Chemical gas leakage source determination using distributed EM algorithm with Gaussian mixture model

Chemical gas leakage source determination with sensor networks has become a research significance in the pollution environmental monitoring and security protection fields, which also known as gas leakage source parameters estimation. In this paper, we proposed a distributed EM algorithm for the chemical gas leakage source determination, and which was based on Gaussian Mixture model. Simulation results show that the proposed EM algorithm could determinate the gas leakage source localization and emission rate, Compare to the central EM algorithm, the distributed EM method was suggested because it can balance the accuracy performance and energy consumption in the sensor network, and it will get a significant reduction in the required numbers of sensor nodes and less energy to achieve the desired performance with less time, all of that was based on the dynamical adjusting scheme for computing sensor nodes selection.