Anomaly Detection using Unsupervised Learning in LTE Mobile Network
Mahmoud Nour, Mina Awad, Mina Kamel, Mostafa Essa, Nashwa Abdelbaki · 2021
Combining the multivariate associated with multiple domains gives the computational edge to solve the complex challenges. We can solve the anomaly detection challenges using the power of the unsupervised learning augmented with the statistical modeling. In this paper, we propose our framework to detect the anomaly cells in a 4G network using the raw counters of 4G mobile network. Our approach is based on modeling the healthy network KPIs in terms of revenue, performance and customer satisfaction using unsupervised learning. We also use the statistical modeling to set the severity of the case to direct the mobile network operator to the health network. This provides flexibility in the operation cost versus the customer perception requirement.