MIMO-Zak-OTFS in Doubly-Selective Channels: I/O Relation Estimation and Detection

In this paper, we are concerned with Zak transform-based orthogonal time frequency space (Zak-OTFS) modulation in a multiple-input multiple-output (MIMO) setting with multiple antennas at the transmitter and receiver. We consider the problem of input-output (I/O) relation estimation and signal detection in MIMO-Zak-OTFS. Our contributions in this paper are as follows. First, we derive the system model for MIMO-Zak-OTFS. Next, we propose an I/O relation estimation scheme for MIMO-Zak-OTFS in a model-free framework. We evaluate the mean square error (MSE) performance of the proposed estimation scheme in Veh-A channel model with fractional delays and Dopplers, for different delay-Doppler (DD) pulse shaping filters, namely, sinc, Gaussian, and Gaussian-sinc filters. Gaussian filter is found to offer better MSE performance compared to sinc and Gaussian-sinc filters, because of the very low side lobes in Gaussian filter. Further, we assess the MIMO-Zak-OTFS signal detection performance using the estimated I/O relation. In this, Gaussian-sinc filter is found to achieve better bit error rate (BER) performance because of the good balancing characteristics of its main lobe and side lobes. We also show that a low-complexity local search based algorithm improves the minimum mean square error (MMSE) detection performance of MIMO-Zak-OTFS.