Modeling of pollutant distribution based on mobile sensor networks.

Pollution monitoring based on wireless sensor networks is becoming highly attractive. This paper presents an effective pollutant distribution modeling approach using a mobile sensor network. As for mobile nodes, energy consumption and link quality between nodes are two key factors. In the proposed approach, we present an autonomous sensing model and an energy-driven motion control scheme, which can make a good trade-off between energy efficiency and modeling accuracy. A comprehensive set of simulations demonstrate that our approach can model the pollutant distribution with less iteration times and higher accuracy. In particular, even for a relatively complex concentration field, the similarity between the reconstructed model and the pollutant distribution can reach 95% through about 20 iterations using 25 mobile sensor nodes. Moreover, we validated the feasibility of the proposed approach through an actual monitoring of water pollutant distribution.