Authors: Hamidreza Mazandarani, Masoud Shokrnezhad, Tarik Taleb

Published on: January 12, 2024

Impact Score: 8.07

Arxiv code: Arxiv:2401.06308

Summary

• What is new: The introduction of a Semantic-Aware Multi-Agent Double and Dueling Deep Q-Learning (SAMA-D3QL) technique for wireless spectrum allocation.
• Why this is important: Existing semantic extraction techniques do not fully address the requirements and characteristics of future 6G-based systems in resource allocation decision-making.
• What the research proposes: A new formulation for multiple access to the wireless spectrum and a SAMA-D3QL technique to optimize the utilization-fairness trade-off with user data correlation.
• Results: SAMA-D3QL outperforms alternative approaches in both single-channel and multi-channel scenarios.

Technical Details

Technological frameworks used: Model-free Multi-Agent Deep Reinforcement Learning (MADRL)

Models used: Semantic-Aware Multi-Agent Double and Dueling Deep Q-Learning (SAMA-D3QL)

Data used: User data correlation, association matrices, and channels

Potential Impact

Telecommunications providers, wireless spectrum regulators, 6G technology developers, and companies planning to leverage federated and dynamic 6G applications

Want to implement this idea in a business?

We have generated a startup concept here: OptiSpectrum.