Identification of Cellular Signal Measurements Using Extreme Learning Machine

Author's Department

Electronics & Communications Engineering Department

Find in your Library

https://doi.org/10.1109/JIOT.2025.3532584

All Authors

Esraa A. Makled Ibrahim Al-Nahhal Octavia A. Dobre Oktay Ureten Hyundong Shin

Document Type

Research Article

Publication Title

IEEE Internet of Things Journal

Publication Date

1-1-2025

doi

10.1109/JIOT.2025.3532584

Abstract

Intelligent radios play a pivotal role in optimizing communication resources for both commercial and military applications. Automatic signal identification (ASI) serves as a crucial component for intelligent radios, with likelihood-based and feature-based ASI algorithms being conventional approaches. Recent studies have explored the integration of machine learning (ML) algorithms for ASI, revealing their enhanced resilience to channel distortions compared to traditional methods. This article proposes the application of an extreme learning machine (ELM), a type of the ML algorithm, for the identification of cellular signals based on over-the-air measurements of power spectral density (PSD). The proposed ELM undergoes evaluation using two distinct datasets of PSDs to assess identification accuracy, with the first dataset utilized for hyperparameter optimization and the second unseen dataset employed to evaluate robustness and generality. The experimental results showcase improved performance in both accuracy and training complexity compared to recent work in the literature.

First Page

15491

Last Page

15500

Link to Full Text
Find this item online

Share

COinS