A Machine Learning Framework for Detection of Sleeping Cells in LTE Network

Abstract

The rapid advancements in telecommunication systems leads to growing data volume and high customer expectations in terms of cost and quality of service. The changing dynamics of radio network usage poses challenges for the operators in terms of optimizing and maximizing network efficiency while reducing maintenance and operational expenditure. Automatic detection of sleeping cell (SC) (i.e. a cell which is not providing normal services to the users) in the network is one way of lowering maintenance cost and improving network performance. This paper presents an intelligent machine learning framework that make use of minimize drive testing (MDT) functionality to gather key performance indicators (KPI's) of the LTE network. These measurements are further projected to a low-dimensional embedding space and are used in conjunction with state of the art learning models to automate the SC detection process.