Discrete event simulations, mobile ad hoc networking and scalability

Mobile Ad hoc NETworks (MANETs) are becoming increasingly popular. Their deployments not only span a wide range of networking and communications environment, but also range from small sized networks (e.g., special operations networks) to large scale networks (e.g., sensor networks, military networks). While the design and analysis for small sized networks is well underway, the counterpart for large sized networks (i.e., networks with 1000's of nodes, which we also refer to as 0(1000) node networks) is far from having been understood. It is not only important that this gap in the understanding of scalability performance is filled, but it is also extremely critical given the fact that the large scale networks that are being deployed (e.g., sensory networks, military networks) need to carry mission critical data. However the challenges in analyzing large scale MANETs (i.e., 0(1000) node networks) are daunting. In this work, we focus on this scalability analysis problem and propose techniques to render such an analysis feasible. More specifically, we combine the power of discrete event simulations (DES), novel memory management (state-of-the-art computing hardware) and analytic modeling in a judicious manner in order to provide a methodology for scalability analysis. We also propose an adaptive algorithm (via analytic modeling), which when combined with DES techniques enhances scalability performance and analysis of MANET routing protocols, significantly.