Title :
Pade approximations of probability density functions
Author :
Amindavar, Hamidreza ; Ritcey, James A.
Author_Institution :
Dept. of Electr. Eng., Washington Univ., Seattle, WA, USA
fDate :
4/1/1994 12:00:00 AM
Abstract :
The analysis of radar detection systems often requires extensive knowledge of the special functions of applied mathematics, and their computation. Yet, the moments of the detection random variable are often easily obtained. We demonstrate here how to employ a limited number of exactly specified moments to approximate the probability density and distribution functions of various random variables. The approach is to use the technique of Pade approximations (PA) which creates a pole-zero model of the moment generating function (mgf). This mgf is inverted using residues to obtain the densities
Keywords :
approximation theory; function approximation; poles and zeros; probability; radar theory; random functions; random processes; signal detection; Pade approximations; detection random variable; distribution functions; moment generating function; moments; pole-zero model; probability density functions; radar detection; Convergence; Distribution functions; Erbium; Function approximation; Laplace equations; Mathematics; Probability density function; Radar detection; Random variables; Taylor series;
Journal_Title :
Aerospace and Electronic Systems, IEEE Transactions on
