Implementing intelligent automation and decision-making within telecom networks through advanced machine learning models

This chapter presents decision-making through intelligent automation methods using machine learning models for telecommunications network performance management, allowing for the assurance of quality of service. The proposed models are aimed at anomaly and change point detection to reduce the operational complexity for operators, using an efficient tool of machine learning and network management. The performance evaluation of these techniques is important for demonstrating and comparing the practical advantages in real-time networks. Telecom network operation processes benefit from automated and semi-automated techniques using intelligent algorithms and reliable software. Their capabilities can assist the ever-increasing number of Managed Service Providers that usually cope with the complexity of very large networks with fewer available labor resources. The time and effort to identify poorly performing network segments must be minimized because they can impact revenue and customer satisfaction. Detecting network issues earlier is a clear advantage for the proactive treatment of a network.

Network management in its different layers also supports applications that are running over the data network platforms. Network performance evaluation and management have become more complex and challenging with the use of several quality of service levels and diverse network resources. The idea of intelligent network management is based on the association of machine learning models to optimize the automation of network maintenance processes. These models can efficiently detect anomalies and change points with a lower rate of false alarms. This efficient automation lightweight cycle for the management process can enhance network privacy and quality of telecommunications networks. The simplified strategic argument is that operators can save money by supporting network operations with a minimum number of skilled personnel. Considering the great improvement of machine learning techniques over the last few years, many important problems can be addressed efficiently and reliably.