This repository contains the companion code, datasets, and documentation for the paper:

Toward Scalable VR-Cloud Gaming: An Attention-aware Adaptive Resource Allocation Framework for 6G Networks
This paper is currently under revision for the Journal of Network and Computer Applications (JNCA).

This work proposes a holistic and scalable resource allocation framework for Virtual Reality Cloud Gaming (VR-CG) systems operating over future 6G networks. VR-CG is an emerging class of immersive, highly interactive applications that impose tight demands on network, computing, and transport resources. Meeting these demands requires intelligent coordination across the end-to-end system.

The paper introduces a multi-stage optimization framework that integrates wireless access, computing–network convergence, semantic communication principles, and transport routing into a unified design. The approach is grounded in 3GPP specifications and evaluated using large-scale scenarios and real data sources.

• 6G-Aligned VR-CG Architecture:
  A full end-to-end system model incorporating Computing Network Convergence (CNC), multi-hop transport, cloud/edge computing nodes, and semantic-aware video processing.

• Attention-Aware Semantic Transmission Pipeline:
  A novel pipeline combining object segmentation, attention prediction, and object-centric encoding to adapt VR-CG video quality according to regions of interest.

• Multi-Stage Optimization Framework:
  A scalable formulation that jointly optimizes:

  • User association
  • Wireless resource allocation
  • VR-CG game-engine placement
  • Adaptive multipath routing
  • Fine-grained PRB scheduling

• Heuristic Algorithms:
  Practical and scalable algorithms designed to solve each stage of the optimization framework with low computational cost and near-optimal performance.

• Large-Scale Evaluation:
  Extensive simulations demonstrating gains in QoE, latency, cost reduction, throughput capability, and scalability compared to baseline and state-of-the-art solutions.