ABSTRACT
In this chapter, the authors briefly discuss the general methods for generating random variates from continuous distributions, the criteria for random variate generation algorithm evaluation and comparisons. Then, they present the state of the art algorithms for exponential variate generation. Most of the random variate generation algorithms fall into one of the following techniques: inverse transformation method, special transformation methods, acceptance/rejection method and the composition or the mixture method. Balakrishnan and Sandhu have recently proposed an alternative method of simulation of progressive Type-II censored samples based on uniform distribution. The important criteria used in judging random variate generation algorithms are speed, simplicity, accuracy and generality of the technique.
