0514975 Dongsoo S. Kim Indiana University
Mobile ad-hoc network (MANET) will be one of the leading technologies in future warfare and emergency rescue to reduce the threats to personnel and to enhance the efficiency of their missions. The vehicular industry is exploiting the MANET technology for enabling road safety and commercial applications. A communication device in the MANET is capable of being a packet forwarder to provide rapid inter-connection without any stationary infrastructure. Mobile nodes are supposed to cooperate in order to achieve overall objective in a reconnaissance mission in the nature of their applications. There are, however, very few research results on cooperative mobility models. Even if there are, their theoretical foundations are not known and the similarity to real movement are neither.
The PI, Prof. Kim, proposes to analyze the characteristics of node movement in ad hoc networks to investigate the theoretical aspects of cooperative node movemett. Among those characteristics are the distance and signal power between mobile nodes, the number of neighbor nodes within a transmission range, the logical link duration, and their cooperating strategies. Utilizing the characteristics, he will develop a new cooperative movement model that closely describe a realistic node movement pattern. In addition, the project will use stochastic process models to investigate an analytic framework of movement models and to describe different mobility models including entity mobility models and group mobility models. The framework will provide a tool to compare two or more dissimilar mobility models in fair ways regardless of detailed deployment environments.
Intellectual Merits Existing mathematical models can describe only a specific mobility model, mostly a random walk model. The PI intends to build a uniform mathematical framework that can characterize different mobility models including random walk, random waypoint, and some group mobility models. Due to the lack of the mathematical foundation, MANET studies focus on exercising extensive simulation using a limited number of mobility models. It is possible to generate biased results validated only for the given mobility models, which leads to an invalid conclusion. The analytic model proposed in the project will give a general insight to the nodal movement and a capability to predict the resulting topological change.
Broad Impact To the large scale of mobile ad hoc networks, the ability to analyze node movement and to predict topological changes due to the movement will give thorough understanding on the availability of logical link and hence the accessibility of multi-hop path in MANET. The project abstracts geometrical properties of node movements so that it gives a generality for comparing different mobility models in a uniform constitution that makes it possible to discover their resemblances and differences.
