Approximate methods for analyzing slotted random-access spread-spectrum networks

Abstract

The authors develop approximate methods for performance evaluation of a general class of slotted Aloha-type random-access spread-spectrum networks. Approximation is used in three levels: the diffusion approximation of the process Q(t) (a variant of the number of backlogged users), the approximation of the random-access channel and the existing receivers by a single server with equivalent queue-size-dependent service rate, and the approximation of the state-dependent composite traffic distribution by a Gaussian random variable with equivalent mean and variance. Applications and comparisons with the precise solutions are given.<>