CN111245718B - Routing optimization method based on SDN context awareness - Google Patents

Routing optimization method based on SDN context awareness Download PDF

Info

Publication number
CN111245718B
CN111245718B CN201911400370.3A CN201911400370A CN111245718B CN 111245718 B CN111245718 B CN 111245718B CN 201911400370 A CN201911400370 A CN 201911400370A CN 111245718 B CN111245718 B CN 111245718B
Authority
CN
China
Prior art keywords
network
sdn
context awareness
context
neural network
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
Application number
CN201911400370.3A
Other languages
Chinese (zh)
Other versions
CN111245718A (en
Inventor
李传煌
庄丹娜
唐豪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Gongshang University
Original Assignee
Zhejiang Gongshang University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang Gongshang University filed Critical Zhejiang Gongshang University
Priority to CN201911400370.3A priority Critical patent/CN111245718B/en
Publication of CN111245718A publication Critical patent/CN111245718A/en
Application granted granted Critical
Publication of CN111245718B publication Critical patent/CN111245718B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H04L45/123—Evaluation of link metrics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/14—Routing performance; Theoretical aspects

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Mathematical Physics (AREA)
  • Biophysics (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Probability & Statistics with Applications (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a routing optimization method based on SDN context awareness, which comprises the following steps: step 1: collecting related scene data of a network and a user, and establishing a network scene data set; step 2: constructing a three-level cascade neural network model consisting of a gate control circulation unit, a self-encoder and a multilayer sensing mechanism; and step 3: inputting the network context data set in the step 1 into a trained graph neural network model for context awareness to obtain a network performance matrix after the context awareness; and 4, step 4: and (4) optimizing the network route through an intelligent algorithm according to the network performance matrix obtained in the step (3). Compared with other route optimization methods, the route optimization method based on SDN context awareness can realize intelligent fitting of actual context and state requirements and improve the resource utilization rate of a network.

