Wireless sensor network based localization and threat estimation of hazardous landfill gas source

Hazardous gases released from subsurface landfill source causes explosion and pulmonary ailments in human beings. This work focuses on wireless ad-hoc multisensor network based localization and threat estimation of such a buried gas source. It is reasonably considered that the gas source effuses methane (CH4) and volatile organic compounds (VOC) over a large indoor environment. Experiments are performed to emulate the subsurface gas diffusion process. The propagation constant of CH4 is also determined experimentally. First, the maximum likelihood (ML) method is used to estimate the most probable gas source location within a target area. It uses gas concentration readings sensed by the multisensor nodes at different predetermined points over the large area. The initial localization belief is next refined using a judiciously chosen subset of sensor nodes with limited mobility. Simulation results indicate that the proposed source localization algorithm offers higher estimation accuracy with reduced hardware using lesser energy resources. Finally, we propose an analysis that helps to estimate the explosion-threat possessed by the landfill source. Since VOC exposure has severe ill effects on human health, a VOC concentration measurement scheme is also proposed here.