Efficient Message-Passing Detection for Multi-Satellite Systems Using OTFS Modulation

In this study, we analyze a low Earth orbit multi-satellite communication system using orthogonal time frequency space (OTFS) modulation. In this context, the choice of the OTFS modulation is driven by its robustness against high Doppler shifts in doubly-selective channels. This scenario is motivated by the need for a more stable and reliable system, which can be achieved through diversity, i.e., allowing each user to be jointly served by multiple satellites. However, numerous challenges arise due to the need to properly combine the different received signals, each one characterized by its own Doppler shift, delay and phase. This paper focuses on soft-output data detection algorithms and proposes a novel message passing (MP)-based approach that leverages both the channel sparsity in the Doppler-delay domain and the particular structure assumed by the equivalent channel matrix, derived from the compact block-wise input-output relation, according to the Forney observation model for linear modulations over additive white Gaussian noise channels. The proposed approach is the first one able to significantly reduce the complexity by acting on the choice of interferers, organized in diagonals, and on the schedule, by prioritizing the strongest elements. We assess the detector's performance by evaluating its pragmatic capacity, i.e., the achievable rate of the channel induced by the signal constellation and the detector soft-output. Simulation results demonstrate that the proposed solution achieves an optimal trade-off between complexity and performance compared to the alternative state-of-the-art solutions.