Abstract
We propose in this paper a novel adaptive approach inspired by fluid dynamics as a distributed, scalable, and robust solution to the deployment problem of mobile sensor networks in unknown environments. Our approach is based on the physical principles of fluids through which we model a mobile sensor network as a fluid body and individual nodes as fluid elements. We achieve desirable properties of effective coverage, scalability, and robustness by virtue of the diffusive and self-spreading behavior of compressible fluids as modeled in our deployment approach. Simulation of our deployment strategy shows that the approach is scalable in terms of environment and network size. It is also robust against localization uncertainty, partial operational failure, and dynamic changes in the landscape. In truly unknown environments where the number of nodes to be deployed cannot be determined a priori, our adaptive deployment scheme guarantees thorough coverage of the environment.