- 2022-04-27 CN CN202210448065.7A patent/CN114938335B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110275758A (en) * | 2019-05-09 | 2019-09-24 | 重庆邮电大学 | A kind of virtual network function intelligence moving method |
| CN112910710A (en) * | 2021-02-08 | 2021-06-04 | 清华大学 | Network flow space-time prediction method and device, computer equipment and storage medium |
| CN113542002A (en) * | 2021-05-27 | 2021-10-22 | 广州杰赛科技股份有限公司 | Slice scheduling method, device, equipment and storage medium for 5G wireless access network |
| CN113923129A (en) * | 2021-09-08 | 2022-01-11 | 中国人民解放军战略支援部队信息工程大学 | VNF demand prediction method and system based on data driving |
| CN113743594A (en) * | 2021-09-30 | 2021-12-03 | 中国联合网络通信集团有限公司 | Network flow prediction model establishing method and device, electronic equipment and storage medium |
| CN113992527A (en) * | 2021-10-26 | 2022-01-28 | 郑州轻工业大学 | Network service function chain online migration method and system |
Non-Patent Citations (5)
| Title |
|---|
| Dynamic Network Slice Scaling Assisted by Prediction in 5G Network;Jinhe Zhou et al.;IEEE;全文 * |
| Publisher: IEEE Cite This PDF;Ziyue Luo et al.;IEEE;全文 * |
| 基于强化学习的5G网络切片虚拟网络功能迁移算法;唐伦;周钰;谭颀;魏延南;陈前斌;;电子与信息学报(03);全文 * |
| 基于流量感知的虚拟网络功能智能编排算法研究;周钰;《中国优秀硕士学位论文全文数据库》;全文 * |
| 虚拟网络功能编排方法研究;袁泉;《中国博士学位论文全文数据库》;正文第4章 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114938335A (en) | 2022-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Intelligent edge computing for IoT-based energy management in smart cities | |
| Chen et al. | Intelligent traffic adaptive resource allocation for edge computing-based 5G networks | |
| CN102256266B (en) | User application-oriented adaptive access network selection device and method | |
| Wang et al. | AI-based cloud-edge-device collaboration in 6G space-air-ground integrated power IoT | |
| CN106934497B (en) | Intelligent community power consumption real-time prediction method and device based on deep learning | |
| Chen et al. | Real-time energy trading and future planning for fifth generation wireless communications | |
| CN113452956B (en) | Intelligent distribution method and system for power transmission line inspection tasks | |
| Piovesan et al. | Joint load control and energy sharing for renewable powered small base stations: A machine learning approach | |
| CN112887142B (en) | Method for generating web application slice in virtualized wireless access edge cloud | |
| CN111654067A (en) | Novel control method and device for power distribution system scheduling in transmission and distribution coordination | |
| CN114422349A (en) | Cloud-edge-end-collaboration-based deep learning model training and reasoning architecture deployment method | |
| CN108834079A (en) | A kind of load balance optimization method based on mobility prediction in heterogeneous network | |
| CN111787123A (en) | Intelligent heat supply network operation and maintenance management system | |
| CN113766544A (en) | Multi-edge cooperation-based power Internet of things slice optimization method | |
| CN115314343A (en) | Source-load-storage resource aggregation control gateway device and load and output prediction method | |
| Muhammad et al. | Deep-reinforcement-learning-based sustainable energy distribution for wireless communication | |
| Li et al. | Communication and computation resource allocation and offloading for edge intelligence enabled fault detection system in smart grid | |
| Temesgene et al. | Dynamic control of functional splits for energy harvesting virtual small cells: A distributed reinforcement learning approach | |
| Wang et al. | Reinforcement learning-based optimization for mobile edge computing scheduling game | |
| Zheng et al. | Learning based task offloading in digital twin empowered internet of vehicles | |
| CN114938335B (en) | Dynamic scheduling method and device for network slice service function | |
| CN114885388A (en) | Multi-service type self-adaptive switching judgment method combined with RSS prediction | |
| Yuan et al. | FedTSE: Low-cost federated learning for privacy-preserved traffic state estimation in IoV | |
| Zorello et al. | Baseband-function placement with multi-task traffic prediction for 5G radio access networks | |
| Tan et al. | Uav control in smart city based on space-air-ground integrated network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |