Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Novel Hierarchical Co-Optimization Framework for Coordinated Task Scheduling and Power Dispatch in Computing Power Networks

Haoxiang Luo, Kun Yang, Qi Huang, Marco Aiello, Schahram Dustdar

TL;DR

The paper tackles the challenge of sustainable, reliable service computing in Computing Power Networks (CPN) amid high renewable energy penetration. It introduces the Two-Stage Co-Optimization (TSCO) framework, combining a day-ahead stochastic unit commitment stage solved via Benders decomposition with a real-time ED coupled to a DRL-based CPN scheduler. The integrated model accounts for CPN task DAGs, heterogeneous hardware power, grid constraints, RES uncertainty, and endogenous carbon intensity, achieving notable reductions in carbon emissions (≈16%), operating costs (≈13%), and RES curtailment (>60%) while maintaining high task success (≈98.5%) and low tardiness (≈12.3s). The results demonstrate the practical viability of cross-domain, model-based planning with AI-driven real-time control for low-carbon, efficient service computing.

Abstract

The proliferation of large-scale AI and data-intensive applications has driven the development of Computing Power Networks (CPN). It is a key paradigm for delivering ubiquitous, on-demand computational services with high efficiency. However, CPNs face dual challenges in service computing. Immense energy consumption threatens sustainable operations. And the integration with power grids also features high penetration of intermittent Renewable Energy Sources (RES), complicating task scheduling while ensuring Quality of Service (QoS). To address these issues, this paper proposes a novel Two-Stage Co-Optimization (TSCO) framework. It synergistically coordinates CPN task scheduling and power system dispatch, aiming to optimize service performance while achieving low-carbon operations. The framework decomposes the complex, large-scale problem into a day-ahead stochastic unit commitment stage and a real-time operational stage. The former is solved using Benders decomposition for computational tractability, while in the latter, economic dispatch of generation assets is coupled with an adaptive CPN task scheduling managed by a deep reinforcement learning agent. It makes carbon-aware decisions by responding to dynamic grid conditions, including real-time electricity prices and marginal carbon intensity. Extensive simulations demonstrate that the TSCO outperforms baseline approaches significantly. It reduces carbon emissions by 16.2% and operational costs by 12.7%, while decreasing RES curtailment by over $60\%$, maintaining a task success rate of 98.5%, and minimizing average task tardiness to 12.3s. This work advances cross-domain service optimization in CPNs.

A Novel Hierarchical Co-Optimization Framework for Coordinated Task Scheduling and Power Dispatch in Computing Power Networks

TL;DR

The paper tackles the challenge of sustainable, reliable service computing in Computing Power Networks (CPN) amid high renewable energy penetration. It introduces the Two-Stage Co-Optimization (TSCO) framework, combining a day-ahead stochastic unit commitment stage solved via Benders decomposition with a real-time ED coupled to a DRL-based CPN scheduler. The integrated model accounts for CPN task DAGs, heterogeneous hardware power, grid constraints, RES uncertainty, and endogenous carbon intensity, achieving notable reductions in carbon emissions (≈16%), operating costs (≈13%), and RES curtailment (>60%) while maintaining high task success (≈98.5%) and low tardiness (≈12.3s). The results demonstrate the practical viability of cross-domain, model-based planning with AI-driven real-time control for low-carbon, efficient service computing.

Abstract

The proliferation of large-scale AI and data-intensive applications has driven the development of Computing Power Networks (CPN). It is a key paradigm for delivering ubiquitous, on-demand computational services with high efficiency. However, CPNs face dual challenges in service computing. Immense energy consumption threatens sustainable operations. And the integration with power grids also features high penetration of intermittent Renewable Energy Sources (RES), complicating task scheduling while ensuring Quality of Service (QoS). To address these issues, this paper proposes a novel Two-Stage Co-Optimization (TSCO) framework. It synergistically coordinates CPN task scheduling and power system dispatch, aiming to optimize service performance while achieving low-carbon operations. The framework decomposes the complex, large-scale problem into a day-ahead stochastic unit commitment stage and a real-time operational stage. The former is solved using Benders decomposition for computational tractability, while in the latter, economic dispatch of generation assets is coupled with an adaptive CPN task scheduling managed by a deep reinforcement learning agent. It makes carbon-aware decisions by responding to dynamic grid conditions, including real-time electricity prices and marginal carbon intensity. Extensive simulations demonstrate that the TSCO outperforms baseline approaches significantly. It reduces carbon emissions by 16.2% and operational costs by 12.7%, while decreasing RES curtailment by over , maintaining a task success rate of 98.5%, and minimizing average task tardiness to 12.3s. This work advances cross-domain service optimization in CPNs.

Paper Structure

This paper contains 41 sections, 8 equations, 5 figures, 4 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Background
  3. Research Motivation
  4. Contributions of the Paper
  5. Paper Structure
  6. Related Works
  7. Computing Power Network Architectures and Scheduling
  8. Integrated Energy System Management
  9. Carbon-Aware Workload Scheduling
  10. Research Gap Summary
  11. System Architecture and Problem Formulation
  12. Computing Power Network Model
  13. CPN Node Model
  14. Task Model
  15. Communication Link Model
  16. ...and 26 more sections

Figures (5)

  • Figure 1: CPN and power grid co-optimization architecture.
  • Figure 2: Two-Stage Co-Optimization (TSCO) framework for CPN and power grid collaborative optimization.
  • Figure 3: Baseline performance comparison. (a) Total operational cost; (b) Total carbon emissions; (c) RES curtailment; (d) CPN job success rate; (d) Average job tardiness.
  • Figure 4: Sensitivity analysis with varying carbon price. (a) Total operational cost; (b) Total carbon emissions; (c) RES curtailment; (d) CPN job success rate; (d) Average job tardiness.
  • Figure 5: Ablation test. (a) Total operational cost; (b) Total carbon emissions; (c) RES curtailment; (d) CPN job success rate; (d) Average job tardiness.