Reducing web page load time by using a dynamic scheduling method based on network environment

Abstract

In recent years, various information has been delivered through web pages using HTTP, making the acceleration of web page display a crucial concern. While HTTP/3 achieves the acceleration of web pages through prioritizing resources, there exists a challenge in establishing an optimal prioritization method. In this study, we addressed this issue by developing a novel approach that dynamically selects scheduling methods based on the network environment, comparing the performance of HTTP/3 prioritization methods across various network conditions. As a result, we demonstrated a reduction in page load times compared to conventional methods, particularly in network environments with low RTTs and low packet loss rates or high RTTs and high packet loss rates.