Diffusive Source Localization Algorithm Based on Decentralized Particle Filter in Cluster Sensor Networks

zhang yong,zhang liyi,meng guangchao,li jigong

DOI: https://doi.org/10.3969/j.issn.1004-1699.2016.08.020

2016-01-01

Abstract:To resolve the diffusive source localization problem in environment monitoring and pollution control,a de-centralized particle filtering method was proposed for the diffusive source localization in dynamic clustering sensornetworks. Based on diffusion model,the decentralized particle filtering algorithm was designed for diffusive sourceparameters estimation. The sensor nodes scheduling and clustering were completed with the optimization of an ener-gy consumption balance model,which include the contents:track variance and energy cost. At last,through iterativeoperation of the above steps,the diffusive source localization could be achieved. The analysis and simulation resultsshow that the proposed method can effectively reduce the energy consumption and balance network lifetime with arelatively high parameters estimation and source localization accuracy.