Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Learning-based Incentive Mechanism for Mobile AIGC Service in Decentralized Internet of Vehicles

Jiani Fan, Minrui Xu, Ziyao Liu, Huanyi Ye, Chaojie Gu, Dusit Niyato, Kwok-Yan Lam

TL;DR

The paper addresses resource-limited AIGC service provisioning in latency-sensitive IoV environments by designing a decentralized incentive mechanism that couples a double-sided auction at each RSU with multi-agent deep reinforcement learning. It introduces a McAfee-based allocation complemented by MAPPO-driven bidding agents to balance supply from virtual machines and demand from IoVs, optimizing global social welfare $SW(t)$ while minimizing latency $L(t)$. Key contributions include the market design for continuous AIGC service allocation, the MADRL-based mechanism, and empirical evaluation showing improved rewards and stable performance over baselines under varying IoV loads. The work offers a scalable, privacy-preserving approach for real-time, locally hosted AIGC services in decentralized IoV networks, reducing transmission latency and enhancing user experience.

Abstract

Artificial Intelligence-Generated Content (AIGC) refers to the paradigm of automated content generation utilizing AI models. Mobile AIGC services in the Internet of Vehicles (IoV) network have numerous advantages over traditional cloud-based AIGC services, including enhanced network efficiency, better reconfigurability, and stronger data security and privacy. Nonetheless, AIGC service provisioning frequently demands significant resources. Consequently, resource-constrained roadside units (RSUs) face challenges in maintaining a heterogeneous pool of AIGC services and addressing all user service requests without degrading overall performance. Therefore, in this paper, we propose a decentralized incentive mechanism for mobile AIGC service allocation, employing multi-agent deep reinforcement learning to find the balance between the supply of AIGC services on RSUs and user demand for services within the IoV context, optimizing user experience and minimizing transmission latency. Experimental results demonstrate that our approach achieves superior performance compared to other baseline models.

A Learning-based Incentive Mechanism for Mobile AIGC Service in Decentralized Internet of Vehicles

TL;DR

The paper addresses resource-limited AIGC service provisioning in latency-sensitive IoV environments by designing a decentralized incentive mechanism that couples a double-sided auction at each RSU with multi-agent deep reinforcement learning. It introduces a McAfee-based allocation complemented by MAPPO-driven bidding agents to balance supply from virtual machines and demand from IoVs, optimizing global social welfare while minimizing latency . Key contributions include the market design for continuous AIGC service allocation, the MADRL-based mechanism, and empirical evaluation showing improved rewards and stable performance over baselines under varying IoV loads. The work offers a scalable, privacy-preserving approach for real-time, locally hosted AIGC services in decentralized IoV networks, reducing transmission latency and enhancing user experience.

Abstract

Artificial Intelligence-Generated Content (AIGC) refers to the paradigm of automated content generation utilizing AI models. Mobile AIGC services in the Internet of Vehicles (IoV) network have numerous advantages over traditional cloud-based AIGC services, including enhanced network efficiency, better reconfigurability, and stronger data security and privacy. Nonetheless, AIGC service provisioning frequently demands significant resources. Consequently, resource-constrained roadside units (RSUs) face challenges in maintaining a heterogeneous pool of AIGC services and addressing all user service requests without degrading overall performance. Therefore, in this paper, we propose a decentralized incentive mechanism for mobile AIGC service allocation, employing multi-agent deep reinforcement learning to find the balance between the supply of AIGC services on RSUs and user demand for services within the IoV context, optimizing user experience and minimizing transmission latency. Experimental results demonstrate that our approach achieves superior performance compared to other baseline models.
Paper Structure (16 sections, 6 equations, 5 figures)

This paper contains 16 sections, 6 equations, 5 figures.

Table of Contents

  1. Introduction
  2. System Overview and Problem Definition
  3. Market Design
  4. RSUs
  5. Virtual Machines
  6. IoVs
  7. Problem Definition
  8. MADRL-based Service Allocation Mechanism
  9. Mechanism Design
  10. Multi-agent Reinforcement Learning-based Mechanism
  11. Observation
  12. Action
  13. Reward
  14. Valuation Function
  15. Evaluation Results
  16. ...and 1 more sections

Figures (5)

  • Figure 1: IoVs competing for local AIGC services provided at the nearest RSU.
  • Figure 2: Reward versus number of IoVs.
  • Figure 3: Social welfare versus number of IoVs.
  • Figure 4: Budget cost versus number of IoVs.
  • Figure 5: Transmission latency (ms) versus number of IoVs.