Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Decentralized Orchestration Architecture for Fluid Computing: A Secure Distributed AI Use Case

Diego Cajaraville-Aboy, Ana Fernández-Vilas, Rebeca P. Díaz-Redondo, Manuel Fernández-Veiga, Pablo Picallo-López

Abstract

Distributed AI and IoT applications increasingly execute across heterogeneous resources spanning end devices, edge/fog infrastructure, and cloud platforms, often under different administrative domains. Fluid Computing has emerged as a promising paradigm for enhancing massive resource management across the computing continuum by treating such resources as a unified fabric, enabling optimal service-agnostic deployments driven by application requirements. However, existing solutions remain largely centralized and often do not explicitly address multi-domain considerations. This paper proposes an agnostic multi-domain orchestration architecture for fluid computing environments. The orchestration plane enables decentralized coordination among domains that maintain local autonomy while jointly realizing intent-based deployment requests from tenants, ensuring end-to-end placement and execution. To this end, the architecture elevates domain-side control services as first-class capabilities to support application-level enhancement at runtime. As a representative use case, we consider a multi-domain Decentralized Federated Learning (DFL) deployment under Byzantine threats. We leverage domain-side capabilities to enhance Byzantine security by introducing FU-HST, an SDN-enabled multi-domain anomaly detection mechanism that complements Byzantine-robust aggregation. We validate the approach via simulation in single- and multi-domain settings, evaluating anomaly detection, DFL performance, and computation/communication overhead.

Decentralized Orchestration Architecture for Fluid Computing: A Secure Distributed AI Use Case

Abstract

Distributed AI and IoT applications increasingly execute across heterogeneous resources spanning end devices, edge/fog infrastructure, and cloud platforms, often under different administrative domains. Fluid Computing has emerged as a promising paradigm for enhancing massive resource management across the computing continuum by treating such resources as a unified fabric, enabling optimal service-agnostic deployments driven by application requirements. However, existing solutions remain largely centralized and often do not explicitly address multi-domain considerations. This paper proposes an agnostic multi-domain orchestration architecture for fluid computing environments. The orchestration plane enables decentralized coordination among domains that maintain local autonomy while jointly realizing intent-based deployment requests from tenants, ensuring end-to-end placement and execution. To this end, the architecture elevates domain-side control services as first-class capabilities to support application-level enhancement at runtime. As a representative use case, we consider a multi-domain Decentralized Federated Learning (DFL) deployment under Byzantine threats. We leverage domain-side capabilities to enhance Byzantine security by introducing FU-HST, an SDN-enabled multi-domain anomaly detection mechanism that complements Byzantine-robust aggregation. We validate the approach via simulation in single- and multi-domain settings, evaluating anomaly detection, DFL performance, and computation/communication overhead.
Paper Structure (27 sections, 17 equations, 9 figures, 1 table, 1 algorithm)

This paper contains 27 sections, 17 equations, 9 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Orchestration in the Cloud–Edge Continuum
  4. Security in (Decentralized) Federated Learning
  5. System Model
  6. Agnostic Decentralized Orchestration Architecture for Fluid Computing
  7. Architecture Overview
  8. Framework Workflow
  9. Phase I -- Placement Planning
  10. Phase II -- Job Scheduling
  11. Phase III -- Fluid Runtime Execution
  12. Decentralized and Multi-Domain Orchestration Plane
  13. Use Case of Security Enhancement in Multi-Domain Decentralized Federated Learning
  14. From Proposed Architecture to Use Case
  15. System Setting
  16. ...and 12 more sections

Figures (9)

  • Figure 1: Schematic representation of Fluid Computing paradigm as a unified platform of computing and communication resources that subsumes the main computing paradigms (Mist, Edge, Fog and Cloud). The left-to-right arrow at the bottom indicates increasing delay as data and tasks move toward the cloud, while the right-to-left arrow indicates decreasing computing capabilities as tasks move closer to end-user devices.
  • Figure 2: Agnostic multi-domain orchestration architecture for multi-tenant Fluid Computing environments, enabling decentralized coordination across administrative domains.
  • Figure 3: Considered scenario for multi-domain DFL deployment with SDN-enabled security-enhancement mechanism
  • Figure 4: Overview of the SDN-enabled anomaly detection workflow (FU-HST algorithm)
  • Figure 5: Comparison of the global F1 and Accuracy among all rounds in different DFL topologies with 3 malicious nodes and a range of 20 to 100 nodes.
  • ...and 4 more figures