Description

Routing optimization method based on SDN context awareness
Technical Field
The invention relates to a context network technology, a deep learning technology and a route optimization technology, in particular to a route optimization method based on SDN context awareness.
Background
With the continuous development and innovation of network technology, network traffic shows an explosive growth trend, and services are more and more diversified. The existing internet takes a distributed theory as a foundation stone, and develops three baseline technologies of transmission, forwarding and routing based on the distributed theory, so that the existing internet has high expandability and robustness. However, the existing IP architecture has single resource management, and its routing mechanism and best-effort service features cannot adapt to dynamic changes of demand and environment, so that the inherent internet architecture cannot meet flexible and variable application requirements.
Route control and optimization is an important aspect of improving network quality of service. Routing affects many key performance metrics associated with the network, such as throughput, latency, and utilization. The complex and variable network scenario makes it increasingly difficult to implement timely and reliable network services, and effective resource allocation and scheduling techniques are needed to solve this problem. Accurate network states and situations obtained through means of analysis, prediction, adjustment and the like can optimize the route forwarding service, so that resource waste is reduced, the overall performance of the network is improved, and diversified application service requirements are met.
The concept of the 'scenario network' is firstly proposed by information engineering university of China Wu Jiang Xing academy of China, Wu Jiang province in 2015, the design criteria of traditional network best effort causes a single network structure, and the problems of low flexibility, poor efficiency, low experience and the like exist. The segment is the 'scene', namely the sum of the network service object and the environment background, the scene network emphasizes 'taking users as the center', actively adopts the scene fitting and learning technology aiming at the scene, and changes the network function and structure to provide more appropriate service for the users.
The Software Defined Network (SDN) has more flexible network control and implementation capability, and the centralized verification type automatic configuration and dynamic global control of the SDN can provide a new platform for realizing network innovation technology. Therefore, the network context is sensed through deep learning in the SDN, and then the network routing is optimized by adopting an intelligent algorithm, so that intelligent fitting of actual context and state requirements can be realized, and the resource utilization rate of the network is improved.
Disclosure of Invention
Aiming at the technical shortage of the current intelligent context awareness-based network routing optimization, the invention provides a routing optimization method based on SDN context awareness, which aims to accurately evaluate and predict routing constraints such as network performance and the like based on a deep learning method, and replans the network routing through an intelligent algorithm on the basis of an obtained network performance matrix, so that the purposes of fitting business service requirements and improving the utilization rate of network resources are achieved.
The technical scheme adopted by the invention for solving the technical problem is as follows: a routing optimization method based on SDN context awareness comprises the following steps:
step 1: collecting related scene data of a network and a user, and establishing a network scene data set;
step 2: and constructing a graph neural network model which is a three-level cascade neural network model formed by a gated loop unit (GRU), a self-encoder and a multilayer perception mechanism (deep neural network, MLP).
And step 3: training the graph neural network model by adopting an open-source data set, and inputting the network context data set in the step 1 into the trained graph neural network model for context awareness to obtain a network performance matrix after the context awareness;
and 4, step 4: and (4) optimizing the network route through an intelligent algorithm according to the network performance matrix obtained in the step (3). The intelligent algorithm adopts a MADDPG algorithm, a DDPG intelligent body carries out route optimization on a network topological structure by taking perceived time delay and available bandwidth in a network as constraints, and finally provides an optimized route strategy RC for SDN route forwardingt. The specific optimization process is as follows:
and (4.1) taking the network topology C, the traffic state T, the available bandwidth B and the routing configuration RC as input, and outputting the modeled delay data D after the processing by the context awareness model.
And (4.2) abstracting the sensing result and the network basic information into environment state information S, and performing action selection by the DDPG intelligent body according to the obtained state information. The calculation principle of the reward value is set according to task objectives, in the process of network route optimization, a link with low selection delay should obtain high reward, and the reward should be given to the selection with large available bandwidth of the link. Setting the calculation formula of the reward value as R according to the solution thought of the multi-constraint problemt＝-α*Dt+β*Bt
Wherein α, β are constant coefficients, α + β ═ 1, where DtThe reason why the delay value is set to be negative is that the higher the delay is, the smaller the reward is, and the smaller the delay is, the higher the reward value is, i.e., the magnitude of the reward value is inversely proportional to the magnitude of the delay. B istIndicating the size of the available bandwidth of the link, it is clear that the value of the available bandwidth should be proportional to the acquisition of the bonus value, which should be greater the available bandwidth.
Further, the open-source data set adopts a KDN (knowledge Defined networking) network modeling data set.
Further, the network context data set in step 1 includes elements of topology, traffic, routing configuration, and performance indicators, where the performance indicators include delay, jitter, packet loss, and the like.
Further, the graph neural network model in step 2 is based on an MPNN message passing mechanism and a RouteNet architecture, and senses the context state in the network.
Further, the three-level cascade neural network model structure is as follows:
the first layer, namely the message transmission stage, senses the state of links and paths in the SDN through a GRU neural network, the layer takes a scene information feature vector in the network as input, then captures the dependency among the links, the paths and the network routing topology through a GRU neural unit in a hidden layer, and then calculates and updates the state information in the network graph.
And a second layer of self-encoder reconstructs the output obtained by the GRU neural unit and extracts abstract features, the layer is an unsupervised learning process, training can be completed without knowing the label of data, and the layer can extract and reconstruct the features of input data under the unsupervised condition.
The third layer is similar to the hidden structure of the multilayer sensing mechanism to read the state information, and a final result is obtained.
The invention has the beneficial effects that: the practice of optimizing network routing by adopting intelligent context awareness is less at present, and the routing optimization method based on SDN context awareness can realize intelligent fitting on actual context and state requirements and improve the resource utilization rate of a network; the invention constructs a mixed model of a graph neural network and a self-encoder, shows strong reasoning capability of a data format based on a non-Euclidean space, and fully utilizes the situation of a node; the intelligent algorithm of the invention adopts MADDPG algorithm, and effectively deals with the interaction of multiple agents.
