Model-based Deep Learning for Rate Split Multiple Access in Vehicular Communications

• DOI: 10.48550/arxiv.2405.01515
• Publication Date: 2024-05-02
• Authors: Hanwen Zhang; Mingzhe Chen; Alireza Vahid; Haijian Sun
• Keywords: Signal Processing (eess.SP); FOS: Computer and information sciences; Computer Science - Information Theory; Information Theory (cs.IT); FOS: Electrical engineering, electronic engineering, information engineering; Electrical Engineering and Systems Science - Signal Processing

Rate split multiple access (RSMA) has been proven as an effective communication scheme for 5G and beyond, especially in vehicular scenarios. However, RSMA requires complicated iterative algorithms proper resource allocation, which cannot fulfill the stringent latency requirement constrained vehicles. Although data driven approaches can alleviate this issue, they suffer from poor generalizability scarce training data. In paper, we propose a fractional programming (FP) based deep unfolding (DU) approach to address allocation problem weighted sum rate optimization RSMA. By carefully designing penalty function, couple variable update with projected gradient descent algorithm (PGD). Following structure of PGD, embed few learnable parameters each layer DU network. Through extensive simulation, have shown that proposed model-based neural networks similar performance optimal results given by traditional but much lower computational complexity, less data, higher resilience test set out-of-distribution (OOD)