Mobility in mobile ad hoc networks: Models, analyses, and applications

An important feature of mobile ad hoc networks is that nodes in the networks are mobile. This mobility changes the links between nodes and the topologies of the networks. Communications are seriously influenced. It is thus important to study the problem of mobility.;This dissertation studies several mobility issues. These issues are integral parts of mobility research. The first issue is to model nodes' movements in real world. Most researches in mobile ad hoc networks rely on simulation and one foundation of simulators is mobility model. Further, given mobility models, people want to know properties of these models. The knowledge of these properties help people better understand the macro behaviors of networks which in turn can be used to improve networks' performances. Besides mobility itself, people also want to know its applications. The next issues of mobility thus deals with how mobility can be used to help meet people's various requirements of the networks, e.g., better communications, higher performances.;Following these footsteps, we do the following work. We first propose an agenda based mobility model in the intention of providing a more realistic model for network simulation. Each node carries a unique agenda which specifies her when-where-what information and drives her movement. This is the first model that combines human social activities and geographical movements. We then study the node encounter problem in a mobile ad hoc network where nodes only visit certain locations and meet each other only at those locations. We derive results for nodes' encounter probability and encounter delay in various scenarios. These results help us better understand node behaviors in networks. Finally we exploit mobility to help secure data transmission in mobile ad hoc networks. In this scheme, source divides a message into multiple shares and sends them to destination one after another with a certain interval. The destination combines the shares and gets the original message. Due to mobility, shares are routed through different intermediate nodes and those nodes are unlikely to get enough shares to combine the original message. This scheme cleverly utilizes the mobility feature and can be used in delay tolerant scenarios.