Performance Analysis of NOMA Uplink Networks under Statistical QoS Delay Constraints

Abstract: In the fifth generation and beyond (B5G), delayconstraints emerge as a topic of particular interest, e.g. forultra-reliable low latency communications (URLLC) such asautonomous vehicles and enhanced reality. In this paper, westudythe performance of a two-user uplink NOMA network understatistical quality of service (QoS) delay constraints, capturedthrough each user s effective capacity (EC). We propose novelclosed-form expressions for the EC of the NOMA users andshow that in the high signal to noise ratio (SNR) region, the 'strong' NOMA user has a limited EC, assuming the same delayconstraint as the 'weak' user. We demonstrate that for the weakuser, OMA achieves higher EC than NOMA at small values ofthe transmit SNR, while NOMA outperforms OMA in terms ofEC at high SNRs. On the other hand, for the strong user theopposite is true, i.e., NOMA achieves higher EC than OMA atsmall SNRs, while OMA becomes more beneficial at high SNRs.This result raises the question of introducing 'adaptive' OMA /NOMA policies, based jointly on the users delay constraints aswell as on the available transmit power.