Fronthaul and backhaul requirements of flexibly centralized radio access networks

TLDR

Cloud radio access networks offer significant benefits over decentralized architectures but impose demanding fronthaul and backhaul requirements, which flexible centralization can mitigate by adaptively allocating processing tasks based on load, user scenario, and fronthaul availability. The article surveys functional split options and analyzes their specific requirements. The authors compare these requirements with existing fronthaul technologies and discuss the convergence of fronthaul and backhaul solutions. Evaluating aggregated fronthaul traffic demonstrates the benefits of flexible centralization and provides guidelines to prevent over‑ or under‑dimensioning of the fronthaul network.

Abstract

Cloud radio access networks promise considerable benefits compared to decentralized network architectures, but they also put challenging requirements on the fronthaul and backhaul network. Flexible centralization can relax these requirements by adaptively assigning different parts of the processing chain to either the centralized baseband processors or the base stations based on the load situation, user scenario, and availability of fronthaul links. In this article, we provide a comprehensive overview of different functional split options and analyze their specific requirements. We compare these requirements to available fronthaul technologies, and discuss the convergence of fronthaul and backhaul technologies. By evaluating the aggregated fronthaul traffic, we show the benefits of flexible centralization and give guidelines on how to set up the fronthaul network to avoid over- or under-dimensioning.