Traffic Engineering Based On Wardrop Equilibrium In SON

With the development of the Internet, the increasingly high requirements on the Quality ofService, and the unsatisfactory result of the efforts in the QoS made by the IETF, the trafficengineering came into being with important significance. Traffic Engineering is a kind ofnetwork engineering technology which can map the data flow to the actual physical link,optimize cyber source to achieve service performance requirements of a specific application andachieve macro-regulation and micro control. Traffic engineering techniques which main functionis load balancing and network recovery can control the optimal balance point between thenetwork-oriented and application-oriented aspect, namely it can improve the utilization ofnetwork resources and the QoS requirements of application-specific services or specific user atthe same time.Service overlay network (SON) is built on the basis of the existing infrastructure networkand operated by a third party. In the past few years, the service overlay network has received agreat deal of attention, SON research hasalso been a rapid development. QoS routing problemhas been the focus of the study, and the introduction of game theory provides a new direction foroverlay network. The Wardrop theory was originally used in the study of the trafficnetwork, later has important applications in the field of road planning and so on. In recent years,the study of communication networks has started to pay attention on the Wardrop theory, hopingto carve out a new direction. Facts have proved that Wardrop model has a good adaptability tocommunication network, and has an important innovative research in the basic network.This article first summarizes the background and current situation of the service overlaynetwork, and analyzes the problems and difficulties encountered by the overlay network, thenintroduces the QoS routing mechanism of service overlay network and Wardrop equilibriumtheory and makes focus on the application of the Wardrop equilibrium theory in communicationnetworks, next gives the UE (User Equilibrium) and SO (System Optimization) model. Thewardrop model in the multi-user and multi-path conditions is put forward combined with theconcept of the routing game (RG) in the traffic engineering. Study on the routing game model, arouting algorithm is given based on Wardrop equilibrium of SON. The algorithm can ensureusers’ quality of service needs, meet the user quality of service and balance the system loadbalancing as well. It improves the utilization rate of cyber source, has a high theoretical and practical value.In order to prove that the impedance function we are given can meet the performancerequirements of the system, this paper makes a comparative analysis of Wardrop equilibriumconditions and optimal social cost function SC (OPT) of the congestion game model andbottleneck game model, at the same time gives an analysis of the routing game model under thecondition of multi user multi path and the Wardrop equilibrium harmonic ratio. A kind ofadaptive weight conversion of routing algorithms, namely MMARA routing algorithm ispresented under the analysis on the IP network routing strategy combined with CNLS withoutreference impedance function mode. This algorithm can guarantee the user Quality of Servicedemand and satisfy the flow balancing requirements. Through this, it realizes to maximize thebenefits of SON service providers and improves the quality of network service on the whole. Anexperiment is performed to demonstrate that MMARA routing algorithm performance hasgreatly improved in the network load capacity and meets the requirement of balancing system.