Sum Spectral Efficiency Maximization for Uplink Multi-Layer RSMA With Finite SIC

In this letter, we investigate the sum spectral efficiency (SE) maximization for multi-layer rate-splitting multiple access (ML-RSMA) systems. In ML-RSMA, the data stream of each user is split into sub-streams, which are individually encoded into sub-signals and transmitted with different power levels. In the decoding stage, to reduce the SIC complexity, we propose a finite successive interference cancellation (SIC) scheme, viz., only performs SIC at the receiver after successfully decoding all sub-signals pertaining to a particular layer. Based on the finite-SIC scheme, we formulate a sum SE maximization problem with the power control coefficients of all sub-signals being the variables and propose to use an iterative quadratic transform method to tackle it with low complexity. The simulation results verify the performance of the proposed ML-RSMA with finite-SIC architecture compared to traditional ML-RSMA.