Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Modular DTaaS Architecture for Predictive Slice Management in 6G Systems

Tuğçe Bilen, Mehmet Özdem

TL;DR

The paper introduces Digital Twin as a Service (DTaaS) for predictive, SLA-centric slice management in 6G networks. It defines per-slice digital twins (SDTs) that ingest multi-domain telemetry and use predictive models to forecast traffic and SLA risks, enabling proactive provisioning and closed-loop control at the edge. The architecture comprises telemetry, twin intelligence, and orchestration layers, with modular, API-driven twin services and a twin-driven provisioning agent. Evaluation in a Python SimPy-based simulator shows DTaaS significantly improves SLA compliance, reduces over-provisioning, and lowers end-to-end latency compared with reactive and centralized baselines, suggesting scalable benefits for 6G orchestration.

Abstract

The sixth generation (6G) of wireless networks will require fundamentally new orchestration paradigms to meet stringent requirements for ultra-low latency, high reliability, and pervasive intelligence. Network slicing emerges as a key enabler to support diverse services with customized quality-of-service (QoS) guarantees. However, dynamic and fine-grained slice management poses significant challenges in terms of real-time provisioning, SLA assurance, and cross-layer observability. In this paper, we propose a novel Digital Twin as a Service (DTaaS) framework that embeds per-slice digital twins (SDTs) into the orchestration loop. Each SDT maintains a synchronized, real-time representation of its slice, leveraging multi-domain telemetry and deep sequential models to predict traffic evolution and SLA risks. The framework introduces modular intelligence layers, programmable interfaces, and edge-embedded decision-making to enable proactive provisioning, adaptive scaling, and closed-loop SLA assurance. Mathematical formulations for fidelity measurement, predictive control, and optimization objectives are provided to ensure rigor and transparency. Evaluation results demonstrate that DTaaS significantly improves SLA compliance ratio, reduces resource over-provisioning, and lowers average SLA violation probability, offering a scalable and reliable orchestration approach for 6G networks.

A Modular DTaaS Architecture for Predictive Slice Management in 6G Systems

TL;DR

The paper introduces Digital Twin as a Service (DTaaS) for predictive, SLA-centric slice management in 6G networks. It defines per-slice digital twins (SDTs) that ingest multi-domain telemetry and use predictive models to forecast traffic and SLA risks, enabling proactive provisioning and closed-loop control at the edge. The architecture comprises telemetry, twin intelligence, and orchestration layers, with modular, API-driven twin services and a twin-driven provisioning agent. Evaluation in a Python SimPy-based simulator shows DTaaS significantly improves SLA compliance, reduces over-provisioning, and lowers end-to-end latency compared with reactive and centralized baselines, suggesting scalable benefits for 6G orchestration.

Abstract

The sixth generation (6G) of wireless networks will require fundamentally new orchestration paradigms to meet stringent requirements for ultra-low latency, high reliability, and pervasive intelligence. Network slicing emerges as a key enabler to support diverse services with customized quality-of-service (QoS) guarantees. However, dynamic and fine-grained slice management poses significant challenges in terms of real-time provisioning, SLA assurance, and cross-layer observability. In this paper, we propose a novel Digital Twin as a Service (DTaaS) framework that embeds per-slice digital twins (SDTs) into the orchestration loop. Each SDT maintains a synchronized, real-time representation of its slice, leveraging multi-domain telemetry and deep sequential models to predict traffic evolution and SLA risks. The framework introduces modular intelligence layers, programmable interfaces, and edge-embedded decision-making to enable proactive provisioning, adaptive scaling, and closed-loop SLA assurance. Mathematical formulations for fidelity measurement, predictive control, and optimization objectives are provided to ensure rigor and transparency. Evaluation results demonstrate that DTaaS significantly improves SLA compliance ratio, reduces resource over-provisioning, and lowers average SLA violation probability, offering a scalable and reliable orchestration approach for 6G networks.

Paper Structure

This paper contains 19 sections, 6 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. The Proposed DTaaS Architecture
  4. Design Aspects
  5. System Overview
  6. DTaaS-Based Slice Management Model
  7. Slice Lifecycle with Digital Twins
  8. Twin-Driven Provisioning and Scaling
  9. Adaptive Reconfiguration and Reallocation
  10. Closed-Loop SLA Assurance
  11. Performance Evaluation
  12. Simulation Environment
  13. Baseline Methods
  14. Evaluation Metrics and Parameters
  15. Simulation Reslts
  16. ...and 4 more sections

Figures (2)

  • Figure 1: The proposed modular DTaaS architecture.
  • Figure 2: Evaluation results.