Performance Analysis of OTFS Modulation with Receive Antenna Selection

TL;DR

This paper evaluates the diversity performance of OTFS modulation with receive antenna selection, highlighting the impact of phase rotation on diversity gains in SIMO, MIMO, and space-time coded OTFS systems.

Contribution

It provides a comprehensive diversity analysis of OTFS with antenna selection, revealing the effects of phase rotation on diversity extraction in various system configurations.

Findings

Without phase rotation, SIMO and MIMO OTFS with RAS are rank deficient.

Phase rotation enables SIMO and STC-OTFS with RAS to achieve full diversity.

MIMO-OTFS with RAS achieves full DD diversity with phase rotation but not full receive diversity.

Abstract

In this paper, we analyze the performance of orthogonal time frequency space (OTFS) modulation with antenna selection at the receiver, where $n_s$ out of $n_r$ receive antennas with maximum channel Frobenius norms in the delay-Doppler (DD) domain are selected. Single-input multiple-output OTFS (SIMO-OTFS), multiple-input multiple-output OTFS (MIMO-OTFS), and space-time coded OTFS (STC-OTFS) systems with receive antenna selection (RAS) are considered. We consider these systems without and with phase rotation. Our diversity analysis results show that, with no phase rotation, SIMO-OTFS and MIMO-OTFS systems with RAS are rank deficient, and therefore they do not extract the full receive diversity as well as the diversity present in the DD domain. Also, Alamouti coded STC-OTFS system with RAS and no phase rotation extracts the full transmit diversity, but it fails to extract the DD diversity. On the other hand, SIMO-OTFS and STC-OTFS systems with RAS become full-ranked when phase rotation is used, because of which they extract the full spatial as well as the DD diversity present in the system. Also, when phase rotation is used, MIMO-OTFS systems with RAS extract the full DD diversity, but they do not extract the full receive diversity because of rank deficiency. Simulation results are shown to validate the analytically predicted diversity performance.