On importance sampling in digital communications. I. Fundamentals

A simulation algorithm design strategy based on the combination of event simulation, conditional importance sampling, and asymptotically optimal biasing of Gaussian noise inputs is discussed. The utility of this approach is illustrated by presenting numerical results for a satellite channel model that includes uplink and downlink noise sources, a travelling-wave tube amplifier (TWTA) nonlinearity, and intersymbol interference (ISI) from both uplink and downlink filtering. An overview and comparison of various simulation design strategies and some results on the optimization of general mean translation and variance scaling biasing schemes for nonlinear systems are presented.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>