NOMA versus OMA in Finite Blocklength Regime: Link-Layer Rate Performance

TL;DR

This paper compares the latency and link-layer rate performance of NOMA and OMA in finite blocklength regimes over Rayleigh fading channels, highlighting conditions where each scheme excels.

Contribution

It provides a novel analysis of effective capacity for NOMA and OMA under delay constraints in finite blocklength, including closed-form expressions and impact of power allocation.

Findings

At low SNR, OMA with better channels outperforms NOMA.

NOMA surpasses OMA in total rate at high SNR with loose delay constraints.

Fixed power allocation influences NOMA's effective capacity.

Abstract

In this paper, we investigate the latency performance of non-orthogonal multiple access (NOMA) and orthogonal multiple access (OMA) technologies in finite blocklength regime. In the comparative study, we derive the achievable effective capacity of two-user NOMA and its OMA counterpart under delay quality-of-service constraints. We then obtain closed-form expressions for the achievable effective capacity of the weak and strong users in both scenarios considering transmissions over Rayleigh fading channels. Numerical results are provided. In particular, it is shown that at low signal-to-noise ratios (SNRs), the OMA user with better channel condition outperforms both NOMA users. \blue{We also evaluate the impact of fixed power allocation scheme on the achievable effective capacity of two-user NOMA.} The comparative analysis of the total link-layer rate shows that at high SNRs, the total link-layer rate of NOMA with finite blocklength outperforms the one of OMA when the delay exponent is loose.