EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Machine Learning Approach for Signal Detection Using Rational Function Curve Fitting in Slow Flat Fading MIMO Communications

Hamid Nourollahi, S. Abolfazl Hosseini, Ali Shahzadi

Engineering Reports · 2026-10 · 8 (10)

DOI: 10.1002/eng2.71103 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Engineering Reports, online erschienen am 16.10.2026. Verfasst von Hamid Nourollahi und 2 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Ensuring reliable signal detection and accurate channel estimation is fundamental to the success of modern communication systems. While various methods have been introduced to address these challenges, classical theoretical detection schemes primarily rely on comparing the received signal with a predetermined transmitter signal constellation. In this paper, we propose a novel scheme for signal detection in MIMO communication, leveraging curve fitting (CF) and supervised machine learning. This approach classifies received signals without relying on explicit knowledge of the signal constellation, modulation type, channel estimation, or the specific transmitter‐receiver structure of an N t × N r MIMO system. Specifically, a specific rational function is fitted to the samples of each received symbol waveform, where the coefficients of the numerator and denominator polynomials are extracted as distinctive features. We apply the proposed method to MIMO‐QPSK modulation systems transmitting over slow, flat Rayleigh fading channels with independent MIMO links. In this process, each corrupted received symbol is sampled, and various rational fraction curves with different polynomial degrees are fitted to the signal samples. Subsequently, an artificial neural network (ANN) classifier utilizes the polynomial coefficients to classify the fitted curves. Simulation results demonstrate that the proposed classification method achieves a lower error rate compared to classical maximum likelihood (ML), minimum mean square error (MMSE), and zero forcing (ZF) methods in a slow, flat Rayleigh fading channel.

Zitieren

APA

Nourollahi, H., Hosseini, S. A., & Shahzadi, A. (2026). A Machine Learning Approach for Signal Detection Using Rational Function Curve Fitting in Slow Flat Fading MIMO Communications. Engineering Reports, 8(10). https://doi.org/10.1002/eng2.71103

Vancouver

Nourollahi H, Hosseini SA, Shahzadi A. A Machine Learning Approach for Signal Detection Using Rational Function Curve Fitting in Slow Flat Fading MIMO Communications. Engineering Reports. 2026;8(10). doi:10.1002/eng2.71103

BibTeX
@article{Nourollahi2026,
  author = {Nourollahi, Hamid and Hosseini, S. Abolfazl and Shahzadi, Ali},
  title = {A Machine Learning Approach for Signal Detection Using Rational Function Curve Fitting in Slow Flat Fading MIMO Communications},
  journal = {Engineering Reports},
  year = {2026},
  volume = {8},
  number = {10},
  doi = {10.1002/eng2.71103}
}
Kostenlos starten →