Model-Driven Deep Learning Assisted Detector for OTFS With Channel Estimation Error

Orthogonal Time Frequency Space (OTFS) technique has received significant research attentions as a potential solution for ensuring robust and reliable communication in high-mobility mobile communication scenarios in recent years. In this letter, we propose a model-driven deep learning based orthogonal approximate message passing (OAMP) detector to harvest the time and frequency diversity promised by OTFS modulation and improve the detection performance in the presence of channel estimation error. The intelligent detector is designed by unfolding the OAMP detector with adding several learnable parameters. Simulation results demonstrate that, with much reduced complexity and much fewer parameters, the proposed detector can achieve better error performance under channel estimation errors compared with other nonlinear detectors.