Predictive Modeling of Data Transmission Throughput in IEEE 802.11 Networks Using Diverse Machine Learning Frameworks: A Comparative Analysis

With the advent of AI-powered communications, efforts are being made in academia and standardization bodies to keep up with the growing complexity of future 5G and beyond networks. Despite its widespread use, enhancing network throughput in IEEE 802.11 remains challenging due to the fluctuating nature of wireless configurations and the existence of multiple interfering factors. This research attempts to enhance the performance of IEEE 802.11 networks using machine learning algorithms for calculating their throughput. Throughput prediction is critical for optimizing network performance and resource utilization, especially in scenarios where factors like interference, signal attenuation, and network congestion can significantly impact data transmission rates. In this paper, ML algorithms such as K-Nearest Neighbors (KNN), Support Vector Regression (SVR), and Linear Regression (LR) are used to characterize the relationship between the input parameters and the throughput. The results obtained illustrate that the proposed approach effectively selects relevant features, improves throughput prediction accuracy, and reduces computational overhead. Overall, this research contributes to advance the discipline of wireless communication by providing a reliable approach for forecasting throughput in IEEE 802.11 networks.