Strategic Resource Optimization in 6G Multi-Tenant Slicing Networks Using Deep Q-Learning

International Journal of Computer Science and Technology | Vol. 4, No. 6, June 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000221

Article Type: Original Research Paper

Title: Strategic Resource Optimization in 6G Multi-Tenant Slicing Networks Using Deep Q-Learning

Names of Authors: Chinedu Aliyu¹, Elena Rostova²

Authors’ Affiliations: ¹Department of Telecommunications Engineering, University of Nigeria, Nsukka, Nigeria; ²Network Systems Research Lab, Helsinki Institute of Technology, Helsinki, Finland

Abstract: The transition toward 6G mobile infrastructures requires the simultaneous management of highly distinct network slices tailored for ultra-reliable low-latency systems and massive machine-type deployments. Static network resource allocations result in severe underutilization or localized bandwidth exhaustion during peak periods. This paper presents a dynamic multi-tenant network slicing optimization model utilizing a Deep Q-Network (DQN) framework. The model tracks live performance telemetry parameters including instantaneous bitrates, queue sizes, and hardware utilization over a continuous time frame. It dynamically scales physical radio resources and virtual routing pathways among independent network operators based on localized demand variables. We simulated this network structure on an enterprise network platform with 20 distinct base stations and 1,000 mobile terminal nodes. The experimental results show a 34.7% improvement in total data throughout efficiency over standard static slicing models. Total network handoff latency was maintained below a strict boundary of 4.5 ms under high mobility stress conditions. A standard statistical evaluation confirmed the significance of this throughput improvement across testing runs (p < 0.01).

Keywords: 6G Infrastructure, Network Slicing, Deep Q-Learning, Resource Management, Multi-Tenant Systems, Telemetry

Manuscript Timeline: Received: February 10, 2023; Revised: April 02, 2023; Accepted: May 12, 2023; Published: June 06, 2023

Citation: Aliyu, C., & Rostova, E. (2023). Strategic Resource Optimization in 6G Multi-Tenant Slicing Networks Using Deep Q-Learning. International Journal of Computer Science and Technology, 4(6), 1–8. DOI: 10.46882/2023/IJCST/000221

Subscribe to read the full article: https://internationalscholarsjournals.org/subscribe-to-read