NOMA Power Minimization of Downlink Spectrum Slicing for eMBB and URLLC Users

TL;DR

This paper compares NOMA and OMA for downlink spectrum slicing in 5G, showing NOMA reduces power consumption for serving eMBB and URLLC users with limited CSI.

Contribution

It introduces a power minimization framework for NOMA in 5G downlink, leveraging partial CSI to improve resource sharing and reduce power usage.

Findings

NOMA outperforms OMA in power efficiency across tests

Limited CSI of eMBB can still benefit URLLC users in NOMA

Numerical results confirm NOMA's lower power consumption

Abstract

Spectrum slicing of the shared radio resources is a critical task in 5G networks with heterogeneous services, through which each service gets performance guarantees. In this paper, we consider a setup in which a Base Station (BS) should serve two types of traffic in the downlink, enhanced mobile broadband (eMBB) and ultra-reliable low-latency communication (URLLC), respectively. Two resource allocation strategies are compared: non-orthogonal multiple access (NOMA) and orthogonal multiple access (OMA). A framework for power minimization is presented, in which the BS knows the channel state information (CSI) of the eMBB users only. Nevertheless, due to the resource sharing, it is shown that this knowledge can be used also to the benefit of the URLLC users. The numerical results show that NOMA leads to a lower power consumption compared to OMA for every simulation parameter under test.