Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Adaptive Context-Aware Multi-Path Transmission Control for VR/AR Content: A Deep Reinforcement Learning Approach

Shakil Ahmed, Saifur Rahman Sabuj, Ashfaq Khokhar

TL;DR

This work addresses the stringent QoS requirements of AR/VR streaming by recognizing that traditional TCP/MPTCP struggle in dynamic, multi-path environments. It introduces Adaptive Context-Aware Multi-Path Transmission Control (ACMPTC), a DRL-based, multi-agent framework that jointly optimizes path selection, congestion control, and bandwidth allocation in real time, guided by a feedback loop and a utility-based objective. The approach combines a formal ACMPTC model with per-stream DRL agents (A2C) that operate with localized state, actions, and rewards to achieve scalable, decentralized optimization, alongside an explicit complexity analysis. Simulation on a 360-degree video dataset demonstrates ACMPTC's superior QoS under AR/VR-like conditions, outperforming TCP and standard MPTCP in throughput, latency, and resilience to congestion. Overall, ACMPTC offers a principled, data-driven path management solution with practical implications for high-bandwidth, low-latency AR/VR delivery in future multi-path networks.

Abstract

This paper introduces the Adaptive Context-Aware Multi-Path Transmission Control Protocol (ACMPTCP), an efficient approach designed to optimize the performance of Multi-Path Transmission Control Protocol (MPTCP) for data-intensive applications such as augmented and virtual reality (AR/VR) streaming. ACMPTCP addresses the limitations of conventional MPTCP by leveraging deep reinforcement learning (DRL) for agile end-to-end path management and optimal bandwidth allocation, facilitating path realignment across diverse network environments.

Adaptive Context-Aware Multi-Path Transmission Control for VR/AR Content: A Deep Reinforcement Learning Approach

TL;DR

This work addresses the stringent QoS requirements of AR/VR streaming by recognizing that traditional TCP/MPTCP struggle in dynamic, multi-path environments. It introduces Adaptive Context-Aware Multi-Path Transmission Control (ACMPTC), a DRL-based, multi-agent framework that jointly optimizes path selection, congestion control, and bandwidth allocation in real time, guided by a feedback loop and a utility-based objective. The approach combines a formal ACMPTC model with per-stream DRL agents (A2C) that operate with localized state, actions, and rewards to achieve scalable, decentralized optimization, alongside an explicit complexity analysis. Simulation on a 360-degree video dataset demonstrates ACMPTC's superior QoS under AR/VR-like conditions, outperforming TCP and standard MPTCP in throughput, latency, and resilience to congestion. Overall, ACMPTC offers a principled, data-driven path management solution with practical implications for high-bandwidth, low-latency AR/VR delivery in future multi-path networks.

Abstract

This paper introduces the Adaptive Context-Aware Multi-Path Transmission Control Protocol (ACMPTCP), an efficient approach designed to optimize the performance of Multi-Path Transmission Control Protocol (MPTCP) for data-intensive applications such as augmented and virtual reality (AR/VR) streaming. ACMPTCP addresses the limitations of conventional MPTCP by leveraging deep reinforcement learning (DRL) for agile end-to-end path management and optimal bandwidth allocation, facilitating path realignment across diverse network environments.
Paper Structure (43 sections, 23 equations, 5 figures, 1 table, 1 algorithm)

This paper contains 43 sections, 23 equations, 5 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Background and Significance
  3. Related Work and Contribution
  4. Evolution of MPTCP
  5. Advances in DRL for Network Optimization
  6. Gaps in Current Research
  7. Contributions
  8. Theoretical Framework: MPTCP
  9. MPTCP sub-flows and Throughput
  10. Traffic Allocation and Optimization
  11. Limitations of MPTCP
  12. ACMPTC Model
  13. Network Model
  14. Dynamic Path Selection, Assignment, and Adjustments
  15. Initial Path Assignment
  16. ...and 28 more sections

Figures (5)

  • Figure 1: The source device utilizes multiple network paths (Wi-Fi, 4G, 5G, Ethernet) to send data to the server. The MPTCP protocol optimizes the use of these paths by managing traffic across sub-flows, improving throughput and fault tolerance.
  • Figure 2: AR/VR devices send data to the ACMPTC server, which, with insights from the DRL agent, dynamically optimizes data transmission paths to enhance AR/VR streaming quality
  • Figure 3: Network metrics analysis
  • Figure 4: Network performance insights
  • Figure 5: Network dynamics examination