Research On Service Performance Analysis Model And Delay Deterministic Transmission Algorithm Of Edge Networks

With the rapid development of the Internet of Things and mobile Internet,diversified services have higher and higher requirements for communication networks,such as high bandwidth,high reliability,low delay and other diversified requirements.Therefore,MEC,SDN and TSN come into being.The introduction of SDN technology can bring more flexibility to the resource allocation of MEC networks,but also leads to the randomness of service routing.The application of TSN technology in MEC networks can provide a deterministic Qo S guarantee for time-triggered traffic in MEC.However,in order to preserve the flexibility of TSN technology,the TSN working group did not stipulate its routing and scheduling policy.Therefore,the main researches of this thesis are the performance analysis models of services in MEC networks with the introduction of SDN technology and the delay deterministic transmission algorithms based on joint optimization of routing and scheduling.The main work of this thesis is as follows:Firstly,this thesis establishes analysis models for service flows in the MEC network based on SNC.Taking actual communication scenarios into consideration,there are Vo IP service,video service and FTP service in MEC networks with decreasing priority.Therefore this thesis uses SNC to establish the analysis models for these three service flows respectively.Due to the random nature of routing in the MEC network,probability parameters are introduced into derived expressions in the proposed analysis models to characterize this randomness.Based on the analysis models,this thesis derives arrival processes,interference processes and service processes of these three data-flows.And then the closed-form expressions showing the the relationship between delay,backlog upper bounds and violation probabilities of the data-flows are derived.The simulation and analysis results show that the delay,backlog upper bounds of data flows are monotonically increasing with the length of routing and only affected by the change of the probability parameters of the interference flows with higher priority than the through flow.Secondly,this thesis proposes a delay deterministic transmission guarantee algorithm based on joint optimization of routing and scheduling.Based on graph theory,this thesis constructs the MEC network model with TSN technology and the data model of TT flows in the MEC network.Taking the characteristics of TSN switches and TT flows into consideration,the constraint conditions and formulas when the TT flows are schedulable and the optimization object that minimizes the unschedulability rate of TT flows are obtained.The complex optimization problem is decomposes into two parts: routing and scheduling.This thesis proposes to choose the most appropriate routing path for TT flows by considering the length of routing and the residual bandwidth of the links in the this routing path comprehensively in the routing part.And in the scheduling part,this thesis optimizes the unschedulability of TT flows based on the PSO(Particle Swarm Optimization)and GA(Genetic Algorithm)respectively.Finally this thesis shows the optimization effect of PSO and GA from two aspects: optimization results and the time used to calculate.And the relationship between the schedulability of TT flows and the network size are also explored from the two aspects.