Cross PHY/APP Layer User Grouping and Power Allocation for Uplink Multiantenna NOMA Video Communication Systems

Abstract

The previous suboptimal user pairing and optimal power control for uplink multiantenna NOMA systems considers the physical-layer-only to maximize the sum data rates. For video communications, the objective is changed to minimize the sum video distortion and then maximize the peak signal-to-noise ratio (PSNR). For uplink multiantenna NOMA systems, we propose 1) a novel weak set user selection scheme to maximize the ratio of signal power and the interference to the strong set users, 2) a novel iterative user substitution scheme based on video distortion in the application layer that increases the PSNR, and the video quality, 3) a novel optimal power allocation derived to minimize the video distortion in the application layer. 1) considers the physical-layer only and 2) and 3) consider the application layer too. The numerical results indicate that the proposed cross-layer scheme outperforms the previous physical-layer-only by 1.1 dB, and is only 0.8 dB away from the exhaustive upper bound in PSNR with 20 users and SNR = 15 dB. The complexity of the proposed scheme is slightly greater than that of the previous physical-layer-only scheme, but still much lower than that of the exhaustive upper bound.