Maximizing Energy and Spectral Efficiency Tradeoff in MISO-RSMA Systems Under Coarse Quantization

Using low-resolution quantizers reduces power consumption increasing energy efficiency (EE) but lowers spectral efficiency (SE) in wireless communication systems. To compensate the SE degradation, we consider rate-splitting multiple access (RSMA) which reduces inter-user interference by introducing a common stream in the downlink multi-user multi-antenna system with low-resolution digital-to-analog converters. We propose a precoding method that optimizes a weighted SE and EE. Key challenges that come with optimizing the trade-off between the SE and EE include transmit power optimization, a minimum rate constraint for a RSMA common stream, and quantization error mitigation. To tackle these challenges, we employ the Dinkelbach method and separate the optimization problem into two parts: optimizing the precoding direction and the transmit power. In terms of the precoding direction, we approximate the SE as a smooth function and derive the first-order optimality condition. Viewing this as an eigenvalue problem, we efficiently identify the principal eigenvector which corresponds to the best stationary point. For transmit power, we utilize a gradient descent algorithm. Both the precoding direction and the transmit power are updated in an alternating manner. Simulations validate the advantages of the proposed algorithm over other state-of-the-art algorithms.