Accurate identification of low-level radiation sources with crowd-sensing networks

Abstract

The use of crowd-sensing networks is a promising and lowcost way for identifying low-level radiation sources, which is greatly important for the security protection of modern cities. However, it is challenging to identify radiation sources based on the inaccurate crowdsensing measurements with unknown sensor efficiency, due to uncontrollable nature of users. Existing methods mainly concentrate on wireless sensor network, where the sensor efficiency is available. To address this problem, inspired by EM (Expectation Maximization) method, we propose an iterative truthful-source identification algorithm. It alternately iterates between sensor efficiency estimation and truthful-source identification, gradually improving the identification accuracy. The extensive simulations and theoretical analysis show that, our method can converge into the maximum likelihood of crowd-sensing measurements, and achieve much higher identification accuracy than the existing methods.