Drawings
FIG. 1 is a diagram of overall route optimization;
FIG. 2 is a diagram of a context aware architecture;
FIG. 3 is a flow chart of route optimization;
fig. 4 is a graph of the average delay contrast for different optimization schemes.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
The network context is sensed through deep learning in the SDN network, and then the network routing is optimized through an intelligent algorithm, so that intelligent fitting of actual context and state requirements can be achieved, and the resource utilization rate of the network is improved. The practice of optimizing network routing by adopting intelligent context awareness is less at present, so the invention aims to accurately evaluate and predict routing constraints such as network performance and the like based on a deep learning method, and replans the network routing by an intelligent algorithm on the basis of an obtained network performance matrix, thereby achieving the purposes of meeting business service requirements and improving the utilization rate of network resources.
The invention adopts the idea of point-to-surface in the network route optimization, realizes the global optimization by optimizing the local network, and can lead the nodes to adjust the network forwarding strategy according to the network situation of the nodes, thereby achieving the purpose of optimizing the network performance.
As shown in fig. 1, the present invention provides a routing optimization method based on SDN context awareness, which includes the following steps:
step 1: collecting relevant scene data of a network and a user, and establishing a network scene data set, wherein the network scene data set comprises the elements of topology, flow, routing configuration and performance indexes (such as delay, jitter, packet loss and the like);
step 2: and (3) constructing a graph neural network model, wherein the graph neural network model is a three-level cascade neural network model formed by a gated loop unit (GRU), a self-encoder and a multilayer perception mechanism, as shown in fig. 2.
The first layer, namely the message transmission stage, senses the state of links and paths in the SDN through a GRU neural network, the layer takes a scene information feature vector in the network as input, then captures the dependency among the links, the paths and the network routing topology through a GRU neural unit in a hidden layer, and then calculates and updates the state information in the network graph.
And a second layer of self-encoder reconstructs the output obtained by the GRU neural unit and extracts abstract features, the layer is an unsupervised learning process, training can be completed without knowing the label of data, and the layer can extract and reconstruct the features of input data under the unsupervised condition.
The third layer is similar to the hidden structure of the multilayer sensing mechanism to read the state information, and a final result is obtained.
And step 3: training the graph neural network model by adopting an open-source data set, and inputting the network context data set in the step 1 into the trained graph neural network model for context awareness to obtain a network performance matrix after the context awareness;
and 4, step 4: and (4) optimizing the network route through an intelligent algorithm according to the network performance matrix obtained in the step (3). The intelligent algorithm adopts a MADDPG algorithm, a DDPG intelligent body carries out route optimization on a network topological structure by taking perceived time delay and available bandwidth in a network as constraints, and finally provides an optimized route strategy RC for SDN route forwardingt. The specific optimization process is as follows:
and (4.1) taking the network topology C, the traffic state T, the available bandwidth B and the routing configuration RC as input, and outputting the modeled delay data D after the processing by the context awareness model.
And (4.2) abstracting the sensing result and the network basic information into environment state information S, and performing action selection by the DDPG intelligent body according to the obtained state information. The calculation principle of the reward value is set according to task objectives, in the process of network route optimization, a link with low selection delay should obtain high reward, and the reward should be given to the selection with large available bandwidth of the link. According to the solution thought of the multi-constraint problem, the reward is givenThe value calculation formula is set to Rt＝-a*Dt+β*Bt
Wherein α, β are constant coefficients, α + β ═ 1, where DtThe reason why the delay value is set to be negative is that the higher the delay is, the smaller the reward is, and the smaller the delay is, the higher the reward value is, i.e., the magnitude of the reward value is inversely proportional to the magnitude of the delay. B istIndicating the size of the available bandwidth of the link, it is clear that the value of the available bandwidth should be proportional to the acquisition of the bonus value, which should be greater the available bandwidth.
The present invention will be further described with reference to the following examples.
The invention adopts an open-source KDN (knowledge Defined networking) network modeling data set generated by OMNet + + by Krzysztoff Rusek et al, which is composed of data collected under three different network topologies of NSFNet, GEANT2 and SYNTH50, wherein the data set comprises basic network scene elements: topology, traffic, routing configuration, performance indicators (e.g., delay, jitter, packet loss, etc.). The data set is composed of a routing configuration file and a network state file, wherein the routing configuration file is a matrix, the network state file stores data in a key value pair mode, and characteristic values can be obtained according to parameter indexes. According to the data set, sample data is stored and read through the TFRecord format in the tensrflow framework.
The routing optimization algorithm based on SDN intelligent context awareness is as follows:
Figure BDA0002347318140000051
Figure BDA0002347318140000061
as shown in fig. 3, when optimizing a network, taking a path in the network as an example, first, a context awareness model calculates a state information matrix according to obtained basic information by link analysis, and then the DDPG performs action selection according to a calculation result of context awareness and a source-destination pair of the path, where each state is a Markov transition process with a finite state, and the action selection is to select a next-hop link. The selection of a series of actions between source-destination pairs is therefore a satisfactory routing strategy.
As shown in fig. 4, RouteNet decreases the latency by 18.39% and GGAE decreases by 25.05% at TI of 12. At TI 15, RouteNet and GGAE reduced the delay by 27.42%, 33.56%, respectively. Comparing experimental data, the situation awareness model can be found to have excellent performance in a routing optimization scheme based on the situation awareness due to the strong data modeling capability of the situation awareness model. Compared with a shortest path algorithm and a RouteNet algorithm, the established context awareness model is remarkably shown in a scheme of carrying out routing optimization through SDN intelligent context awareness, and network delay can be effectively reduced through predicted data.
The above-described embodiments are intended to illustrate rather than to limit the invention, and any modifications and variations of the present invention are within the spirit of the invention and the scope of the appended claims.

