Track genomics, neuroscience and molecular biology.
Search papers, labs, and topics across Lattice.
This paper introduces a physics-driven self-supervised graph learning framework to optimize distortion-aware precoders for satellite systems, directly addressing nonlinear power amplifier (PA) distortion without requiring labeled data. The framework jointly designs the precoder and admission control to minimize PA power consumption while satisfying user quality-of-service (QoS) constraints under PA nonlinearity in a MISO satellite system. The proposed GNN-based pipeline achieves performance comparable to SCA-based methods but with significantly reduced runtime and a broader operational range.
Self-supervised GNNs can directly learn distortion-aware precoders for satellite systems, outperforming traditional optimization methods in speed and operating range while maintaining QoS.
This work develops a physics-driven self-supervised graph learning framework that directly optimizes the distortion-aware precoder under nonlinear correlated power amplifier (PA) distortion, without relying on expensive optimizer-generated labels for a user-centric satellite system. We minimize nonlinear PA consumption subject to user quality-of-service (QoS) constraints under PA nonlinearity for a downlink multi-input-single-output (MISO) satellite system. Our pipeline jointly designs the precoder and admission control that minimize nonlinear PA consumed power. In particular, a graph neural network (GNN) for the design of a distortion-aware precoder is optimized end-to-end in a self-supervised manner using a differentiable Itô-Hermite model of the true modified Rapp amplifier model. The precoder output by the GNN is used to initialize a low-complexity distortion-aware signal-to-interference-distortion (SINDR) tracking algorithm that jointly optimizes for power and admission control. Our proposed physics-based pipeline closely tracks a successive convex approximation (SCA) based solution in consumed power performance with the advantage of faster runtime and a wide operating range covered during training.