Mobile network data, aka mobile phone data, stand as a promising data source for the production of official statistics. Several results already show their potential. However, the configuration of an end-to-end industrialised statistical production process using this source in combination with other data still needs further work to achieve usual quality standards in Official Statistics. The statistical production process is a complex process, which entails the need to deal with many highly interrelated different aspects. Some of these have been recently approached in the first ESSnet on Big Data (2016-2018) and are currently under further research in the second ESSnet on Big Data (2018-2020) like the geolocation of network events, the creation of a data model for statistical exploitation, the inference to the target population and the proposal of quality indicators. Here we present ongoing work in the use of hierarchical models in the inference from mobile phone data sets to the target population under analysis with the combination of auxiliary information. Our work adapts already existing proposals with administrative data to estimate the population size. We illustrate the use of these models with synthetic aggregate mobile network data about a closed population providing a proof of concept of this framework to approach the representativity issue with this new data source.