Optimized Active Measurement Configurations for Monitoring WebRTcNetwork Behavior

This paper introduces a novel sampling method, customized stratified-cluster sampling, to optimize the active measurement configuration for simulating WebRTC network behavior, thereby enhancing measurement accuracy and ensuring consistency with actual WebRTC network behavior. The optimization aims to align the measurement results closely with the real-world performance of WebRTC by focusing on the packet inter-arrival interval's mean and standard deviation as key network indicators. Through a comparative experiment with non-optimized measurement configurations, our optimized approach demonstrates a higher degree of similarity to the true WebRTC network behavior across both statistical dimensions and time series features.