Studies On Mluticast Switching Simulation Of Self-similar Traffic

With the rapid development of communication technology, multimedia is usedmore and more widely, and the proportion of multicast communication in thecommunication network is gradually increased; the real network traffic has thecharacteristic of the self-similar.Therefore,the research and performance evaluationof the self similar traffic multicast switching structure,contribute to the analysis ofself similar network delay, throughput and performance influence.The first part is about the analysis of the self-similar traffic stream flownetwork self-similarity, including self-similar process and its property self-similartraffic stream heavy-tailed feature and the discussion of the ON/OFF model and keyparameters, through direct superposition of N independent heavy tailed distributioncharacteristics of Pareto ON/OFF source synthesis of self similar traffic flowmethod, in order to establish the self similar traffic model. Secondly, based on theFIFOMS algorithm Input-buffer multicast switching structure, analyze and solvemulticast switching head-of-line blocking, to reduce the use amount of buffer space;Detailed description of the multicast scheduling algorithm request and approval oftwo scheduling process and the cell to send, the iterative algorithm, theestablishment of the input buffer multicast switch fabric model. Again in thesimulation software, we can build a simulation network to achieve the Bernoullisource bursty source and self-similar source input buffer multicast switch fabricperformance simulation. Finally, in different sources, different multicast and Hurstindex of self similarity source conditions, the input buffer switch structure of theaverage throughput, average delay and Captain and other performance indicatorsare analyzed.By the simulation, study the exchange structure performance of the self-similarsource and Bernoulli source, bursty traffic source in the input buffer multicastswitching structure; provide a reference for the performance analysis traffic inswitched networks which the multicast service of is being increased more and more.