Page 56 - 2024S

UEC Int’l Mini-Conference No.52 49

Optimizing Wireless Networks for eXtended Reality (XR):

Implementing 3GPP XR Traffic Models with TCP-ECN

MuhammadReza HIDAYAT , Satoshi OHZAHATA , Ryo YAMAMOTO
∗
∗
Department of Information Science and Technology
The University of Electro-Communications
Tokyo, Japan

Keywords: eXtended Reality (XR), Congestion Control, Quality of Experience (QoE), Wireless Net-
works, 3GPP XR Traffic Model.

Abstract
This research focuses on implementing a 3GPP downlink XR traffic model for wireless networks
using TCP-ECN. The study explores the challenges of integrating eXtended Reality (XR) applications,
which encompass Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR), into Wi-Fi
networks. XR applications demand high bandwidth and low latency, particularly in multi-user scenarios,
which can strain WiFi networks. The research proposes a priority-based congestion control mechanism
using explicit congestion notification (ECN) marking to optimize traffic management and ensure fairness
for multiple XR users. This approach aims to address Wi-Fi limitations such as bandwidth constraints
and high latency in dense environments. The study utilizes the 3GPP XR traffic model, implemented in
the NS3 simulation environment, to evaluate the proposed method. The traffic model includes different
streams for AR, VR, and Cloud Gaming (CG), each with specific characteristics for control, video,
and audio data. The network setup in NS3 simulates realistic conditions with increasing numbers of
XR users, measuring performance and fairness. The research employs Quality of Experience (QoE)
metrics and Jain’s Fairness Index to assess the effectiveness of the proposed method.By comparing
scenarios with and without ECN for XR traffic, the study aims to demonstrate the benefits of the
proposed approach in supporting multiple XR applications simultaneously, identifying areas for network
performance improvement, and ensuring fair treatment of all applications.

∗ The author is supported by (AiQuSci) MEXT Scholar-
ship.