Модель обслуживания трафика одноадресных и многоадресных соединений высокочастотной сети 5G

Abstract

Multicasting is widely used in conventional wired and wireless networks as it allows significantly improving resource utilization in presence of users interested in the same content. However, the support of this type of service in prospective 5G New Radio (NR) systems has received only little attention so far. NR systems operating in millimeter wave (mmWave) frequency, a feature of which is the need for direct visibility between devices involved in the connection. One of the tasks in NR systems is the problem of finding the optimal location of access points to provide a network with a stable connection. In this paper, merging the tools of queuing theory and stochastic geometry we develop a model of 5G NR base station (BS) serving a mixture of unicast and multicast traffic. We validate our model against computer simulations using multicast/unicast session drop probabilities and system resource utilization as metrics of interest. Our numerical results illustrate that the presence of multicast type of traffic severely compromises performance of unicast sessions. Furthermore, this effect is amplified when the inter-site distance (ISD) between BSs increases. Thus, to satisfy prescribed performance guarantees in terms of unicast and multicast session drop probabilities, explicit resource reservation mechanism at NR BS might be required.