5G-NR Cross Layer Rate Adaptation for VoIP and Foreground/Background Applications in UE

Abstract

The recommended bit rate (RBR) is assigned by gNodeB to the user equipment (UE) using MAC control element (CE) entity to provide bit rate information in 5G New Radio (NR). At the UE, the bit rate information is passed on to the upper layers; i.e., transport or application, for a specific logical channel either in uplink or downlink. However, based on specific application rate of target, UE does not know how to efficiently utilize and distribute RBR/throughput in lower and upper layer, respectively considering the specific logical channel. To address these problems, we propose a cross layer rate adaptation (CLRA) mechanism for UE. CLRA consists of two parts. In the first part, CLRA utilizes RBR received from gNodeB to compute throughput at lower layer. In the second part, CLRA distributes the throughput in upper layer received from lower layer based on specific applications rate of target. CLRA provides an intelligent mechanism to distribute throughput among foreground/ background applications and voice over internet protocol (VoIP) application considering a learning based codec adaptation. We conduct experiments with Samsung Galaxy S8 device and simulations to validate CLRA mechanism for applications in 5G NR.