5G RAN Slicing with Load Balanced Handovers

2025-05-30

AI Summary

In 5G network slicing, multi-cell load balancing faces critical challenges—heterogeneous slice characteristics complicate overload/underload detection, while simultaneously ensuring QoS isolation and high resource utilization remains intractable. Method: This paper proposes a cross-slice global load-balancing criterion and introduces RadioWeaver, the first RAN slicing framework integrating dynamic quota allocation with intelligent service-cell reselection. RadioWeaver jointly optimizes quota distribution and handover decisions via trajectory-driven large-scale simulation and OpenRAN-based real-world validation. Contribution/Results: Compared to baseline approaches, RadioWeaver improves handover success rate by 16%–365%, significantly enhances slice-level QoS isolation, and boosts overall spectrum and computational resource utilization. It establishes a deployable, holistic load-balancing paradigm for sliced RAN architectures.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Hardware-aware ML
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Virtualization and resource management in Web systems and infrastructures
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

With increasing density of small cells in modern multi-cell deployments, a given user can have multiple options for its serving cell. The serving cell for each user must be carefully chosen such that the user achieves reasonably high channel quality from it, and the load on each cell is well balanced. It is relatively straightforward to reason about this without slicing, where all users can share a global load balancing criteria set by the network operator. In this paper, we identify the unique challenges that arise when balancing load in a multi-cell setting with 5G slicing, where users are grouped into slices, and each slice has its own optimization criteria, resource quota, and demand distributions, making it hard to even define which cells are overloaded vs underloaded. We address these challenges through our system, RadioWeaver, that co-designs load balancing with dynamic quota allocation for each slice and each cell. RadioWeaver defines a novel global load balancing criteria across slices, that allows it to easily determine which cells are overloaded despite the fact that different slices optimize for different criteria. Our evaluation, using large-scale trace-driven simulations and a small-scale OpenRAN testbed, show how RadioWeaver achieves 16-365% better performance when compared to several baselines.

Problem

• Balancing load in 5G RAN slicing with diverse slice criteria
• Defining overloaded vs underloaded cells in multi-slice environments
• Co-designing load balancing and dynamic quota allocation for slices

Innovation

• 5G RAN slicing with load balancing
• Dynamic quota allocation per slice
• Novel global load balancing criteria