2023/12/28 by Hnin Pann Phyu, Phyu, Hnin Pann, Razvan Stanica +5
Computer Science · #Advanced Technologies in Various Fields #FOS: Computer and information sciences #IoT and Edge/Fog Computing #Networking and Internet Architecture (cs.NI) #Software-Defined Networks and 5G
paper · pdf · doi:10.48550/arxiv.2312.17179
openalex publication_date 2023/12/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Network slicing is essential for transforming future telecommunication networks into versatile service platforms, but it also presents challenges for sustainable network operations. While meeting the requirements of network slices incurs additional energy consumption compared to non-sliced networks, operators strive to offer diverse 5G and beyond services while maintaining energy efficiency. In this study, we address the issue of slice activation/deactivation to reduce energy consumption while maintaining the user quality of service (QoS). We employ Deep Contextual Multi-Armed Bandit and Thompson Sampling Contextual Multi-Armed Bandit agents to make activation/deactivation decisions for individual clusters. Evaluations are performed using the NetMob23 dataset, which captures the spatio-temporal consumption of various mobile services in France. Our simulation results demonstrate that our proposed solutions provide significant reductions in network energy consumption while ensuring the QoS remains at a similar level compared to a scenario where all slice instances are active.