Claims (3)

1. A routing optimization method based on SDN context awareness is characterized by comprising the following steps:
step 1: collecting related scene data of a network and a user, and establishing a network scene data set; the network scenario data set comprises the elements of topology, flow, routing configuration and performance indexes, wherein the performance indexes comprise delay, jitter and packet loss;
step 2: constructing a graph neural network model, wherein the graph neural network model is a three-level cascade neural network model structure formed by a gated cyclic unit (GRU), a self-encoder and a multilayer perception mechanism, and comprises the following steps:
a first layer, namely a message transmission stage, performs state perception on links and paths in the SDN through a GRU neural network, wherein the layer takes a context information feature vector in the SDN as input, then grasps a dependency relationship among the links, the paths and a network routing topology through a GRU neural unit in a hidden layer of the GRU neural network, and then calculates and updates state information of the paths and the links in the network;
the second layer of self-encoder reconstructs the output obtained by the GRU neural unit and extracts abstract features, the layer is an unsupervised learning process, training can be completed without knowing the label of data, and the layer can extract and reconstruct the features of input data under the unsupervised condition;
the third layer adopts a hidden structure of a multilayer sensing mechanism to read the state information to obtain a final result;
and step 3: training the graph neural network model by adopting an open-source data set, and inputting the network context data set in the step 1 into the trained graph neural network model for context awareness to obtain a network performance matrix after the context awareness;
and 4, step 4: optimizing the routing of the network through an intelligent algorithm according to the network performance matrix obtained in the step 3; the intelligent algorithm adopts a multi-main-body depth certainty strategy gradient algorithm MADDPG, the DDPG intelligent body carries out route optimization on a network topological structure by taking the perceived time delay and available bandwidth in the network as constraints, and finally provides an optimized route strategy RC for SDN route forwardingt(ii) a The specific optimization process is as follows:
(4.1) taking a network topology C, a traffic state T, an available bandwidth B and a routing configuration RC as inputs, and outputting modeled delay data D after processing the inputs through a context awareness model;
(4.2) abstracting the sensing result and the network basic information into environment state information S, and performing action selection by the DDPG intelligent agent according to the obtained state information; according to the solution thought of the multi-constraint problem, the calculation formula of the reward value is as follows, and the value of the reward value is inversely proportional to the delay value;
Rt＝-α*Dt+β*Bt
wherein α, β are constant coefficients, α + β ═ 1, where DtRepresenting a delay value, BtThe value of the available bandwidth is proportional to the acquisition of the reward value, and the larger the available bandwidth is, the larger the reward value is.
2. The SDN context awareness-based route optimization method of claim 1, wherein the open-source dataset is a KDN (knowledge Defined networking) network modeling dataset.
3. The SDN context awareness-based route optimization method according to claim 1, wherein in step 2, the graph neural network model is based on a message passing mechanism of a message passing neural network MPNN and a RouteNet architecture, and perceives a context state in the network.
CN201911400370.3A 2019-12-30 2019-12-30 Routing optimization method based on SDN context awareness Active CN111245718B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911400370.3A CN111245718B (en) 2019-12-30 2019-12-30 Routing optimization method based on SDN context awareness

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911400370.3A CN111245718B (en) 2019-12-30 2019-12-30 Routing optimization method based on SDN context awareness

Publications (2)

Publication Number Publication Date
CN111245718A CN111245718A (en) 2020-06-05
CN111245718B true CN111245718B (en) 2022-03-29

Family

ID=70879495

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911400370.3A Active CN111245718B (en) 2019-12-30 2019-12-30 Routing optimization method based on SDN context awareness

Country Status (1)

