Target Location in Self-organized Sensor Networks Based on Perfect Sampling

Abstract

An equivalent center of concentric circle model in target location based on self-organized sensor networks (ad-hoc) is derived. The model is a description of many different sensors in the area where there is one or more monitored target, and the number of sensors is assumed to be the certain stochastic distribution. By defining a proposal distribution about unknown parameters, the maximum posterior likelihood function is taken. Perfect sampling scheme has been applied to induce the runtime of coalesce of chain. We come to the samples from the posterior distribution of targets being monitored. Numerical simulation results indicate that the developed method is reasonable.