Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI Model Placement for 6G Networks under Epistemic Uncertainty Estimation

Liming Huang, Yulei Wu, Juan Marcelo Parra-Ullauri, Reza Nejabati, Dimitra Simeonidou

TL;DR

This work tackles AI model placement for 6G NFV under epistemic uncertainty arising from varying compute, storage, and performance of AI models. It introduces an uncertainty-aware sequence-to-sequence placement model that integrates an Orthonormal Certificate (OC) uncertainty estimator and Gaussian fuzzy representations, trained via neural combinatorial optimization in a reinforcement learning framework. The approach outputs placements for each AI model in a service chain while minimizing energy and satisfying resource, bandwidth, latency, and SLA constraints. Empirical results show strong NS request acceptance across diverse service chains and network sizes, highlighting the method's robustness and practical potential for deploying AI-driven VNFs in future networks.

Abstract

The adoption of Artificial Intelligence (AI) based Virtual Network Functions (VNFs) has witnessed significant growth, posing a critical challenge in orchestrating AI models within next-generation 6G networks. Finding optimal AI model placement is significantly more challenging than placing traditional software-based VNFs, due to the introduction of numerous uncertain factors by AI models, such as varying computing resource consumption, dynamic storage requirements, and changing model performance. To address the AI model placement problem under uncertainties, this paper presents a novel approach employing a sequence-to-sequence (S2S) neural network which considers uncertainty estimations. The S2S model, characterized by its encoding-decoding architecture, is designed to take the service chain with a number of AI models as input and produce the corresponding placement of each AI model. To address the introduced uncertainties, our methodology incorporates the orthonormal certificate module for uncertainty estimation and utilizes fuzzy logic for uncertainty representation, thereby enhancing the capabilities of the S2S model. Experiments demonstrate that the proposed method achieves competitive results across diverse AI model profiles, network environments, and service chain requests.

AI Model Placement for 6G Networks under Epistemic Uncertainty Estimation

TL;DR

This work tackles AI model placement for 6G NFV under epistemic uncertainty arising from varying compute, storage, and performance of AI models. It introduces an uncertainty-aware sequence-to-sequence placement model that integrates an Orthonormal Certificate (OC) uncertainty estimator and Gaussian fuzzy representations, trained via neural combinatorial optimization in a reinforcement learning framework. The approach outputs placements for each AI model in a service chain while minimizing energy and satisfying resource, bandwidth, latency, and SLA constraints. Empirical results show strong NS request acceptance across diverse service chains and network sizes, highlighting the method's robustness and practical potential for deploying AI-driven VNFs in future networks.

Abstract

The adoption of Artificial Intelligence (AI) based Virtual Network Functions (VNFs) has witnessed significant growth, posing a critical challenge in orchestrating AI models within next-generation 6G networks. Finding optimal AI model placement is significantly more challenging than placing traditional software-based VNFs, due to the introduction of numerous uncertain factors by AI models, such as varying computing resource consumption, dynamic storage requirements, and changing model performance. To address the AI model placement problem under uncertainties, this paper presents a novel approach employing a sequence-to-sequence (S2S) neural network which considers uncertainty estimations. The S2S model, characterized by its encoding-decoding architecture, is designed to take the service chain with a number of AI models as input and produce the corresponding placement of each AI model. To address the introduced uncertainties, our methodology incorporates the orthonormal certificate module for uncertainty estimation and utilizes fuzzy logic for uncertainty representation, thereby enhancing the capabilities of the S2S model. Experiments demonstrate that the proposed method achieves competitive results across diverse AI model profiles, network environments, and service chain requests.
Paper Structure (20 sections, 8 equations, 3 figures, 4 tables)

This paper contains 20 sections, 8 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. VNF Placement
  4. Uncertainty Learning
  5. Problem Formulation
  6. Problem Statement
  7. Mathematical Formulation
  8. Optimizing AI Model Placement with Uncertainty Estimation
  9. Sequence-to-Sequence Model for Placement
  10. Uncertainty Estimation and Combination
  11. Uncertainty Estimation
  12. Fuzzy Representation for Uncertainties
  13. Uncertainty Combination
  14. Neural Combinatorial Optimization
  15. Evaluation
  16. ...and 5 more sections

Figures (3)

  • Figure 1: The framework of our AI placement model.
  • Figure 2: The learning framework for the S2S model with RL environment.
  • Figure 3: Comparisons with NCO 8945291, Gecode schulte2006gecode and FF kumaraswamy2019bin with 128 NS requests in (a) 10 hosts and (b) 20 hosts.