Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Online Learning for Autonomous Management of Intent-based 6G Networks

Erciyes Karakaya, Ozgur Ercetin, Huseyin Ozkan, Mehmet Karaca, Elham Dehghan Biyar, Alexandros Palaios

TL;DR

The paper tackles conflict-prone, intent-based management in autonomous 6G networks by proposing a hierarchical online learning framework based on multi-armed bandits. Child CLs independently target specific KPI intents, while a parent CL aggregates their local estimates to resolve cross-service conflicts using a Fed2-UCB strategy, enabling adaptive resource allocation under dynamic KPI targets and bandwidth limits. The authors formulate the problem with an objective to maximize the cumulative reward over a horizon while minimizing pseudo-regret $R(T)$ and present Algorithms 1 (Child CL) and 2 (MABCR) to realize the hierarchy. Through a Python-based emulator with URLLC and video services, they show that the MABCR approach outperforms greedy and Thompson sampling in SLA adherence and PRB efficiency, benefiting from the stabilizing influence of the parent evaluator. The work demonstrates a practical pathway to scalable, autonomous management of intent-based networks in 6G, with potential for real-world deployment and further exploration of network configurations.

Abstract

The growing complexity of networks and the variety of future scenarios with diverse and often stringent performance requirements call for a higher level of automation. Intent-based management emerges as a solution to attain high level of automation, enabling human operators to solely communicate with the network through high-level intents. The intents consist of the targets in the form of expectations (i.e., latency expectation) from a service and based on the expectations the required network configurations should be done accordingly. It is almost inevitable that when a network action is taken to fulfill one intent, it can cause negative impacts on the performance of another intent, which results in a conflict. In this paper, we aim to address the conflict issue and autonomous management of intent-based networking, and propose an online learning method based on the hierarchical multi-armed bandits approach for an effective management. Thanks to this hierarchical structure, it performs an efficient exploration and exploitation of network configurations with respect to the dynamic network conditions. We show that our algorithm is an effective approach regarding resource allocation and satisfaction of intent expectations.

Online Learning for Autonomous Management of Intent-based 6G Networks

TL;DR

The paper tackles conflict-prone, intent-based management in autonomous 6G networks by proposing a hierarchical online learning framework based on multi-armed bandits. Child CLs independently target specific KPI intents, while a parent CL aggregates their local estimates to resolve cross-service conflicts using a Fed2-UCB strategy, enabling adaptive resource allocation under dynamic KPI targets and bandwidth limits. The authors formulate the problem with an objective to maximize the cumulative reward over a horizon while minimizing pseudo-regret and present Algorithms 1 (Child CL) and 2 (MABCR) to realize the hierarchy. Through a Python-based emulator with URLLC and video services, they show that the MABCR approach outperforms greedy and Thompson sampling in SLA adherence and PRB efficiency, benefiting from the stabilizing influence of the parent evaluator. The work demonstrates a practical pathway to scalable, autonomous management of intent-based networks in 6G, with potential for real-world deployment and further exploration of network configurations.

Abstract

The growing complexity of networks and the variety of future scenarios with diverse and often stringent performance requirements call for a higher level of automation. Intent-based management emerges as a solution to attain high level of automation, enabling human operators to solely communicate with the network through high-level intents. The intents consist of the targets in the form of expectations (i.e., latency expectation) from a service and based on the expectations the required network configurations should be done accordingly. It is almost inevitable that when a network action is taken to fulfill one intent, it can cause negative impacts on the performance of another intent, which results in a conflict. In this paper, we aim to address the conflict issue and autonomous management of intent-based networking, and propose an online learning method based on the hierarchical multi-armed bandits approach for an effective management. Thanks to this hierarchical structure, it performs an efficient exploration and exploitation of network configurations with respect to the dynamic network conditions. We show that our algorithm is an effective approach regarding resource allocation and satisfaction of intent expectations.
Paper Structure (10 sections, 3 equations, 4 figures, 4 tables, 2 algorithms)

This paper contains 10 sections, 3 equations, 4 figures, 4 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Problem Formulation
  3. Proposed Method
  4. Algorithm for Child CL
  5. Multi arm bandit based Conflict Resolution (MABCR)
  6. Numerical Analyses
  7. Simulation Environment
  8. Simulation Scenarios
  9. Results
  10. Conclusions

Figures (4)

  • Figure 1: General structure of our approach for two service system case.
  • Figure 2: Each plot above shows the results for scenario given Table-I with confidence interval 95$\%$ for the Scenario-I.
  • Figure 3: Each plot above shows the results for scenario given Table-II with confidence interval 95$\%$ for the Scenario-II
  • Figure 4: Each plot above shows the results for scenario given Table-III with confidence interval 95$\%$ for the Scenario-III.