Uplink Sensing in Vehicular Networks with OTFS Signaling: A LSTM-based Approach

The integrated sensing and communication (ISAC) has been considered as an enabling technology in the next generation wireless communications. In this paper, we propose a novel uplink sensing scheme based on orthogonal time frequency space (OTFS) modulation and deep learning (DL) to realize the ISAC framework in vehicular networks. Specifically, the initial estimates of motion-related parameters (e.g., angles and locations) are generated by the multiple signal classification (MUSIC) and threshold-based algorithms with the assistance of OTFS modulation. Subsequently, we develop a long short-term memory (LSTM)-based method to implicitly learn the features of historical channels and predict the trajectories of the vehicles. Finally, simulation results show that the proposed method can achieve satisfactory sensing performance with low signal overhead.