Parallel simulation of high-speed network multiplexers

The simulation of statistical multiplexers in high-speed networks requires excessively long simulation runs when extremely small cell loss ratios are considered. We present an approach that allows the fast parallel simulation of finite capacity statistical multiplexers driven by bursty ON/OFF sources. Our approach is based on two techniques: (1) the derivation of cell level statistics from a burst level simulation; and (2) the use of time-parallel methods without fix-up phases for the burst level simulation. Both techniques contribute to the overall simulation speedup. The latter technique allows speedup factors that are linear with respect to the number of processors. An implementation of the method using 16 processors on a KSR 1 multiprocessor simulates from two million to over one billion cell arrivals per second, or up to five orders of magnitude faster than an efficient conventional sequential simulation.<<ETX>>

[1]  H. Ammar,et al.  Time warp simulation using time scale decomposition , 1992, TOMC.

[2]  John Zahorjan,et al.  Parallel simulation of performance Petri nets , 1991 .

[3]  Philip Heidelberger,et al.  Effective Bandwidth and Fast Simulation of ATM Intree Networks , 1994, Perform. Evaluation.

[4]  David M. Nicol,et al.  Simultaneous parallel simulations of continuous time Markov chains at multiple parameter settings , 1991, 1991 Winter Simulation Conference Proceedings..

[5]  Albert G. Greenberg,et al.  Algorithms for unboundedly parallel simulations , 1991, TOCS.

[6]  Michael R. Frater Application of fast simulation techniques to systems with correlated noise , 1992, WSC '92.

[7]  Albert G. Greenberg,et al.  Efficient Parallel Algorithms for Linear Recurrence Computation , 1982, Inf. Process. Lett..

[8]  Marie Cottrell,et al.  Large deviations and rare events in the study of stochastic algorithms , 1983 .

[9]  Philip Heidelberger,et al.  Parallel trace-driven cache simulation by time partitioning , 1990, 1990 Winter Simulation Conference Proceedings.

[10]  Jean Walrand,et al.  A quick simulation method for excessive backlogs in networks of queues , 1989 .

[11]  David M. Lucantoni,et al.  A Markov Modulated Characterization of Packetized Voice and Data Traffic and Related Statistical Multiplexer Performance , 1986, IEEE J. Sel. Areas Commun..

[12]  Yi-Bing Lin,et al.  Parallel Trace-Driven Simulation for Packet Loss in Finite-Buffered Voice Multiplexers , 1993, Parallel Comput..

[13]  Jain J. Wang,et al.  Determining initial states for time-parallel simulations , 1993, PADS '93.

[14]  R. M. Fujimoto,et al.  Parallel discrete event simulation , 1989, WSC '89.

[15]  Yi-Bing Lin,et al.  A time-division algorithm for parallel simulation , 1991, TOMC.

[16]  D.M. Cohen,et al.  A simulation study of video teleconferencing traffic in ATM networks , 1993, IEEE INFOCOM '93 The Conference on Computer Communications, Proceedings.

[17]  François Baccelli,et al.  Parallel simulation of stochastic Petri nets using recurrence equations , 1993, TOMC.

[18]  Jain J. Wang,et al.  Approximate time-parallel simulation of queueing systems with losses , 1992, WSC '92.