Discrete ZAK-OTFS With/Without IRS: Design of Low-Complexity LMMSE Receiver and Optimal Reflection Coefficients
IEEE Transactions on Wireless Communications · 2026 · 25 · 25504-25519
DOI: 10.1109/twc.2026.3745121 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Wireless Communications, 2026. Verfasst von Rakesh Kumar Yadav und 3 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben. Gefördert von: Anusandhan National Research Foundation (ANRF), Government of India.
Abstract (Originalsprache)
Recently, in contrast to multi-carrier orthogonal time frequency space (MC-OTFS), discrete Zak-transform-based OTFS (DZT-OTFS) has gained attention as a potential alternative owing to its superior performance for high Doppler scenario and lower complexity. In this work, our initial aim is to design the low-complexity linear equalizer for DZT-OTFS system. Unlike MC-OTFS for integer Doppler and delay indices, the delay–Doppler (DD) channel matrix in DZT-OTFS for fractional Doppler and delay indices does not exhibit sparsity, thereby resulting in a challenging low-complexity receiver design problem in DD domain. Here, we consider two scenarios, firstly, without limiting the effective pulse at the output of matched filter (MF), we observe non-sparse DD channel matrix. Next, with a sufficient large number of delay bins, the effective pulse can be limited to only a few significant non-zero values, yielding a highly sparse DD channel matrix. For these two scenarios, we propose low-complexity linear minimum mean-squared error (LMMSE) equalizer by developing novel 1) Cholesky-based decomposition algorithm, and 2) forward and backward substitution techniques. In addition, by amalgamating an intelligent reflecting surface (IRS) with OTFS system, one can achieve improved bit-error-rate (BER) and spectral efficiency, as it adapts the propagation environment in high Doppler scenarios. Therefore, our another objective is to propose a novel IRS-aided DZT-OTFS system by designing optimal IRS coefficients to maximize uplink achievable-rate (AR) under imperfect channel state information (CSI). Here, we formulate AR based optimization problem, which can be solved using proposed coordinate-ascent (CA) based technique. Through simulation results, we initially demonstrate the efficacy of the proposed LMMSE equalizers for DZT-OTFS system by comparing its BER performance as well as the computational calculations with the conventional and state-of-the-art equalizers. Moreover, we present the AR and BER performances of the IRS-aided DZT-OTFS system under the proposed optimization framework and compare the results with the settings: 1) DZT-OTFS without IRS, 2) randomly selected IRS phases, 3) with perfect CSI conditions, and 4) existing state-of-the-art techniques.
Förderung
- Anusandhan National Research Foundation (ANRF), Government of India
Zitieren
APA
Yadav, R. K., Mukherjee, D., Mishra, H. B., & Mukhopadhyay, S. (2026). Discrete ZAK-OTFS With/Without IRS: Design of Low-Complexity LMMSE Receiver and Optimal Reflection Coefficients. IEEE Transactions on Wireless Communications, 25, 25504-25519. https://doi.org/10.1109/twc.2026.3745121
Vancouver
Yadav RK, Mukherjee D, Mishra HB, Mukhopadhyay S. Discrete ZAK-OTFS With/Without IRS: Design of Low-Complexity LMMSE Receiver and Optimal Reflection Coefficients. IEEE Transactions on Wireless Communications. 2026;25:25504-25519. doi:10.1109/twc.2026.3745121
BibTeX
@article{Yadav2026,
author = {Yadav, Rakesh Kumar and Mukherjee, Debasmita and Mishra, Himanshu B. and Mukhopadhyay, Samrat},
title = {Discrete ZAK-OTFS With/Without IRS: Design of Low-Complexity LMMSE Receiver and Optimal Reflection Coefficients},
journal = {IEEE Transactions on Wireless Communications},
year = {2026},
volume = {25},
pages = {25504--25519},
doi = {10.1109/twc.2026.3745121}
}
EUVIMED