Parallel simulation of statistical multiplexers (Q1901404)

The simulation of high-speed telecommunication systems such as ATM (Asynchronous Transfer Mode) networks has generally required excessivley long run times. This paper reviews alternative approaches using parallelism to speed up simulations of discrete event systems, and telecommunication networks in particular. Subsequently, a new simulation method is introduced for the fast parallel simulation of a common network element, namely, a work-conserving finite capacity statistical multiplexer of bursty ON/OFF sources arriving on input links of equal peak rate. The primary performance measure of interest is the cell loss ratio, due to buffer overflows. The proposed method is based on two principal techniques: (i) the derivation of low-level (cell level) statistics from a higher level (burst level) simulation and (ii) parallel execution of the burst level simulation program. For the latter, a time-division parallel simulation method is used where simulations operating at different intervals of simulated time are executed concurrently on different processors. Both techniques contribute to the overall speedup. Furthermore, these techniques support simulations that are driven by traces of actual network traffic in to standard models for source traffic. An analysis of this technique is described, indicating that it offers great potential for delivering good performance.