A Critical Learning Period-Aware Incentive Mechanism for Federated Learning

Critical learning periods (CLPs) in federated learning (FL) represent early stages where low-quality contributions (e.g., sparse training data availability) can permanently impair learning outcomes. Yet, strategies to motivate clients with highquality contributions to join the model training process and share model updates during CLPs remain underexplored. Additionally, existing incentive mechanisms in FL treat all training periods equally, which consequently fails to motivate clients to participate early. In this paper, we propose a time-aware incentive mechanism, called R3T, to encourage client involvement, especially during CLPs in FL. We first characterize the cloud utility as a function of client's time and system capabilities, effort, joining time, and reward. Then, we analytically derive the optimal contract and devise a CLP-aware mechanism to incentivize early participation and efforts while maximizing cloud utility, even under information asymmetry. By providing the right reward at the right time, R3T can attract the highest-quality contributions during CLPs. Simulation studies show that R3T increases cloud utility and is more economically effective than benchmarks.