InSlicing: Interpretable Learning-Assisted Network Slice Configuration in Open Radio Access Networks

2025-02-21

AI Summary

To address the weak adaptability of network slicing configuration in dynamic Open Radio Access Network (Open RAN) environments and the lack of interpretability in existing performance models, this paper proposes an end-to-end interpretable framework for slice performance modeling and joint optimization. We innovatively introduce the Kolmogorov–Arnold Network (KAN) for the first time to model slice performance functions, transforming black-box models into transparent, white-box representations. The framework synergistically combines the global search capability of genetic algorithms with the local refinement precision of trust-region gradient methods, ensuring both solution reliability and analytical tractability. Deeply integrated with standardized Open RAN interfaces, it enables real-time resource decision-making. Extensive experiments across multiple scenarios demonstrate over 25% reduction in operational costs, while delivering auditable, verifiable, and traceable slice resource allocation rationale—significantly enhancing system adaptability and management transparency.

Technology Category

• Search and Optimization: Algorithm Configuration
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Machine Learning: Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Search and Retrieval-Augmented AI: Web search models and ranking

Abstract

Network slicing is a key technology enabling the flexibility and efficiency of 5G networks, offering customized services for diverse applications. However, existing methods face challenges in adapting to dynamic network environments and lack interpretability in performance models. In this paper, we propose a novel interpretable network slice configuration algorithm (emph{InSlicing}) in open radio access networks, by integrating Kolmogorov-Arnold Networks (KANs) and hybrid optimization process. On the one hand, we use KANs to approximate and learn the unknown performance function of individual slices, which converts the blackbox optimization problem. On the other hand, we solve the converted problem with a genetic method for global search and incorporate a trust region for gradient-based local refinement. With the extensive evaluation, we show that our proposed algorithm achieves high interpretability while reducing 25+% operation cost than existing solutions.

Problem

• Interpretable network slice configuration
• Dynamic 5G network adaptation
• Performance cost reduction

Innovation

• Interpretable network slice configuration
• Kolmogorov-Arnold Networks integration
• Hybrid genetic and gradient optimization