A Geometric Moment Bounding Algorithm

Abstract

There are many important problems in the field of communications theory whose solution is the expectation of a function of a random variable. Examples include linear interference problems such as intersymbol interference and co-channel interference. In these cases, it is often not computationally feasible to evaluate the expectation exactly. This paper presents an algorithm that will compute tight upper and lower bounds to generalized moments of a broad class of random variables. The procedure is based on an isomorphism theorem from Game Theory. The technique is easily understood while yielding excellent results for this class of communication problems.