Service flow modelling in the telecom cloud
Document typeMaster thesis
Rights accessOpen Access
In telecom cloud infrastructures, a wide variety of network elements can be monitored to retrieve for many purposes, such as improving network performance and end user experience. Such wide and intense monitoring entails collecting huge volumes of data that needs to be transferred and stored, as well as being analyzed and fast processed to achieve near real-time performance. Therefore, Big Data techniques for data collection, pre-processing, and analysis and visualization have been recently proposed to provide a fully Big Data-backed ecosystem for telecom operators. This project tackles the problem of service traffic flow modelling in the telecom cloud. A simulation and modelling procedure targeting at obtaining predictive models for realistic service traffic flows is developed. Distinct data analytics approaches can be emulated with the objective of evaluating the performance of distributed and centralized monitoring and modelling deployments.