Efficient Signal Demodulation Based on Multistage Interference Separation for MIMO-OTFS

In typical communication scenarios with numerous scatters, the Multiple-Input Multiple-Output (MIMO) Orthogonal Time Frequency Space (OTFS) system encounters various dimensional interference, such as inter-Doppler interference (IDI), inter-antenna interference (IAI), and inter-symbol interference (ISI). These interference present significant challenges to effective signal demodulation at the receiver. To address these challenges, we propose an iterative symbol detector (ISD) based on multistage interference separation in the time-space (TS) domain, leveraging its inherent sparsity. Specifically, we integrate an equalizer before the ISD to provide initial estimates and mitigate IDI. In each iteration, the ISD reconstructs and cancels IAI and ISI, then applies the Gauss-Seidel (GS) iteration method to solve the TS domain equations following interference cancellation. Simulation results demonstrate the effectiveness of the proposed ISD, showing its ability to significantly reduce interference in the MIMO-OTFS system. It outperforms benchmark algorithm in 4-QAM and 16-QAM and enhanced demodulation performance.