Adaptive statistical learning of cellular users behavior

The wide-spread usage of cellular communication in recent years provides opportunities for studying social and environmental phenomena on a global level, using measurements collected anyway by the cellular network. In this paper we demonstrate how certain aspects of social behavior can be studied using statistical analysis of cellular transmission path loss data. We suggest several applications and, in particular, we present a method for dynamically estimating the percentage of indoor vs. outdoor usage in cellular sectors, by applying an innovative mixed Gaussian model for the accumulated path loss measurements. The method is tested with real data collected by a commercial cellular network from a large number of sectors. In addition to the indoor vs. outdoor usage, we demonstrate how path loss data can be used for real-time estimation of the coverage area of cellular sectors, which provides valuable information for cellular network planning, optimization and operation.