Combined Multi-Agent and Centralized Resource Allocation in Cloud Radio Access Networks

ABSTRACT

Cloud Radio Access Networks (C-RAN) have emerged as a promising architecture to meet the increasing demands for high data rates and efficient resource utilization in next-generation wireless systems. However, resource allocation in C-RAN remains a challenging task due to the dynamic network conditions, interference management, and the need for scalable coordination among distributed units. Traditional centralized approaches offer global optimization but suffer from high computational complexity and latency, while fully distributed methods lack global awareness, leading to suboptimal performance.

This work proposes a hybrid resource allocation framework that combines multi-agent distributed intelligence with centralized coordination to leverage the strengths of both paradigms. In the proposed approach, multiple agents deployed at remote radio heads make local decisions based on real-time channel conditions, while a central controller performs periodic global optimization to ensure fairness and overall system efficiency. The framework integrates adaptive power allocation, user association, and interference mitigation strategies.

The methodology involves modeling the system as a cooperative multi-agent environment, where agents interact with the network using learning-based or heuristic policies, supported by a central unit that refines decisions through global network state information. Simulation results demonstrate that the proposed hybrid approach achieves improved spectral efficiency, reduced latency, and better load balancing compared to purely centralized or distributed schemes.

Overall, the proposed combined multi-agent and centralized resource allocation scheme provides a scalable and efficient solution for future C-RAN deployments, enabling enhanced network performance under dynamic and dense wireless environments.

Keywords: Cloudradio access network, multi-agent system, resource allocation.