AImers-6G: AI-Driven Region-Temporal Resource Provisioning for 6G Immersive Services
Published in IEEE Wireless Communications, 2023
✨ Proposed Method
This paper introduces Almers-6G, an AI-driven approach designed for resource provisioning in 6G immersive services across both large and small regions. In large regions, the framework utilizes a context-immersive learning-based Lyapunov optimization algorithm—powered by an LSTM network—to allocate heterogeneous resources and satisfy user perception experiences. Meanwhile, for small regions, the paper proposes a secure Blockchain-based Double Dutch Auction (BDDA) mechanism to handle the matching and pricing determination of these resources dynamically.
📊 Experimental Results
Superior Efficiency: The Almers-6G scheme achieves the lowest average service energy consumption and latency compared to baselines.
Latency Reduction: Outperforms Queue Priority (QP) and Random methods in total latency by 16.20% and 36.56%, respectively.
Energy Savings: Reduces service energy consumption by 4.14% (vs. QP) and 17.77% (vs. Random).
Market Optimization: The BDDA approach improves social welfare by 11.02% and 10.53% when compared to traditional Double Auction (DA) and Double Dutch Auction (DDA) methods.
🤝 Collaborating Institutions
Tianjin University; Guangming Laboratory of Artificial Intelligence and Digital Economy (SZ); Tianjin University of Finance and Economics
Recommended citation: Qiu Chao, Zheyuan Chen, Xiaoxu Ren, Ziming Dai, Cheng Zhang, and Xiaofei Wang. "AImers-6G: AI-driven region-temporal resource provisioning for 6G immersive services." IEEE Wireless Communications 30, no. 3 (2023): 196-203.
Download Paper