Exploiting mobility in large scale ad hoc wireless networks

Mobility is generally viewed as a major impediment in the control and management of large scale wireless networks. However, researchers have recently looked at mobility in a different way, trying to take advantage of it instead of protecting from it. In the literature, mobility is generally modeled as independent random moves by each node. However, in reality, mobile nodes tend to show some degree of correlated (group oriented) motion behavior. This feature, once detected and understood, can be exploited to help improve network performance, in particular, scalability. We present several schemes for exploring mobility patterns, in particular group mobility. We show, via simulations, that the physical group mobility behavior of mobile nodes can be autonomously recognized and discovered by network layer routing schemes. It can then be utilized to reduce routing overhead and achieve good scalability in large scale ad hoc networks.