Font Size: a A A

Personality-aware VNF Deployment And Routing

Posted on:2021-02-04Degree:MasterType:Thesis
Country:ChinaCandidate:R M YangFull Text:PDF
GTID:2428330620968135Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Network function virtualization is an emerging technology that decouples network functions from dedicated hardware devices,increasing the flexibility and scalability of network services.For profitable virtual network function(VNF)providers,the major challenge they face is how to design effective VNF deployment and routing schemes to optimize profits.In this paper,we first build a user satisfaction model to predict the personalized service demand of users.Then,based on personalized service demand,we design a profit-driven VNF deployment and routing algorithm to maximize the profits of VNF providers while ensuring user satisfaction.The main research contributions of this paper are as follows.1.This paper explores the influence of user personality on the preference between service quality and service price through questionnaires.Then,based on questionnaires,this paper constructs a user satisfaction prediction model to achieve the prediction of user satisfaction with the given user personality,service quality,and service price.The model can also predict the service demand of users with different personality for service quality and service price.2.This paper proposes a genetic algorithm-based VNF deployment scheme.First,we formulate the VNF deployment problem.Then,we adopt the genetic algorithm to obtain the optimal VNF deployment scheme for the VNF provider to maximize the profit.3.This paper presents a reinforcement learning-based routing selection scheme.First,we formulate the routing selection problem.Then,we use the reinforcement learning method to obtain the optimal routing selection scheme for the VNF provider to maximize the profit.The experimental results show that our user satisfaction prediction model can accurately predict user satisfaction.The genetic algorithm-based VNF deployment algorithm and reinforcement learning-based routing algorithm presented in this paper can increase the profit by at least 198.1%,76.8% and 38.0% respectively as compared to the Random algorithm,QAAV(Qo S-aware Adaptive Allocation of VNF)algorithm and RSA(Restrictive Search Algorithm).
Keywords/Search Tags:Profit Optimization, User Personality, User Satisfaction, VNF Deployment, Routing Selection, Genetic Algorithm, Reinforcement Learning
PDF Full Text Request
Related items