View article

Smartphones and other devices connected to a mobile network typically create billions of measurement samples every day. Those measurements are currently used for instantaneous resource allocation and link adaptation. There is much room for using the measurements also for network optimization with big data analytics. The measurements give valuable insight into the service quality experienced by devices in their locations. This paper illustrates how machine learning algorithms can be applied to identify main interference issues in mobile networks. Subsequent optimization can then lead to lower interference levels, enhanced user throughputs and improved success rates.