Performance Evaluation of Video Streaming Applications with Target Wake Time in Wi-Fi 6

TL;DR

This paper evaluates Wi-Fi 6's Target Wake Time feature for optimizing video streaming by analyzing traffic patterns and proposing a synthetic traffic generator, demonstrating maintained QoS under congestion.

Contribution

It introduces a synthetic video streaming traffic generator and a two-stage TWT scheduling approach, advancing resource management in Wi-Fi 6 networks.

Findings

TWT can sustain QoS for streaming traffic under congestion

Synthetic traffic model accurately mimics real-world streaming patterns

TWT duty cycle and MF jointly optimize scheduling

Abstract

The Target Wake Time (TWT) feature, introduced in Wi-Fi 6, was primarily meant as an advanced power save mechanism. However, it has some interesting applications in scheduling and resource allocation. TWT-based resource allocation can be used to improve the user experience for certain applications, e.g., VoIP, IoT, video streaming, etc. In this work, we analyze the packet arrival pattern for streaming traffic and develop a synthetic video streaming traffic generator that mimics real-world streaming traffic. We propose a two-stage approach where we calculate the TWT duty cycle in the first step. In the subsequent step, we determine the Multiplication Factor(MF), which jointly dictates the required TWT schedule for the synthetic traffic model. Initial testing shows that key QoS metrics can be met for sustained performance of synthetic traffic upon enabling TWT, even in the presence of peak background congestion in the network.