Incentive-Aware Federated Averaging with Performance Guarantees under Strategic Participation
For federated learning systems, this work addresses the problem of strategic client participation by providing a method with performance guarantees, though the approach is incremental.
The paper proposes an incentive-aware federated averaging method where clients dynamically adjust their dataset sizes via a Nash equilibrium-seeking update rule to balance learning payoff and data contribution cost. Experiments on MNIST and CIFAR-10 show competitive global model performance with stable participation strategies.
Federated learning (FL) is a communication-efficient collaborative learning framework that enables model training across multiple agents with private local datasets. While the benefits of FL in improving global model performance are well established, individual agents may behave strategically, balancing the learning payoff against the cost of contributing their local data. Motivated by the need for FL frameworks that successfully retain participating agents, we propose an incentive-aware federated averaging method in which, at each communication round, clients transmit both their local model parameters and their updated training dataset sizes to the server. The dataset sizes are dynamically adjusted via a Nash equilibrium (NE)-seeking update rule that captures strategic data participation. We analyze the proposed method under convex and nonconvex global objective settings and establish performance guarantees for the resulting incentive-aware FL algorithm. Numerical experiments on the MNIST and CIFAR-10 datasets demonstrate that agents achieve competitive global model performance while converging to stable data participation strategies.