Reservoir Computing in Wireless Receive Processing: An Information-Theoretic Perspective
IEEE Journal on Selected Areas in Information Theory · 2026 · 7 · 146-160
DOI: 10.1109/jsait.2026.3683606 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Journal on Selected Areas in Information Theory, 2026. Verfasst von Shashank Jere und 3 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Gefördert von: National Science Foundation.
Abstract (Originalsprache)
Deep learning is playing a transformational role in the physical layer of wireless communications. While demonstrating remarkable empirical performance in applications such as MIMO detection and receive processing, the fundamental reasons underlying this superlative performance remain insufficiently understood. In this work, we advance the development of Explainable AI (xAI) within the physical layer by grounding the analysis of deep learning methods in estimation-theoretic and information-theoretic principles. Specifically, we focus on wireless receive processing, namely channel equalization through reservoir computing (RC)—a recurrent neural network (RNN) framework that has shown strong empirical receive processing performance in both Orthogonal Frequency Division Multiplexing (OFDM) and Orthogonal Time Frequency Space (OTFS) systems. Within RC, we focus on the echo state network (ESN) and its extension, the windowed ESN (WESN). Building on our prior work that establishes the signal processing foundations of the ESN in receive processing, we conduct an exact analytical characterization of the equalization performance of the ESN and WESN and compare them against the optimum Wiener equalizer in a wireless fading channel with inter-symbol interference (ISI). The validity of the theoretical results is established through extensive numerical evaluations.
Förderung
- National Science Foundation
Zitieren
APA
Jere, S., Zheng, L., Khan, U. S., & Liu, L. (2026). Reservoir Computing in Wireless Receive Processing: An Information-Theoretic Perspective. IEEE Journal on Selected Areas in Information Theory, 7, 146-160. https://doi.org/10.1109/jsait.2026.3683606
Vancouver
Jere S, Zheng L, Khan US, Liu L. Reservoir Computing in Wireless Receive Processing: An Information-Theoretic Perspective. IEEE Journal on Selected Areas in Information Theory. 2026;7:146-160. doi:10.1109/jsait.2026.3683606
BibTeX
@article{Jere2026,
author = {Jere, Shashank and Zheng, Lizhong and Khan, Ummay Sumaya and Liu, Lingjia},
title = {Reservoir Computing in Wireless Receive Processing: An Information-Theoretic Perspective},
journal = {IEEE Journal on Selected Areas in Information Theory},
year = {2026},
volume = {7},
pages = {146--160},
doi = {10.1109/jsait.2026.3683606}
}
EUVIMED