Atm Network, Discrete Event Simulation Methods And Forecast Congestion Algorithm Design,

The Asynchronous Transfer Mode(ATM) is the key technology to implement B-ISDN. However, the buffer queues extremely increase and the congestion appears when many kinds of traffics or multiple connections arrive at a bottle-necked node at the same time, or high-speed users access to the low-speed links. As a result, the round trip delay rapidly raises, and heavy congestion may cause collapse. So it is a primary task to design the efficient scheme to control congestion in order to increase the utilization of the resource and improve the quality of service.In this paper, the dynamical behavior of the ATM network is analyzed by using discrete event system method and Petri net. Then integrated with EFCI algorithm and ERICA algorithm, the event-drived mechanism and programme flow chart within the ATM network are exposed. In order to further understand the discpline, the implementation of the source, the switch and the destination algorithms are provided by using OPNET.Because the existing algorithms didn' t consider network transmission delay, the predictive-based EFCI algorithm(P-EFCI) is presented by using basic control theories. The OPNET simulation results demonstrate that the performance of P-EFCI algorithm is superior to the original one.An improved algorithm is proposed based on generalized predictive control, which enhances the robustness of closed-loop systems, and guaranteed the systems stability, and realizes the fairness of bandwidth allocation in steady state. The effectiveness of the proposed methods is demonstrated by the simulation results.In a word, combining with EFCI and ERICA algorithm, the algorithm execution mechanism is exposed by analyzing the discrete event behavior of end system. On the basis of the discussion, the extended EFCI algorithm and predictive congestion control algorithm are presented, and the stability and fairness of the proposed method is proved. Simulation results indicate that the proposed scheme in this paper can obtain better network performance.