Simulation of the self-similar least-action walk model

Abstract

ABSTRACT Simulation of ad hoc networks requires the use of a realistic human mobility model to give credibility to computer-based performance evaluations. To this end, a number of models have appeared in the literature from which the self-similar least-action walk model (SLAW) stands out as an approach mainly intended to incorporate several statistical properties of human motion. Although it is generally accepted that this model is able to produce realistic traces, so far not enough attention has been given towards creating an implementation with an upfront clear design that is also reliable and computationally efficient. This paper addresses these issues by presenting the design, implementation, and validation of a SLAW simulator. This simulator follows the discrete-event simulation paradigm for generation of mobility traces and is flexible enough for incorporating future walker models. This simulator can be used, among other purposes, to perform computer-based studies of networking technologies where human mobility plays a major role.