Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) have emerged as a promising technology for achieving full-space coverage. Prior works on STAR-RISs mostly assumed the full and instantaneous channel state information (CSI) is available, which, however, is practically difficult to obtain due to the large number of elements. To address it, we investigate STAR-RIS aided NOMA systems, where two efficient two-timescale transmission protocols are proposed for different channel setups to maximize the average sum-rate. Specifically, 1) for line-of-sight (LoS) dominant channels, we propose the beamforming-then-estimate (BTE) Protocol, where the long-term STAR-RIS coefficients are optimized based on the statistical CSI, while the short-term power allocation at the base station (BS) is designed based on the effective channels; 2) for the rich scattering environment, we propose an alternative partition-then-estimate (PTE) Protocol, where the BS determines the long-term STAR-RIS surface-partition strategy; then the BS estimates the instantaneous subsurface channels and designs its power allocation and STAR-RIS phase-shifts accordingly. Simulation results validate the superiority of our proposed transmission protocols as compared to various benchmarks. It is shown that the BTE Protocol outperforms the PTE Protocol when the number of STAR-RIS elements is large and/or the LoS channel components are dominant, and vice versa.
翻译:为了解决这个问题,我们调查STAR-RIS(STAR-RIS)协助的NOMA(Star-RIS)系统,其中提出了两个高效的双尺度传输协议,供不同频道设置使用,以最大限度地实现平均和总和。具体地说,1,对于视线(LES)主要管道,我们提议采用BAM-当时估计(BTE)议定书,其中长期的STAR-RIS(CSI)系数是根据统计CIS(CSI)获得的全速和即时信道状态信息(CSI),但实际上由于大量要素,很难获得这些信息。为了解决这个问题,我们调查STAR-RIS(NOM)协助的短期电力分配系统,因为根据有效渠道设计了两个高效的双尺度传输协议,为尽量扩大平均比例(PTE)议定书,其中BS确定长期STRAR-RIS(LES)地面分割战略;然后,BS(BS)根据统计,长期的次表估计长期地下估计(BST-RIS)系数系数系数系数系数系数系数,并据此显示STRA-RVRVA(S-RRRRV)系统的大型结构结构结构结构结构结构结构结构结构,即显示SIT-RVLVILVILVILVILA-RVLA-RA-RA-RR)的大型配置,因此显示S-RVIP-RVA-RVA-RVA-RA-RA-RA-RVA-RVA-RVA-RVA-RVDRVT-RVDRVT-RB。