Multipath Scheduling for 5G Networks: Evaluation and Outlook
H. Wu, Giuseppe Caso, Simone Ferlin, Özgü Alay, Anna Brunström
Abstract
The fifth generation (5G) of cellular networks aims at providing very high data rates, ultra-reliable low latency, and massive connection density. As one of the fundamental design trends toward these objectives, 5G exploits multi-connectivity (i.e., the concurrent use of multiple access networks), where multipath transport protocols have emerged as key technology enablers. The scheduler of a multipath transport protocol determines how to distribute the data packets onto different paths and has a critical impact on the protocol performance. Within this context, we present in this article the first empirical evaluation of state-of-the-art multipath schedulers based on real 5G data, for both static and mobile scenarios. Furthermore, we introduce M-Peekaboo, which builds on a state-of-the-art learning-based multipath scheduler and extends its usage to 5G networks. Our results illustrate the benefits of employing a learning-based multipath scheduler for 5G networks and motivate further studies of advanced learning schemes that can adapt more quickly to the path conditions, and take into account the emerging features and requirements of 5G and beyond networks.