A study of LTE network performance based on data analytics and statistical modeling

Various simulation based approaches are developed to study wireless network performance and capacity traditionally. However, these simulation based approaches has its own limitations in supervised and unsupervised learning. Since big data techniques become more available, using big data techniques to understood the huge amount of the LTE network measurements and diagnosis data, evaluate and predict the LTE network capacity and performance becomes a very promising approach. This paper investigates the LTE network performance from the aspect of data analytics and statistical modeling. In this paper, we develop a methodology of data analytics and modeling to evaluate LTE network performance based on traffic measurements and service growth trends. A relational algorithm is developed to obtain the relationship between LTE network resources and LTE KPIs and a forecasting algorithm is developed to trend the network resource consumptions based on traffic and service growth. The numerical results show a high accuracy, robustness, and reliability of the proposed methodology.