Improving Routing Scalability through Mobile Geographic Hashing in MANETs

Mobile wireless ad hoc networks (MANETs) have important applications including military command and control, collaborative and distributed computing, emergency services and sensing networks. Scalable routing in large networks in the presence of mobility is a challenging problem due to the frequent topology changes. A class of routing protocols that utilize location information are particularly scalable as they route using local state. However, the scalability of such a geographic routing protocol critically depends on the scalability of an associated location service which performs global tracking of destination node locations. This talk shows how the technique of mobile geographic hashing can be leveraged to provide a lightweight, efficient and robust means of building scalable location services.

I will first discuss how mobile geographic hashing can be used to build a location service (GHLS) that enables scalable unicast routing in large mobile wireless ad hoc networks. We compare GHLS to the state-of-the-art location services. Surprisingly, our results show that for practical sized networks, complex, asymptotically scalable location services are not necessarily the best choice. In contrast, GHLS provides a more robust and scalable location service. The second part of the talk discusses how mobile geographic hashing helps to build a location service for tracking group membership to enable scalable multicast routing in large mobile wireless ad hoc networks. Our multicast protocol HRPM constructs a lightweight virtual hierarchy using mobile geographic hashing to manage group membership information and uses the stateless properties of geographic routing to deliver data down the multicast trees. Both our protocols demonstrate improved scalability compared to current state-of-the-art unicast and multicast routing protocols and together show the effectiveness of mobile geographic hashing in constructing scalable routing protocols in MANETs.