Research On Mulitcast Routing Problem In Software Defined Networking

As a new network architecture,Software Defined Networking(SDN)effectively compensates for the shortcomings of the traditional networks and provides more opportunities for network innovation.As the multicast services has been quite mature in traditional networks,SDN also needs to accommodate multicast applications to improve its network services capability.Therefore,how to design and deploy multicast services has become one of the major highlights in the field of SDN.Multicast routing problem is one of the issues for offering multicast services.Quality of Service(QoS)reflects to what extent network services are able to meet the demand of end users.In general,QoS involves in a number of network properties,such as bandwidth and delay,where their threshold values are defined as constraints.How to address QoS multicast routing problem has become one of the frontier research topics in today's academic field.At present,people use Evolutionary Algorithm(EA)to solve QoS multicast routing problem,including Ant Colony Optimization(ACO)algorithm.However,there are too many defects in the existing ACOs for example too time-consuming,weak global search ability etc.So,in this paper,a new ACO is proposed to tackle the bandwidth and delay constrained multicast routing problem.The new algorithm uses a new ant routing rules and a pheromone update strategy,which makes up the defects of the traditional ACO algorithms and significantly enhances the local and global search abilities.In this paper,a number of multicast scenarios are generated to evaluate the performance of the proposed algorithm.The simulation results show that the proposed ACO outperforms the existing ACO algorithms in terms of the running time and the quality of best solutions obtained.However,relatively speaking,the proposed ACO also consumes a considerable amount of computational time on each instance.On the other hand,Fruit Fly Optimization(FFO)as an emerging EA that is fast yet effective.This paper,for the first time,applies FFO to address the bandwidth-and-delay-constrained multicast routing problem.We develop a new type of fruit fly position form so that FFO is well adapted for the multicast routing problem concerned in this paper.In order to enhance FFO's optimization ability,a new smell-based search strategy and a new vision-based search strategy are devised.Compared to a number of widely used EAs,simulation results demonstrate that the superiority of the proposed FFO with respect to the running time,the convergence speed and the quality of the solution.This paper designs and implements a QoS multicast services system for SDN.The system uses Mininet simulation tool,Ryu controller and OpenFlow protocol and technical standard to form a system with five modules,including the multicast group management,QoS multicast routing,topology management,loop control and request scheduling.This system adopts the proposed ACO and FFO in the process of multicast routing.calculation.The simulation results show that the system can provide multicast services in SDN.In addition,it is also a valued reference regarding the deployment of multicast services in the context of SDN.