Iterative Algorithms for Performance Evaluation of Wireless Networks with Guard Channels

Abstract

Analytical and modeling work performed since the late eighties have led to nonlinear equations that relate important parameters used for performance evaluation of wireless networks. These nonlinear equations have no closed-form solution and iterative algorithms are used to find numerical solutions. However, iterative algorithms that are not designed to detect and overcome nonconvergence situations may fail to converge. We present an iterative algorithm that numerically solves six dependent nonlinear equations. The algorithm always converges and obtains values of blocking probability, pb, and forced termination probability, pft, at any desired level of accuracy. We then used this algorithm to numerically show that for a given pair of values of pb and pft, there is an optimal number of guard channels that supports a maximal new-call arrival rate.