Researchers have developed CoSlice, a novel framework utilizing coordinated multi-agent reinforcement learning for resource allocation in 5G network slicing within Open Radio Access Networks (O-RAN). This approach employs preference-conditioned agents for rapid allocation and an LLM-based coordinator for slower, global objective adaptation. The system is formulated as a multi-objective partially observable stochastic game, and experimental results demonstrate its effectiveness in reducing SLA violations compared to existing methods. AI
IMPACT This framework could improve efficiency and reliability in future 5G network deployments.
RANK_REASON The cluster contains a research paper detailing a novel framework for network resource allocation.
Read on arXiv cs.MA (Multiagent) →
- 5G
- CoSlice
- Open Radio Access Network
- ORANSlice
- Preference Correlated Equilibrium
- RFSIM: A feature based image quality assessment metric using Riesz transforms
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →