A Novel Slice Reconfiguration Method for Achieving QoS Guaranteeing and OPEX Saving in 5G Networks

Document Type

Article

Source of Publication

IEEE Transactions on Mobile Computing

Publication Date

10-22-2025

Abstract

Network slicing is recognized as one effective method to provide tailored services for various vertical industries in 5G and beyond. However, the traditional static slice configuration pattern struggles to address the issue of time-varying service loads. Thus leading to the degraded quality of service (QoS) and inefficient network resource utilization. Slice reconfiguration paves the way for achieving QoS-guaranteed and cost-efficient management of network slices by adjusting slice configurations when service load varies. In this paper, we propose a slice reconfiguration (SR) framework that enables network service providers (NSPs) to perform vertical/horizontal scaling operations to cope with the challenges brought by time-varying service load. We first model the slice reconfiguration as an mixed-integer programming problem to minimize NSP's service operating expenses (OPEX). However, its inherent non-convexity challenges direct optimization. Consequently, we reformulate it into a mixed-integer quadratically constrained programming-based SR problem (MIQCP-SR) enabling provably exact solutions. Given the prohibitive computational complexity of MIQCP-SR, we further develop one two-step slice reconfiguration (TS-SR) method: first, determine the function deployment, and second, execute the resource allocation. Extensive simulations show that our MIQCP-SR and TS-SR can effectively adapt the slice configuration to the changed service load. Compared with static configuration based on peak loads, MIQCP-SR and TS-SR reduce OPEX by 38% while still guaranteeing QoS requirements. Additionally, TS-SR outperforms other baselines and achieves 270 times faster solving speed than MIQCP-SR with only a 0.7% increase in cost.

ISSN

1536-1233

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Disciplines

Computer Sciences

Keywords

Dynamic Slice Load, Network Slicing, Operational Expenditure (OPEX), Quality of Service (QoS), Slice Reconfiguration

Scopus ID

105019931076

Indexed in Scopus

yes

Open Access

no

Share

COinS