Satisfaction-Aware Incentive Scheme for Federated Learning in Industrial Metaverse: DRL-Based Stackbelberg Game Approach

2025-02-10

AI Summary

To address low participant willingness and the trade-off between model quality and training latency in federated learning (FL) for industrial metaverse applications, this paper proposes a satisfaction-aware two-stage Stackelberg game-based incentive mechanism. We innovatively design a node satisfaction function integrating data volume, information freshness (Age of Information, AoI), and training delay, embedded within a server–node utility framework. This work is the first to combine satisfaction awareness with Stackelberg game theory and further devises a dynamic incentive policy via deep reinforcement learning (DRL). Experimental results demonstrate that, under identical budget constraints, the proposed mechanism improves system utility by ≥23.7% without compromising model accuracy, thereby significantly enhancing the sustainability, practicality, and robustness of FL in heterogeneous industrial IoT (IIoT) environments.

Technology Category

• Multiagent Systems: Mechanism Design
• Machine Learning: Distributed Machine Learning & Federated Learning
• Game Theory and Economic Paradigms: Mechanism Design

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Industrial Metaverse leverages the Industrial Internet of Things (IIoT) to integrate data from diverse devices, employing federated learning and meta-computing to train models in a distributed manner while ensuring data privacy. Achieving an immersive experience for industrial Metaverse necessitates maintaining a balance between model quality and training latency. Consequently, a primary challenge in federated learning tasks is optimizing overall system performance by balancing model quality and training latency. This paper designs a satisfaction function that accounts for data size, Age of Information (AoI), and training latency. Additionally, the satisfaction function is incorporated into the utility functions to incentivize node participation in model training. We model the utility functions of servers and nodes as a two-stage Stackelberg game and employ a deep reinforcement learning approach to learn the Stackelberg equilibrium. This approach ensures balanced rewards and enhances the applicability of the incentive scheme for industrial Metaverse. Simulation results demonstrate that, under the same budget constraints, the proposed incentive scheme improves at least 23.7% utility compared to existing schemes without compromising model accuracy.

Problem

• Balancing model quality and training latency
• Optimizing federated learning system performance
• Enhancing node participation with incentive schemes

Innovation

• DRL-Based Stackelberg Game
• Satisfaction Function Optimization
• Industrial Metaverse Federated Learning