DaringFed: A Dynamic Bayesian Persuasion Pricing for Online Federated Learning under Two-sided Incomplete Information

TL;DR

This work tackles incentive design for Online Federated Learning under dynamic Two-sided Incomplete Information (TII) by proposing DaringFed, which blends Bayesian persuasion signaling of server-provided communication resources with a dynamic pricing rule for rewards. It models the server-client interaction as a Bayesian persuasion game, proves the existence (and uniqueness) of a Bayesian Persuasion Nash Equilibrium, and develops an approximate DaringFed mechanism under TII with provable bounds, using an upper confidence bound strategy to estimate client resource distributions. Theoretical contributions include BPNE existence and a near-optimal design under TII with a bound of $2\xi$, along with Algorithm 1 for discrete signaling and pricing. Empirically, DaringFed improves OFL performance on real datasets by up to $16.99\%$ in accuracy and up to $12.6\%$ in server utility on synthetic data, validating its practical potential for real-time, resource-constrained FL deployments.

Abstract

Online Federated Learning (OFL) is a real-time learning paradigm that sequentially executes parameter aggregation immediately for each random arriving client. To motivate clients to participate in OFL, it is crucial to offer appropriate incentives to offset the training resource consumption. However, the design of incentive mechanisms in OFL is constrained by the dynamic variability of Two-sided Incomplete Information (TII) concerning resources, where the server is unaware of the clients' dynamically changing computational resources, while clients lack knowledge of the real-time communication resources allocated by the server. To incentivize clients to participate in training by offering dynamic rewards to each arriving client, we design a novel Dynamic Bayesian persuasion pricing for online Federated learning (DaringFed) under TII. Specifically, we begin by formulating the interaction between the server and clients as a dynamic signaling and pricing allocation problem within a Bayesian persuasion game, and then demonstrate the existence of a unique Bayesian persuasion Nash equilibrium. By deriving the optimal design of DaringFed under one-sided incomplete information, we further analyze the approximate optimal design of DaringFed with a specific bound under TII. Finally, extensive evaluation conducted on real datasets demonstrate that DaringFed optimizes accuracy and converges speed by 16.99%, while experiments with synthetic datasets validate the convergence of estimate unknown values and the effectiveness of DaringFed in improving the server's utility by up to 12.6%.

Figures (4)

• Figure 1: An overview of OFL.
• Figure 2: Testing the accuracy of OFL with the proposed DaringFed on (a) MNIST, (b) Fashion-MNIST, (c) FEMNIST, and (d) CIFAR-10.
• Figure 3: Convergence on (a) $\hat{\theta}$ and (b) $\gamma$.
• Figure 4: Improvements on (a) $\hat{\theta}$ and (b) $c_s$.

Theorems & Definitions (10)

• Definition 1
• Definition 2
• Definition 3
• Definition 4
• Definition 5
• Lemma 1
• Theorem 1
• Lemma 2
• Theorem 2
• Theorem 3