Towards Next Generation Immersive Applications in 5G Environments Conference: ; ;

Rohail Asim, Ankit Bhardwaj, Lakshmi Subramanian, Yasir Zaki · 2025

Read Towards Next Generation Immersive Applications in 5G Environments Conference: ; ; by Rohail Asim, Ankit Bhardwaj, Lakshmi Subramanian, Yasir Zaki (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-20050.txt.

Introduction

Rohail Asim and three co-authors present a framework for shared virtual reality over 5G in this two thousand twenty-five paper. Multi-user immersive apps serve teamwork, entertainment, and training. Wireless networks strain to meet their twin demands: huge bandwidth and tiny delay. Their framework, Hera, works at two levels. A top layer handles streaming and keeps users in sync. A lower layer controls send rates with an eye on quality of experience.

The rate control adapts to shifting wireless conditions while treating users fairly. Tests show Hera holds quality better than leading rival schemes. It keeps delay low even on rough 5G links. The work targets engineers who build immersive apps for mobile networks. It shows app logic and network control working as one.

Catalog record

CREATORRohail Asim, Ankit Bhardwaj, Lakshmi Subramanian, Yasir Zaki
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-20050

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Towards Next Generation Immersive Applications in 5G Environments Conference: ; ; Rohail Asim, Ankit Bhardwaj, Lakshmi Subramanian, Yasir Zaki Abstract . The Multi-user Immersive Reality (MIR) landscape is evolving rapidly, with applications spanning virtual collaboration, entertainment, and training. However, wireless network limitations create a critical bottleneck, struggling to meet the high-bandwidth and ultra-low latency demands essential for next-generation MIR experiences. This paper presents Hera, a modular framework for next-generation immersive applications, comprising a high-level streaming and synchronization layer for AR/VR systems and a low-level delay-based QoE-aware rate control protocol optimized for dynamic wireless environments. The Hera framework integrates application-aware streaming logic with a QoE-centric rate control core, enabling adaptive video quality, multi-user fairness, and low-latency communication across challenging 5G network conditions. We demonstrate that Hera outperforms existing state-of-the-art rate control algorithms by maintaining up to 66% lower latencies with comparable throughput performance, higher visual quality with 50% average bitrate improvements in our analysis, and improved fairness. By bridging the gap between application-level responsiveness and network-level adaptability, Hera lays the foundation for more scalable, robust, and high-fidelity multi-user immersive experiences. 1. Introduction The landscape of Multi-user Immersive Reality (MIR) technology, encompassing both Virtual Reality (VR) and Augmented Reality (AR), is undergoing a transformative period driven by substantial investments from industry leaders like Meta and Apple ( Inc. 2024a ; Inc. 2024b ; Inc. 2024c ; Times 2024 ; Inc. 2024d ) . With the rapid growth of high-bandwidth MIR applications, the demand for ultra-fast and low-latency wireless communication has surged across both indoor and outdoor environments. In indoor settings, high-performance MIR applications often rely on Wi-Fi technologies such as IEEE 802.11ad (WiGig) and IEEE 802.11ay, which operate in the 60 GHz frequency band to enable multi-gigabit data rates. Similarly, outdoor MIR applications increasingly leverage 5G networks, particularly millimeter-wave (mmWave) 5G, to support real-time streaming and interaction. However, despite their potential for high-speed wireless connectivity, both Wi-Fi and 5G networks present fundamental challenges, including limited range, susceptibility to signal attenuation, and high bandwidth variability, all of which threaten the seamless performance of MIR applications. Figure 1 . Real-world MIR application ecosystem. The end users can be indoors or outdoors and are connected to the application server by a 5G wireless channel. A critical yet often overlooked aspect of the infrastructure supporting multi-user immersive reality (MIR) applications illustrated in Figure 1 is whether existing systems, including transport protocols and application-level streaming architectures, can deliver the Quality of Experience (QoE) required for seamless interaction in dynamic wireless environments. Despite advances in mobile networking, the performance of MIR applications remains constrained not only by the limitations of traditional congestion control protocols but also by the lack of integration between network adaptation mechanisms and the application logic responsible for streaming, synchronization, and rendering. Addressing these issues requires a holistic framework that not only optimizes rate control at