Sensing Aided OTFS Massive MIMO Systems: Compressive Channel Estimation

Orthogonal time frequency space (OTFS) modulation has gained increasing interest due to its robustness to highDoppler fading channels. In this paper, we focus on MIMOOTFS systems that enjoy the high spectral efficiency of MIMO and the Doppler-resilience of OTFS. Interestingly, the MIMOOTFS wireless channel in the angle-delay-Doppler domain has a strong connection to the physical communication environment. In particular, the strong delay, Doppler, and angle taps of the channel are determined by the distance, velocity, and direction of the mobile users and the reflectors/scatterers in the environment. Therefore, prior information of the physical communication environment can be potentially utilized to aid and improve various communication tasks. To investigate this novel direction, we propose to exploit radars to obtain sensing information of the physical communication environment and leverage this information to aid the channel estimation for MIMO-OTFS systems. First, we formulate the MIMO-OTFS channel estimation problem as a sparse recovery problem. Then we utilize the radar sensing information to extract the strong angle-delay-Doppler taps of the sparse channel. We evaluate our radar-aided sparse channel recovery approach using co-existing radar and communication data generated by an accurate 3D ray-tracing framework. Simulation results show that the proposed channel estimation using radar sensing outperforms the conventional sparse recovery algorithms that do not utilize prior information of the communication environment.