Research Of QoS Multicast Routing Based On Improved Ant Colony Algorithm

In recent years,with the rapid development of information technology,many new applications have appeared on internet,particularly multimedia application with high-bandwidth demand.It requires a mass of transmission of audio and video which take high bandwidth,such as video on demand,network meetings.Traditional data communication adopts technology of unicast or broadcasting.These methods cause the excessive resources waste of host computer and network,and also cause problems of sharp bandwidth consumption and network congestion.In order to alleviating these problems,computer network is required to has QoS(Quality of Service) multicast communication capabilities.Multicast is a new network transmission technology of point-to-multipoint or multipoint-to-multipoint.It can control network traffic and reduce load of host computer.In this paper,after analyzing the bandwidth,delay,delay jitter and packet loss rate and cost constraints of QoS multicast routing problem,we put forward the QoS multicast routing optimization method based on improved ant colony algorithm on the basis of abstracting QoS multicast routing model.Aimming at the flaws of ant colony algorithm,such as the long time calculation and easily falling into the local optimal solution,we improve the ants pheromone update strategy and the next jump node selection strategy.Then it is simulated in matlab.The simulation results show the improved ant colony algorithm can effectively find a good performance multicast trees, and meet the business needs of the quality of service.After software design and simulation,FPGA is adopted to implement QoS multicast routing based on improved ant colony algorithm.In the process of implementation,the system is divided into pheromone update module of ants,next step node selection module of ants,random number generator modules and so on.VHDL is used to describe the each basic functional module.In order to improve design efficiency,we use IP for memory design,and use DSP builder for mathematical computing.Timing control is the core of the whole system.Here the timing control of each module is realized by the single-process moore state machine.After completing design of each function module,we carry out function simulation in quartusⅡ.The top module is designed by components case statement after accurate simulation.Finally, we carry out function simulation and timing simulation for the system respectively.In the whole design process,the speed of ant colony algorithm is greatly enhanced because of gaving full play to FPGA parallel computing capability and the application of pipeline technology.The simulation results show that QoS multicast routers based on FPGA run faster about an order of magnitude than the software.