Networks of queues come up often as types for a large choice of congestion phenomena. Discrete occasion simulation is usually the one to be had capacity for learning the habit of advanced networks and lots of such simulations are non­ Markovian within the experience that the underlying stochastic approach can't be repre­ sented as a continual time Markov chain with countable nation house. according to illustration of the underlying stochastic means of the simulation as a gen­ eralized semi-Markov approach, this publication develops probabilistic and statistical tools for discrete occasion simulation of networks of queues. The emphasis is at the use of underlying regenerative stochastic method constitution for the layout of simulation experiments and the research of simulation output. the obvious methodological benefit of simulation is that during precept it's acceptable to stochastic structures of arbitrary complexity. In perform, although, it is usually a decidedly nontrivial subject to acquire from a simulation details that's either helpful and actual, and to acquire it in a good demeanour. those problems come up basically from the inherent variability in a stochastic approach, and it is important to hunt theoretically sound and computationally effective tools for conducting the simulation. except implementation reflect on­ ations, very important matters for simulation relate to effective equipment for producing pattern paths of the underlying stochastic method. the layout of simulation ex­ periments, and the research of simulation output.