Country Link
CN (1) CN111245718B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112202672B (en) * 2020-09-17 2021-07-02 华中科技大学 Network route forwarding method and system based on service quality requirement
CN113158543B (en) * 2021-02-02 2023-10-24 浙江工商大学 Intelligent prediction method for software defined network performance
CN113156979B (en) * 2021-05-27 2022-09-06 浙江农林大学 Forest guard patrol path planning method and device based on improved MADDPG algorithm
CN113489654B (en) * 2021-07-06 2024-01-05 国网信息通信产业集团有限公司 Routing method, device, electronic equipment and storage medium
CN114362175B (en) * 2022-03-10 2022-06-07 山东大学 Wind power prediction method and system based on depth certainty strategy gradient algorithm
CN115225561B (en) * 2022-08-15 2022-12-06 南京邮电大学 Route optimization method and system based on graph structure characteristics
CN116055378B (en) * 2023-01-10 2024-05-28 中国联合网络通信集团有限公司 Training method and device for traffic scheduling strategy generation model
CN118214708A (en) * 2024-05-15 2024-06-18 陕西智网驿成信息科技有限公司 Routing configuration strategy intelligent analysis system based on unsupervised learning

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109218049A (en) * 2017-06-30 2019-01-15 华为技术有限公司 A kind of control method, relevant device and system
CN110611619A (en) * 2019-09-12 2019-12-24 西安电子科技大学 Intelligent routing decision method based on DDPG reinforcement learning algorithm

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109218049A (en) * 2017-06-30 2019-01-15 华为技术有限公司 A kind of control method, relevant device and system
CN110611619A (en) * 2019-09-12 2019-12-24 西安电子科技大学 Intelligent routing decision method based on DDPG reinforcement learning algorithm

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Learning and Generating Distributed Routing Protocols Using Graph-based Deep Learning;Geyer F等;《Proc of the 2018 Workshop on Big Data Analytics and Machine Learning for Data Communication Networks》;20180820;第40-45页 *
基于深度强化学习的软件定义网络QoS优化;兰巨龙等;《通信学报》;20191207;第40卷(第12期);第60-67页 *

Also Published As

Publication number Publication date
CN111245718A (en) 2020-06-05

Similar Documents

Publication Publication Date Title
CN111245718B (en) Routing optimization method based on SDN context awareness
Liu et al. DRL-R: Deep reinforcement learning approach for intelligent routing in software-defined data-center networks
CN113158543B (en) Intelligent prediction method for software defined network performance
CN112491714B (en) Intelligent QoS route optimization method and system based on deep reinforcement learning in SDN environment
Yao et al. Machine learning aided load balance routing scheme considering queue utilization
Quang et al. Multi-domain non-cooperative VNF-FG embedding: A deep reinforcement learning approach
Kim et al. Ant colony based self-adaptive energy saving routing for energy efficient Internet
CN112650581A (en) Cloud-side cooperative task scheduling method for intelligent building
Sun et al. Accelerating convergence of federated learning in MEC with dynamic community
Xu et al. An actor-critic-based transfer learning framework for experience-driven networking
Cong et al. A deep reinforcement learning-based multi-optimality routing scheme for dynamic IoT networks
Li et al. Traffic modeling and optimization in datacenters with graph neural network
CN114710439B (en) Network energy consumption and throughput joint optimization routing method based on deep reinforcement learning
Zhou et al. Multi-task deep learning based dynamic service function chains routing in SDN/NFV-enabled networks
Saravanan et al. Energy efficient optimization algorithms for MANET
Zhang et al. DSOQR: Deep Reinforcement Learning for Online QoS Routing in SDN‐Based Networks
Yang et al. Comparative analysis of routing schemes based on machine learning
CN104009916A (en) Delay tolerant network energy-efficient routing scheme based on social attribute forwarding
Dai et al. Iqor-lse: An intelligent qos on-demand routing algorithm with link state estimation
Liu et al. Scalable deep reinforcement learning-based online routing for multi-type service requirements
Vendramin et al. CGrAnt: a swarm intelligence-based routing protocol for delay tolerant networks
Ren et al. End-to-end network SLA quality assurance for C-RAN: a closed-loop management method based on digital twin network
CN111555978B (en) SDN routing arrangement method with energy saving and service quality guarantee functions in smart grid
Rao et al. A deep learning-based constrained intelligent routing method
CN115499365A (en) Route optimization method, device, equipment and medium

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