RSFLoc: Robust Smart Fingerprint Localization-Based Hybrid PSO– ANN Approach in Complex Environment
Management Strategies and Engineering Sciences · 2026 · 8 (4) · 1-13
DOI: 10.61838/msesj.327 ↗Kurz gefasst
Fachartikel in der Zeitschrift Management Strategies and Engineering Sciences, 2026. Verfasst von Rawaa Akram Mohammad Ali und 3 weiteren Autor:innen.
Abstract (Originalsprache)
The rapid proliferation of Internet of Things (IoT)-enabled wireless localization systems has gained prominence in electrical engineering applications, particularly in domains such as healthcare, where precise indoor positioning is essential for tracking personnel and assets. This paper introduces the Robust Smart Fingerprint Localization (RSFLoc) model, leveraging Received Signal Strength Indicator (RSSI) fingerprinting in wireless communication networks. The model enhances localization accuracy through the optimization of an Artificial Neural Network (ANN) using Particle Swarm Optimization (PSO), addressing key signal processing challenges like multipath propagation, shadowing, and environmental dynamics in indoor environments. In the offline phase, RSSI data from multiple Access Points (APs) across diverse zones are rigorously analyzed to evaluate signal robustness and model path loss (PL) characteristics. The online phase deploys the PSO-optimized ANN on a simulated ESP32 microcontroller interfaced with MATLAB/Simulink for real-time x-y coordinate estimation. Localization results are transmitted via IoT protocols to the ThingSpeak cloud platform, enabling visualization and remote monitoring through a mobile application. Experimental results demonstrate high fidelity between measured and estimated PL curves, with a low Mean Square Error (MSE) of 1.001 dB for RSSI-driven PL modeling (ranging from 3.2723 × 10^{-4} to 1.001 dB). The ANN model's validation MSE achieves 2.0956 m, outperforming training (MSE: 0.0557 m post-optimization) and testing phases due to improved hyperparameter tuning via PSO, which enhances convergence stability. Regression analysis reveals strong linearity (R² > 0.95) between predicted and actual locations, while error histograms indicate an 85% reduction in localization error compared to baseline methods, underscoring the model's efficacy in practical wireless communication systems.
Zitieren
APA
Ali, R. A. M., Majma, N., Al-Nakkash, A. H., & Soltanaghaei, M. (2026). RSFLoc: Robust Smart Fingerprint Localization-Based Hybrid PSO– ANN Approach in Complex Environment. Management Strategies and Engineering Sciences, 8(4), 1-13. https://doi.org/10.61838/msesj.327
Vancouver
Ali RAM, Majma N, Al-Nakkash AH, Soltanaghaei M. RSFLoc: Robust Smart Fingerprint Localization-Based Hybrid PSO– ANN Approach in Complex Environment. Management Strategies and Engineering Sciences. 2026;8(4):1-13. doi:10.61838/msesj.327
BibTeX
@article{Ali2026,
author = {Ali, Rawaa Akram Mohammad and Majma, Negar and Al-Nakkash, Aseel Hameed and Soltanaghaei, Mohammadreza},
title = {RSFLoc: Robust Smart Fingerprint Localization-Based Hybrid PSO– ANN Approach in Complex Environment},
journal = {Management Strategies and Engineering Sciences},
year = {2026},
volume = {8},
number = {4},
pages = {1--13},
doi = {10.61838/msesj.327}
}
EUVIMED