Freshness-aware Incentive Mechanism for Mobile AI-Generated Content (AIGC) Networks
Jinbo Wen, Jiawen Kang, Minrui Xu, Hongyang Du, Zehui Xiong, Yang Zhang, Dusit Niyato
Abstract
Artificial Intelligence-Generated Content (AIGC) is a rapidly growing field that uses advanced AI algorithms to automatically generate, manipulate, and modify data to create diverse and valuable content. Recently, mobile AIGC networks have attracted significant attention. They integrate with mobile edge networks to provide customized and personalized AIGC services in real time. This is achieved by deploying AI algorithms on edge devices, particularly on Unmanned Aerial Vehicles (UAVs) with edge servers in a flexible and dynamic manner for ubiquitous edge intelligence. However, the wide deployment of UAV-enabled AIGC still faces critical challenges, such as issues with data freshness for AIGC fine-tuning and concerns around incentivizing fresh data sharing among UAVs under information asymmetry. In this paper, we first utilize Age of Information (AoI) as a well-accepted data-freshness metric to quantify data freshness for AIGC fine-tuning. Then, we propose an AoI-based contract theory model to incentivize the contribution of fresh data among UAVs. Moreover, we design the optimal contract that is feasible to maximize the expected utility of the base station that is responsible for dispatching UAVs to collaboratively perform AIGC tasks. Finally, numerical results demonstrate the effectiveness of the proposed scheme in mobile AIGC networks.