Forecasting Routes And Self-Adaptation In Multi-Hop Wireless Sensor Networks

Sensor networks find application in many tactical ISR/ISTAR processes and applications. However, these processes and applications depend on reliable collection, distribution and delivery of information that, typically, travels over multiple interconnecting nodes to reach processing centres, and are susceptible to various disruptions such as the ones caused caused by message drops, packet loss and loss of connectivity due to high traffic volumes and noise on the wireless medium. In this paper, we investigate and present approaches to pro-actively adapt routing over such networks by forecasting potential faulty regions of the network based on previous trends and reorganising routing paths. We have prototyped this approach in the ITA Sensor Fabric, an evolving middleware infrastructure for sensor networks. We, further, provide some preliminary results based on simulations.