A Context-Aware Data-Driven Algorithm for Small Cell Site Selection in Cellular Networks

In mobile networks, detecting and eliminating areas with poor performance is key to optimize end-user experience. In spite of the vast set of measurements provided by current mobile networks, cellular operators have problems to pinpoint problematic locations because the origin of such measurements (i.e., user location) is not registered in most cases. At the same time, social networks generate a huge amount of data that can be used to infer population density. In this paper, a data-driven methodology is proposed to detect the best sites for new small cells to improve network performance based on attributes of connections, such as radio link throughput or data volume, in the radio interface. Unlike state-of-the-art approaches, based on data from only one source (e.g., radio signal level measurements or social media), the proposed method combines data from radio connection traces stored in the network management system and geolocated posts from social networks. This information is enriched with user context information inferred from traffic attributes. The method is tested with a large trace dataset from a live Long Term Evolution (LTE) network and a database of geotagged messages from two social networks (Twitter and Flickr).