Home // SENSORCOMM 2025, The Nineteenth International Conference on Sensor Technologies and Applications // View article

Machine Learning-Based Joint TX Power and RX Sensitivity Control for Overlapping Basic Service Set Interference Mitigation in Dense Internet of Things Wireless Networks

Authors:
Jin-Min Lee
Hye-Yeon Shim
Il-Gu Lee

Keywords: Overlapping Basic Service Set; Machine Learning; TX power and RX sensitivity Control; Internet of Things Wireless Networks.

Abstract:
The recent increase in Internet of Things devices and wireless network equipment has led to frequent occurrences of overlapping basic service set environments, where multiple wireless networks share the same or adjacent channels within the same space. In these environments, network quality degrades owing to channel interference. Previous studies have attempted to avoid interference by blocking some links or using time-division methods; however, these methods have limitations in responding to real-time environmental changes and improving overall network throughput and spatial reuse rates. This study proposes a Machine Learning-based joint control technique for TX power and RX sensitivity. This technique is implemented in both centralized and distributed architectures. Each node recognizes the network state, predicts optimal parameters through a Machine Learning model, and applies them to minimize interference. Experimental results demonstrate that the proposed technique achieves up to 47.1% higher effective throughput and 29.6% better measured Signal-to-Interference-plus-Noise-Ratio compared with the conventional technique. The proposed distributed technique demonstrated approximately 46.4% higher effective throughput (21.43 Mbps) than the conventional central technique under low traffic load and maintained relatively high link quality even in environments with increased traffic load. While the proposed distributed method incurred higher control overhead owing to increased computational requirements compared with the conventional distributed method, the distributed architecture enables each Access Point to operate independently, allowing for parallel processing benefits in actual network deployments.

Pages: 14 to 19

Copyright: Copyright (c) IARIA, 2025

Publication date: October 26, 2025

Published in: conference

ISSN: 2308-4405

ISBN: 978-1-68558-304-0

Location: Barcelona, Spain

Dates: from October 26, 2025 to October 30, 2025