Analysis of Multiple Trees on Path Discovery for Beacon-Based Routing Protocols

Routing in Mobile Ad-hoc Networks (MANETS) has proved to be an extremely challenging research problem due to the high frequency of link changes in wireless mobile environments. Most known routing protocols struggle to maintain complex routing structures, e.g., N spanning trees, or use expensive (from the perspective of communications cost) methods for path discovery. These methods incur extreme communications costs in most MANET deployments. In this paper, we investigate the performance of a relatively unstudied class of routing protocols which we refer to as Beacon-Based Routing Protocols. This class of protocols maintain minimum routing structures and use these structures to boot strap path discovery. In a well defined sense, this class of routing protocols achieves optimal performance with respect to minimizing the communications costs associated with on-demand routing. We specifically investigate the impact of multiple tree implementations for path discovery and analyze their impact on the effectiveness of optimal path discovery.