A cache predictive scheme for on-demand routing in Ad-hoc networks

Ad hoe networks play important roles in next generation mobile systems. This kind of wireless networks do not have any wired infrastructure support. In ad hoc networks, nodes change their position frequently and each node can act as a router. Due to high mobility and frequently link breakage, selecting the reliable path is a critical challenge. Some well-known protocols such as DSR take advantages of caching and on-demanding mechanisms to increase performance and reduce overheads. In this paper, we propose a new factor added to routing table to address such problems. By exploiting routing table stability parameter (RTSP) in each node, we can predict fragile routes and thus provide a mechanism to select robust path to increase efficiency. Each node computes RTSP in a distribution mechanism and adds this factor to route request and reply packets. We use Global Mobile Simulation (GloMoSim) environment to simulate our method applied to Dynamic Source Routing (DSR) protocol. Simulation results demonstrate that considering RTSP factor improves throughput and reduce routing protocol overhead with increased mobility.