the transport level but also enables the application to dynamically adapt to changing network conditions to preserve the user experience. QoE Metric Description Startup delay The time users wait before the immersive content begins. Higher startup delay reduces perceived responsiveness at session start. Video bitrate / resolution level The visual clarity of the XR environment. Lower throughput forces the system to lower resolution or increase compression, reducing visual fidelity. Stall / buffering events Interruptions in the immersive experience where video or scene rendering pauses to rebuffer. Caused by throughput falling below content rate requirements. Interaction latency Delay between a user’s action (e.g., moving an object) and the visible response in the shared scene. Directly impacts perceived interactivity and collaboration smoothness. Collaborative fluency index The smoothness and synchronicity of shared actions among users, largely driven by latency. High latency leads to disjointed collaborative interactions. Table 1 . MIR QoE metrics and their user impact. 1.1. Challenges 1.1.1. Application Requirements Currently available MIR applications showcase some of the potential applications enabled by MIR but the Quality of Experience (QoE), in terms of key metrics summarized in Table 1 such as video resolution, interaction latency, and collaborative fluency, for these applications is restricted by various network performance bottlenecks ( Liu et al. 2020 ; Struye et al. 2024 ; Perfecto et al. 2020 ) . To circumvent these bottlenecks, many of these applications only support 2D experiences, which create a large “screen” in the immersive environment. Others that provide an immersive experience are forced to reduce the frames-per-second (FPS) for the application, which reduces how smooth the application feels. Alternatively, the application displays the content at a lower resolution, which reduces the graphical fidelity, or freezes momentarily, breaking the users’ immersion ( Tripathi et al. 2024 ) . For a high-fidelity and comfortable MIR experience, a target frame rate (FPS) of at least 120 FPS is recommended, as lower frame rates can cause motion sickness due to the increased discrepancy between visual input and vestibular perception ( Wang et al. 2023 ) . At the same time, 60 FPS and 90 FPS are often considered as a baseline and target for today’s MIR systems ( Casasnovas et al. 2024 ; Tan et al. 2024 ) . In terms of resolution, a target video quality of at least 4K resolution per eye ( Struye et al. 2022 ) is required to mitigate the “screen-door effect,” a common VR artifact where inter-pixel gaps become visible and compromise the visual fidelity ( Cho et al. 2017 ) . A high-quality, immersive, multi-user immersive reality experience featuring 360-degree 3D video at 120 frames per second (FPS) and 4K resolution per eye necessitates a downlink bandwidth of at least 100 Mbps per user to accommodate the high data rate associated with 360-degree video capture, stereoscopic 3D rendering, and real-time multi-user synchronization. Furthermore, end-to-end latency must remain below 20 ms to ensure a comfortable and responsive user experience, minimizing motion sickness and maximizing the sense of presence. This latency budget encompasses both network transmission delays and any processing overhead. Conventional congestion control algorithms, like BBR ( Cardwell et al. 2016 ) and TCP ( Allman et al. 2009 ) , are often insufficient to satisfy these stringent requirements. While UDP is commonly used for real-time communication due to its low overhead and minimal latency, many multi-user immersive applications require reliable delivery to ensure consistency across users, prevent visual artifacts, and maintain the integrity of complex shared scene states. TCP becomes essential in scenarios where packet loss or out-of-order delivery could disrupt synchronized interactions, collaborative object manipulation, or the seamless rendering of high-fidelity visual elements. BBR, while designed to maximize throughput, can exhibit performance degradation in lossy environments and may not consistently deliver the low latency necessary for VR. Cubic ( Ha et al. 2008 ) , optimized for TCP flows, is similarly challenged by the real-time nature of VR streaming, as its congestion control mechanisms can introduce unacceptable delays, particularly in dynamic network conditions. The high bandwidth demands and strict latency constraints of high-fidelity VR necessitate the exploration and development of specialized network protocols and rate control algorithms tailored for real-time media streaming and interactive applications. 1.1.2. 5G Channel Variability Figure 2 . Cellular channels’ variability analysis. The 5G channel provides higher mean bandwidth, but it comes at the cost of higher bandwidth fluctuations. Data generated through experiments

Read the complete text