Estimation of the number of operating sensors in sensor network

This paper investigates the estimation of the number of operating sensors in a wireless sensor network. The basic model considered is equivalent to an urn model with replacement. For this model, we propose an estimator based on the good-turing nonparametric estimator of the missing mass. We show how this estimator can be applied to other related problems like estimation of class histograms. It is also shown that the estimator proposed is robust to model changes; more exactly its performance degradation is small when it is applied to a batch sampling model that models a receiver with multiple packet reception (MPR) capabilities. A modified estimation method that takes into account the batch sampling model is given and its performance is discussed.