Active RIS Aided NOMA for HAP-MISO Systems

Abstract

This letter investigates an energy-efficient scheme for the downlink active reconfigurable intelligent surface (ARIS) aided non-orthogonal multiple access (NOMA) on high altitude platform multiple-input single-output (HAP-MISO) with lens antenna array, i.e. ARIS aided beamspace HAP-NOMA. Specifically, by jointly considering the line-of-sight and non-line-of-sight components on HAP, we deduce the sparsity of the sparse matrix for the dimension-reduced beamspace HAP-NOMA channel and the sparse matrix is defined as the matrix with at most four nonzero elements, where the power is concentrated in the corresponding direction. The traditional user grouping strategy is employed by the sparsity. Then, we derive the closed-form equivalent (CE) of the spectral efficiency (SE) based on the sparsity, which is the upper bound on SE in practice due to the existing power leakage for beamspace HAP-NOMA. Moreover, the optimization problem is formulated to maximize the CE and the joint alternating precoding algorithm is proposed to attain the active reflecting at the ARIS and the power allocation at the HAP, which is only dependent on statistical channel state information (SCSI). Notably, users in the same beam can be served simultaneously by NOMA and the severe path loss due to the HAP-RIS-user link can be alleviated by ARIS. The integration of NOMA and ARIS would compensate the performance loss caused by beamspace HAP-MISO. Numerical results show that the proposed algorithm can converge faster with a lower complexity and gain better performance.