Predictive Quality of Service in Cellular Networks: Challenges, Framework, and Application in Vehicular Communications

Abstract

Predictive quality of service will allow next-generation cellular networks to improve the management of the existing services and support more demanding use cases. In this article, we discuss the main challenges on the way to practical realizations of predictive quality of service, from fundamental considerations related to prediction horizons and handling uncertainty to practical aspects on data collection and implications of the architectural choice. We also present a general methodology for predictive quality of service that addresses some of the main challenges and may be used to realize predictive quality of service at scale. As a case study, we apply the framework to a vehicular communications setting using a data-set acquired from a deployed private LTE network.