Todorov, Apostol; Stoykova; Zlatev

IMPROVING SIGNAL STRENGTH ESTIMATION IN IOT USING WI-FI NETWORK PERFORMANCE DATA

2023

APPLIED RESEARCHES IN TECHNICS, TECHNOLOGIES AND EDUCATION

Тракийски университет - Стара Загора

1314-8796 [issn]

Article

eng

WiFi networks; signal strength prediction; regression modeling; additional signal features; accuracy improvement; network performance analysis; computational simulations; educational buildings

This study aims to improve signal strength prediction in WiFi networks by incorporating additional signal characteristics and using advanced regression modeling techniques. The research builds on existing work and offers a more accurate approach for predicting signal strength in WiFi networks. Using the RReliefF method, the influence of the main characteristics of the wireless signal on its quality has been established. Data were reduced using the Principal Components Analysis method. They were used in creating a complex predictive model for signal quality in a WiFi network. By analyzing key characteristics of wireless signals, the study achieved an increased prediction accuracy of 94%. The proposed regression model overcomes the previous limitations and offers a reliable method for predicting signal strength. The research goes beyond graphical representations and focuses on regression modeling to present more accurate estimates. The study includes computational simulations and real-world measurements to assess signal strength in an educational building. Furthermore, the research addresses the coverage issues of wireless sensor networks within WiFi networks, aiming to provide effective solutions to existing challenges. By overcoming the limitations, the study contributes to the overall improvement of WiFi networks. The obtained results have practical implications for network planning, optimization, wireless sensor networks, and network performance analysis. The research also paves the way for future studies to explore machine learning techniques, cross-layer analysis, and energy efficiency considerations in the context of WiFi networks. Overall, this study offers valuable information for predicting signal strength and improving WiFi network performance.